ISSN 0352-9045 Journal of Microelectronics, Electronic Components and Materials Vol. 53, No. 1(2023), March 2023 Revija za mikroelektroniko, elektronske sestavne dele in materiale letnik 53, številka 1(2023), Marec 2023 UDK 621.3:(53+54+621+66)(05)(497.1)=00 ISSN 0352-9045 Informacije MIDEM 1-2023 Journal of Microelectronics, Electronic Components and Materials VOLUME 53, NO. 1(185), LJUBLJANA, MARCH 2023 | LETNIK 53, NO. 1(185), LJUBLJANA, MAREC 2023 Published quarterly (March, June, September, December) by Society for Microelectronics, Electronic Components and Materials - MIDEM. Copyright © 2023. All rights reserved. | Revija izhaja trimesečno (marec, junij, september, december). Izdaja Strokovno društvo za mikroelektroniko, elektronske sestavne dele in materiale – Društvo MIDEM. Copyright © 2023. Vse pravice pridržane. Editor in Chief | Glavni in odgovorni urednik Marko Topič, University of Ljubljana (UL), Faculty of Electrical Engineering, Slovenia Editor of Electronic Edition | Urednik elektronske izdaje Kristijan Brecl, UL, Faculty of Electrical Engineering, Slovenia Associate Editors | Odgovorni področni uredniki Vanja Ambrožič, UL, Faculty of Electrical Engineering, Slovenia Arpad Bürmen, UL, Faculty of Electrical Engineering, Slovenia Danjela Kuščer Hrovatin, Jožef Stefan Institute, Slovenia Matija Pirc, UL, Faculty of Electrical Engineering, Slovenia Franc Smole, UL, Faculty of Electrical Engineering, Slovenia Matjaž Vidmar, UL, Faculty of Electrical Engineering, Slovenia Editorial Board | Uredniški odbor Mohamed Akil, ESIEE PARIS, France Giuseppe Buja, University of Padova, Italy Gian-Franco Dalla Betta, University of Trento, Italy Martyn Fice, University College London, United Kingdom Ciprian Iliescu, Institute of Bioengineering and Nanotechnology, A*STAR, Singapore Marc Lethiecq, University of Tours, France Teresa Orlowska-Kowalska, Wroclaw University of Technology, Poland Luca Palmieri, University of Padova, Italy Goran Stojanović, University of Novi Sad, Serbia International Advisory Board | Časopisni svet Janez Trontelj, UL, Faculty of Electrical Engineering, Slovenia - Chairman Cor Claeys, IMEC, Leuven, Belgium Denis Đonlagić, University of Maribor, Faculty of Elec. Eng. and Computer Science, Slovenia Zvonko Fazarinc, CIS, Stanford University, Stanford, USA Leszek J. Golonka, Technical University Wroclaw, Wroclaw, Poland Jean-Marie Haussonne, EIC-LUSAC, Octeville, France Barbara Malič, Jožef Stefan Institute, Slovenia Miran Mozetič, Jožef Stefan Institute, Slovenia Stane Pejovnik, UL, Faculty of Chemistry and Chemical Technology, Slovenia Giorgio Pignatel, University of Perugia, Italy Giovanni Soncini, University of Trento, Trento, Italy Iztok Šorli, MIKROIKS d.o.o., Ljubljana, Slovenia Hong Wang, Xi´an Jiaotong University, China Headquarters | Naslov uredništva Uredništvo Informacije MIDEM MIDEM pri MIKROIKS Stegne 11, 1521 Ljubljana, Slovenia T. +386 (0)1 513 37 68 F. + 386 (0)1 513 37 71 E. info@midem-drustvo.si www.midem-drustvo.si Annual subscription rate is 160 EUR, separate issue is 40 EUR. MIDEM members and Society sponsors receive current issues for free. Scientific Council for Technical Sciences of Slovenian Research Agency has recognized Informacije MIDEM as scientific Journal for microelectronics, electronic components and materials. Publishing of the Journal is cofi- nanced by Slovenian Research Agency and by Society sponsors. Scientific and professional papers published in the journal are indexed and abstracted in COBISS and INSPEC databases. The Journal is indexed by ISI® for Sci Search®, Research Alert® and Material Science Citation Index™. | Letna naročnina je 160 EUR, cena posamezne številke pa 40 EUR. Člani in sponzorji MIDEM prejemajo posamezne številke brezplačno. Znanstveni svet za tehnične vede je podal pozitivno mnenje o reviji kot znanstveno-strokovni reviji za mikroelektroniko, elektronske sestavne dele in materiale. Izdajo revije sofinancirajo ARRS in sponzorji društva. Znanstveno-strokovne prispevke objavljene v Informacijah MIDEM zajemamo v podatkovne baze COBISS in INSPEC. Prispevke iz revije zajema ISI® v naslednje svoje produkte: Sci Search®, Research Alert® in Materials Science Citation Index™. Design | Oblikovanje: Snežana Madić Lešnik; Printed by | tisk: Biro M, Ljubljana; Circulation | Naklada: 1000 issues | izvodov; Slovenia Taxe Percue | Poštnina plačana pri pošti 1102 Ljubljana Journal of Microelectronics, Electronic Components and Materials vol. 53, No. 1(2023) Content | Vsebina Original scientific papers Izvirni znanstveni članki N. Jothi, S. Krishnan, Lakshminarayanan: 3 N. Jothi, S. Krishnan, Lakshminarayanan: Design and Comparative Analysis of Inter Zasnova in primerjalna analiza medsatelitsko- Satellite- Optical Wireless Communication optičnih brezžičnih komunikacij (IS-OWC) za (IS-OWC) for Return to Zero (RZ) & Non-Return modulacijske formate povratek v ničlo (RZ) to Zero (NRZ) Modulation Formats Through in nepovratek v ničlo (NRZ) s tehniko Channel Diversity Technique kanalske raznolikosti I. Rajagopalan, J. Ponnuswamy: 15 I. Rajagopalan, J. Ponnuswamy: Loss Reduction and Reliability Improvement in Zmanjšanje izgub in izboljšanje zanesljivosti v Distributed Network Using HF-SOA Based porazdeljenem omrežju z uporabo optimalne Optimal Installation of DG, SCs and STF namestitve DG, SC in STF na podlagi HF-SOA I. El Gmati, R. Ghayoula: 31 I. El Gmati, R. Ghayoula: Liquid Metal Droplet Tunable RF MEMS Inductor Nastavljiva mikrofluidna RF MEMS tuljava P. Teotia, O. A. Shah: 39 P. Teotia, O. A. Shah: Power and Area Efficient Sense Amplifier Based Energetsko in prostorsko učinkovit senzorski Flip Flop with Wide Voltage and Temperature ojačevalnik na osnovi flip flopa s širokim Upholding for Portable IoT Applications razponom napetosti in temperature za prenosne aplikacije interneta stvari J. Wang, Y. Yan: 49 J. Wang, Y. Yan: The Design of Frequency-tunable Mechanical Načrtovanje mehanskega spojnika z možnostjo Tuning Coupler Based on Coupled nastavitve frekvence na podlagi strukture Line Structure sklopljene linije Front page: Naslovnica: uMPPT system for simultaneous monitoring of a uMPPT sistem za sočasno spremljanje delovanja large number of perovskite solar cells. velikega števila perovskitnih sončnih celic. (M. Jankovec, LPVO) (M. Jankovec, LPVO) 1 2 Original scientific paper https://doi.org/10.33180/InfMIDEM2023.101 Journal of Microelectronics, Electronic Components and Materials Vol. 53, No. 1(2023), 3 – 13 Design and comparative analysis of Inter Satellite- Optical Wireless Communication (IS-OWC) for Return to Zero (RZ) & Non-Return to Zero (NRZ) modulation formats through channel diversity technique Nirmal Jothi1, Santhana Krishnan2, Lakshminarayanan3 1Department of Electronics and Communication Engineering, Vel Tech Rangarajan, Dr.Sagunthala R&D Institute of science and technology, Tamilnadu, India 2Department of Electronics and Communication Engineering, SCAD College of Engineering and technology, Tirunelveli, Tamilnadu, India 3Department of Electronics and Communication Engineering, Francis Xavier Engineering College, Tirunelveli, Tamilnadu, India Abstract: Inter-Satellite Optical Wireless Communication (IS-OWC) is a novel strategy for establishing an inter-connection between two satellites. The IS-OWC is focused on the use of lasers rather than conventional radio and microwave structures for wireless optical communication. Optical wireless communication between satel-lites is being developed by integrating optical wireless communication technology and space technology. IS-OWC can connect satellites in the same orbit or different orbits. When compared to the single channel, the channel diversity strategy produces better results. The channel diversity techniques use the IS-OWC devices, in which several signal paths are available for allowing the Q factor and signal intensity, which is extended or en-larged over a large distance. In this paper, the Q factor and Bit Error Rate (BER) are reviewed with diverse mod-ulation designs for Return to Zero (RZ) and Non-Return to Zero (NRZ) using the channel diversity technique. The simulation is conducted on opti-system-16.0 software with 32768 numbers of samples with a bit rate of 109 bits/ sec. The Q factor attained a maximum in the 8-channel OWC is 19.4385, whereas the Q factor attained min-imum in the 2-channel OWC is 19.4385. Moreover, the number of channels increased may develop the profi-ciency of minimum power inter-satellite relation. Keywords: Optical Wireless Communication (OWC), BER, Q-factor, Return to Zero (RZ), Non-Return to Zero (NRZ). Zasnova in primerjalna analiza medsatelitsko-optičnih brezžičnih komunikacij (IS-OWC) za modulacijske formate povratek v ničlo (RZ) in nepovratek v ničlo (NRZ) s tehniko kanalske raznolikosti Izvleček: Medsatelitska optična brezžična komunikacija (IS-OWC) je nova strategija za vzpostavitev medsebojne povezave med dvema satelitoma. IS-OWC se osredotoča na uporabo laserjev namesto običajnih radijskih in mikrovalovnih struktur za brezžično optično komunikacijo. Optična brezžična komunikacija med sateliti se razvija s povezovanjem tehnologije optične brezžične komunikacije in vesoljske tehnologije. IS-OWC lahko povezuje satelite v isti orbiti ali različnih orbitah. V primerjavi z enim kanalom daje strategija raznolikosti kanalov boljše rezultate. Tehnike kanalske raznolikosti uporabljajo naprave IS-OWC, v katerih je na voljo več signalnih poti za omogočanje faktorja Q in intenzivnosti signala, ki se poveča na veliki razdalji. V tem članku sta raziskana faktor Q in stopnja napake v bitu (BER) z različnimi modulacijskimi zasnovami za vrnitev v nič (RZ) in brez vrnitve v nič (NRZ) z uporabo tehnike kanalske raznolikosti. Simulacija je izvedena s programsko opremo opti-system-16.0 s 32768 vzorci s hitrostjo prenosa 109 bitov na sekundo. Faktor Q, dosežen pri 8-kanalnem OWC, je 19,4385, medtem ko je faktor Q, dosežen pri 2-kanalnem OWC, 19,4385. Poleg tega se lahko s povečanjem števila kana-lov razvije spretnost najmanjše moči medsatelitskega razmerja. Ključne besede: Optične brezžične komunikacije (OWC), BER, faktor Q, povratek v ničlo (RZ), nepovratek v ničlo (NRZ). * Corresponding Author’s e-mail: drnj2023@gmail.com How to cite: N. Jothi et al., “Design and comparative analysis of Inter Satellite- Optical Wireless Communication (IS-OWC) for Return to Zero (RZ) & Non-Return to Zero (NRZ) modulation formats through channel diversity techniqu", Inf. Midem-J. Microelectron. Electron. Compon. Mater., Vol. 53, No. 1(2023), pp. 3–13 3 R. Fathima et al.; Informacije Midem, Vol. 53, No. 1(2023), 3 – 13 1 Introduction broadband communications networking based on con- stellations of tiny satellites are exam-ined and discussed. With the utilization of satellite communication short-ly, The paper’s primary goal is to demonstrate the technical the optical wireless communication mode has undoubt- details of the Starlink network that SpaceX withholds in edly evolved. Individual OWC systems have standards popular-science presentations [13]. due to their outstanding performance [1]. The IS-OWC systems performance multiplexing techniques are used IS-OWC system is the adaption of wireless technolo-gy with extended distance trans-mission. The inter-satellite in the modern period. Present microwave satel-lite sys- link of 10 Gbit/s data rate has modelled with 5000 kilo- tems in maximum bandwidth, minimum size, and least metres of communica-tion range [2, 3]. Therefore, the power demand offered by the IS-OWC system [14]. The validation of com-parative analysis shows that the RZ RF signal is compared with the laser signal providing and NRZ model modulation are changing input power lower loss. Lasers are associ-ated with the RF wavelength stages [4]. The lightwave system preserves a high data [15]. Joining altered positions on the earth’s surface is a rate for the upcoming generations, and the data rate can fascinating ob-jective. The earth station and satellite are grasp user capacity [5]. The network configuration has establish-ing a network. Consequently, an inter-satellite high speed due to the minimal linear and non-linear dis- link (ISL) contains the energetic influence under the op- tortions by increasing the demand of assert lookup for timal communication implication [16]. Owing to propa- robust networks. In the case of optical fibers, chromatic gation loss being weighty, the time lag in extensive diffusion and polarization mode diffusion comprise lin- transmission for Geo stationary Earth Orbit (GEO). The ear and nonlinear damage such as self and cross-phase globular link and communication function are extreme- modulation and so on. [6, 7]. It is necessary to avoid the ly desirable for the Low Earth Orbit (LEO) as well as Me- optimal modula-tion design by the narrow optical spec- dium Earth Orbit (MEO) satellites [17]. Two satellites in trum and not to resist excessive distortion by improving IS-OWC technology act as transmitters and receivers at spatial performance. the propagation chan-nel. The tracking system’s highly accurate signal is required for beacon signal and quad- The self-phase and cross-phase modulation with the rant detector with connected satellite, which makes sure constant optical power modulation format is a smaller that the tracking system, has orientation and correct line amount of susceptible [8]. The amplified spontaneous of sight (LOS) [18]. The major problem of the system is emission (ASE) in case of long haul optical network, closed-loop tracking. Servo motors are addressing the noise enters the image that may reduce and choose the problem and blocking the beacon signals on satellites. optimum modulation format. The modulation format The other satellites are expanding the ephemeris data plays a vital role in commu-nication [9]. Microwave com- to accurately pinpoint and track the other satellites munication system for satellite communication is im- [19]. The IS-OWC system develops link distance into the portant assistance over the optical wireless communica- modulation system that is uti-lized for input power level, tion (OWC) system is occupied with the communication operation wavelength, receiver sensitivity, and diver- world in huge applications [10]. The minimum power sity methods. Some of the techniques are used to com- consumption and extensive distance transmission appli- pensate for the ef-fects that damage the link efficiency cation are set by the low-cost landmark function in the [20, 21]. The minimum BER is obtained by the ISL link; a OWC system and it is tough to organize the extended low transmitter divergence is utilized for delivering the fiber optic cables [11, 12]. Thus, the effectiveness of the greatest power level on the receiver side and re-moves satellite application of the IS-OWC system is compatible that difficulty in power degeneracy. with communication training. The com-munication sys- tem is needed to overcome better efficiency in errone- Any two satellites in this connection are used for IS- ous transmission, power con-straints and noise. OWC if the satellite is in a similar orbit or dissimilar or- bit. The most common orbits used by the satellite are The goal of every link involving a communications satel- LEO, MEO and geostationary earth orbit. On the other lite is to provide the highest-quality signal with the least hand, the tracking system is highly accurate and it is amount of bandwidth and power while utilising the used for the beacon signals on the hand and quadrant most suitable technology. In a small-satellite constel- detector. The accurate line of sight has aligned and en- lation system, the design decreases the system’s com- sures that the linked satellite is prop-agative. The light plexity and implementation costs, lowers the necessary travels at 3 x 108 m/s in data that can send without transmission power, which improves the signal-to-noise delay and attenuation. The maxi-mum speed of data is ratio, and condenses the frequency spectrum. The Star- transmitted over thousands of kilometres with minor link satellite broadband communications network creat- payload with radio frequency (RF) links over the ben- ed by American company SpaceX is evaluated in this pa- efits of optical links. per. The key technical features of this particular area of 4 R. Fathima et al.; Informacije Midem, Vol. 53, No. 1(2023), 3 – 13 The main contributions of this manuscript are sum- In 2018, Sharma et al [23] presented wireless com-mu- marized below, nication in the dynamic field of IS-OWC. In nu-merous - In this manuscript, the design and compara-tive counts of channels under the structure, the Q factor analysis of the IS-OWC for RZ and NRZ modula- value maximizes as the count of channels maximizes tion formats through channel diver-sity tech- with minimized BER value. In the signal received as dis- niques are proposed. similar routes, the strength and ca-pacity of the signal - The IS-OWC is concerned through the em-ploy- increase with numerous counts of the channel have ment of lasers based on traditional ra-dio and mi- numerous counts of transmitter and receiver antenna. crowave systems for the OWC. The performance of the pro-posed method improved - When compared to the single channel, the chan- the low power and in-creased the count of channels. nel diversity strategy produces better results. - The IS-OWC device uses a diversity tech-nique, in In 2018, Khichar et al [24] presented the IS-OWC system which several signal paths are available, allowing in filter and amplifier diversity techniques utilized for the Q-factor and signal intensity to be extended data transmission. The diversity tech-nique of higher Q or increased over a large distance. factor and improved bit rate was suggested with noise - In this manuscript, the Q-factor with BER is esti- power separation and re-ceived noise power amplifica- mated with diverse modulation formats, like RZ tion. A wavelength of 1550 nm was the modulation of and NRZ using the channel diversity technique. NRZ. The system has 30 dBm input power with a 7000 - Here, a channel variety strategy with multi-ple km link distance over the 40 Gbit/s has improved the transmitter and receiver antennas is used to in- data rate. For additional development, under link dis- vestigate the results for various types of adjust- tance and system data rate were evaluated. ment techniques in NRZ and RZ. - Finally, the experiment is simulated with opti- In 2020, Sivakumar et al [25] presented the DSP al- system-16.0 software. The proposed system is gorithm to improve the link range with the coherent simulated using 32768 numbers samples with a detection technique for the IS-OWC system based on bit rate of 109 bits/sec. high-speed single channel PDM-QPSK. Here, the pre- sented method performance in the influence of receiver The rest of this manuscript is structured as: Section pointing error with OSNR system demand to achieve the 2 describes the literature review of various research acceptable level of BER maximizes the pointing error an- papers associated with IS-OWC for channel diversity gle increases. An efficient IS-OWC transmission system schemes. Section 3 illustrates the proposed IS-OWC for with long-distance and high-speed bandwidth is used. RZ and NRZ modulation through the channel diversity technique. Section 4 demonstrates the experimental In 2019, Sri et al [26] presented the revolutionary tech- outcomes and discussion of the pro-posed method. At nique in IS-OWC that establishes genuine com-munica- last, section 5 concludes the manuscript. tion. Here, the RF satellite links in the knowledge trans- mit with a similar speed of around Mbps. Therefore, reaching the highest knowledge rate in optical links for 2 Related Work homes over optical lasers (OLs) is necessary for exploi- tation. The light beam from the OLs offers less prob- Among the various research works on the IS-OWC for ability for detective work, interrupting with diminish- channel diversity techniques, some of the most recent ing the possibility of electronic signal congestion. OISL works are reviewed here, systems were faster and safer information measures in the trans-mission of knowledge. In 2020, Singh et al [22] presented the 3D orthogonal modulation scheme by incorporating the band-width- In 2018, Viswanath et al [27] presented that the trans- efficient with a high-speed IS-OWC link. Here, three mitter power condition was comparatively greater for autonomous 40 Gbit/s information signals were sent a greater value of turbulence at ground level. Here, on a similar wavelength channel that was modulated the power demand for uplink was 8-10 dB as likened through the parameters of the various signals of the to downlink. In the event, that trans-mitter power was optical carrier. The IS-OWC link effi-ciency of the pre- required for non-feasibility, it can be transported with sented method has been re-searched for growing possible level through trans-mitter spatial diversity or point errors with simulation results showing that the MISO approach for the receiver of uplink diversity or presented IS-OWC link was extended until the link dis- SIMO approach for downlink is aperture averaging. The tance was 16,000 in the effect of pointing error was 3.5 practical rele-vance of the laser communications satel- with satisfactory BER. lite launched into geostationary orbit, ground-satellite and satellite-ground links must be recognized. 5 R. Fathima et al.; Informacije Midem, Vol. 53, No. 1(2023), 3 – 13 In 2017, Pradhan et al [28] presented the IS-OWC sys- tem for space and polarization diversity methods. Here, 25 dBm input power and 7.63 Gbit/s data rate were achieved for the 6,000 km for link length. The diversity technique was polarized by the perfor-mance of sys- tem development also evaluated. The system can be evaluated by enhancing link distance with data rate. 3 Proposed methodology All communication schemes involve a transmitter, propagation medium, and receiver. Two or more sat- ellites are developed by an inter-satellite link that one satellite communicates with another using OWC. Both the satellite transmitter and receiver hold that monitor- ing system for decreasing signal misalignment. Here, the IS-OWC method with chan-nel diversity technique, a transmitter, a propagation medium and a receiver make up the IS-OWC system. And then the binary sig- nal is generated by a pseu-do-random bit sequence (PRBS) that is converted into the electric signal, such as Figure 1: Block diagram of the IS-OWC method with the RZ and NRZ in pulse generators. In RZ and NRZ gen- channel diversity technique. erators through carri-er signal (laser light) using modu- lator Mach - Zender (MZ) supports in modulating the performance of transmitter signal, SF describes that voltage. With the help of a power combiner, the optical optical performance of receiver signal, WH de-notes signal is propagated to the 1: N fork that provides inde- that transmitter signal gain, SH denotes that receiver pendent signal copies with N number of OWC. An ava- signal gain, WK indicates the transmitter pointing loss lanche photodiode (APD) is utilized in the receiv-er for factor, SK indicates the receiver point-ing loss factor, converting the optical domain signal with an electric ω represents the operational wave-length, S represents signal. The redundant signal is filtered with Low Pass the link distance among the transmitter and receiver. Bessel Filter (LPBF). For different modula-tion schemes, the BER analyzer performance is as-sessed based on 3.2 Flow chart the Q factor or BER. Figure 1 por-trays the IS-OWC block diagram method with the channel diversity technique. 3.1 Inter satellite-optical wireless communica-tion (IS-OWC) This is utilized for the carrier signal to transmit the sig- nal as laser light. The system is dependent on numer- ous limitations, such as laser light wave-length, trans- mitted power, modulation technique in different types, antenna size of the receiver and transmitted signal. The laser light signal is utilized to deliver wireless connec- tivity between the source and the destination. The free - space is the medium utilized here to take the mes- sage. The receiver sig-nal power level in the IS-OWC system is calculated [18] using equation (1) 2 QS WQW  F S  FWHS  HWKSK   S  (1)  4  Where QS represents the received power, WQ repre- Figure 2: Flow chart of IS-OWC system using channel sents the transmitted power, WF describes the opti-cal diversity technique 6 R. Fathima et al.; Informacije Midem, Vol. 53, No. 1(2023), 3 – 13 3.3 Return to zero modulation format Figure 4 shows the layout diagram of 4 channels OWC for return to zero modulation. Here, the optical signal A pseudorandom pulse generator produces the infor- propagates with a 1: N fork that provides independent mation to be broadcasted in the structure of binary signal replicates with N number of OWC and the sig- data and comprises the layout that desig-nates the nal is converted into four channels with a frequency of transmitted segment. The binary data is served into a 1550 nm and a range of 2500 km. For the modulation return to zero pulse generator, which converts it into scheme, the performance of the BER analyzer rate is the electrical pulses for transmission. RZ pulse gen- 1.96543*10-030. erator output is directed to an MZ modulator. The MZ modulator modulates the elec-trical pulse through a continuous wave laser with 1,550 nm. The main use of the optical antenna is the optical signal output in the MZ modulator that is directed near the n-OWC chan- nel. The IS-OWC link’s receiving end consists of an optical receiving anten-na that receives the incoming signal. The signal is guided to a photodiode called APD that transforms the optical domain signal (ODS) into its Figure 5: Layout diagram of 6-OWC electrical equivalent. A Bessel low-pass filter surveys the APD photodiode, and then high-frequency noise in Figure 5 shows the layout diagram of 6 channels OWC the received signal is eliminated. The Q-factor, BER, and for return to zero modulation. Here, the binary signal SNR of the received signal are calculated using the BER is generated through PRBS that is converted into an analyzer. electric signal via pulse generator RZ. The MZ modula- tor helps in modulating the voltage of the RZ generator with carrier signal in the extinction ratio of 20 dB. For the modulation scheme, the per-formance of the BER analyzer rate is 8.58148*10-052. Figure 3: Layout diagram of 2-OWC Figure 3 shows the layout diagram of 2 channels OWC for return to zero modulation. In this, the bina-ry sig- nal is generated by the PRBS that is converted into an electric signal via pulse generator RZ. The MZ modu- lator helps modulate the voltage of the RZ generator through the carrier signal (laser light) extinction ratio of 20 dB. And then the carrier signal is converted into two channels with a frequency of 1550 nm and a range of 2500 km. With the help of a power combiner, the opti- cal signal has been propa-gated to an optical amplifier Figure 6: Layout diagram of 8-OWC measured with a max-imum output power of 25 dBm. An APD is used in the receiver for converting ODS with Figure 6 shows the layout diagram of 8 channels OWC an electric signal. The redundant signal is filtered out for return to zero modulation. The optical sig-nal propa- using LPBF. For the modulation scheme, the perfor- gates with a 1: N fork that provides inde-pendent signal mance of the BER analyzer rate is. 6.4925*10-015. replicas with N number of OWC, and the signal is con- verted into eight channels with a frequency of 1550 nm and a range of 2500 km. By using the power combiner, the optical signal has been propagated to the optical amplifier measured with the maximum output power of 25 dBm. An APD is used in the receiver for converting ODS with an electric signal. The redundant signal is fil- tered using LPBF. For the modulation scheme, the per- formance of the BER analyzer rate is 3.29942*10-066. Figure 4: Layout diagram of 4-OWC 7 R. Fathima et al.; Informacije Midem, Vol. 53, No. 1(2023), 3 – 13 3.4 Non return to zero modulation layouts ulation scheme, the performance of the BER analyzer rate is 8.86894*10-064. The binary signal is made by the PRBS that is trans- formed using an electrical signal via an NRZ pulse gen- erator. The NRZ generator output is modulated by the carrier signal utilizing the MZ modulator, which is prop- agated optically to Fork 1: N, it pro-vides the N number of the optical wireless channel and separates the signal imitations utilizing the power combiner. An APD is used in the receiver for converting ODS with an electric sig- nal. The removal of an unwanted signal is used by the LPBF. Then the use of the BER analyzer is a view of the Figure 10: Layout diagram of 8-OWC results of the Q factor or BER form. The efficiency of the proposed method improves the number of channels, Figure 10 shows the layout diagram of 8 channels OWC and transmitters and increases the receiver antenna. for non-return to zero modulation. Here, the perfor- mance of the BER analyzer rate is 1.67178*10-088. 4 Results and discussion This section describes the implementation of the experi- ment and carried out the details. The simula-tion is per- formed in opti-system-16.0 software. And the simulation Figure 7: Layout diagram of 2-OWC parameters utilized under experi-ments are tabulated in Table 1. Here, table 1 por-trays the simulation parame- Figure 7 shows the layout diagram of 2 channels OWC ters advanced to im-plement the proposed IS-OWC sys- for non-return to zero modulation. For the modulation tem. Continuous wave laser (CWL) utilized a source with scheme, the performance of the BER analyzer rate is 193.1 THz transmitted frequency. The proposed method 1.34606*10-020. is simulated with 32768 samples with 109 bits/sec bit rate. The transmitter frequency 100 GHz spacing with the transmitted power is 15 dBm is occupied and 1550 nm wavelength is used for simulation. 4.1 Performance evaluation The performance measures such as gain, pointing loss, Figure 8: Layout diagram of 4-OWC bit error rate (BER) are discussed below, Figure 8 shows the layout diagram of 4 channels OWC 4.1.1 Gain for non-return to zero modulation. For the modulation The antenna gain is the measure of effectiveness that scheme, the performance of the BER analyzer rate is is maximum with the antenna can emit the de-livered 5.53168*10-044. power by the transmitter side, and the transmitter is given [23] by equation 2 2 H '   d  W  W  ( )   2   The gain of the receiver side antenna is given by equa- tion 3, 2 H '   d  S   ( )   3 S Figure 9: Layout diagram of 6-OWC   Figure 9 shows that the layouts diagram of 6 chan-nel Where, dW represents the transmitter telescope diam- OWC for non-return to zero modulation. For the mod- eter, dS denotes the diameter of the destina-tion of tel- escope. 8 R. Fathima et al.; Informacije Midem, Vol. 53, No. 1(2023), 3 – 13 Table 1: Simulation parameters  BER 1  erfd P    (6) Parameter Value 2  2  Simulation Window e P2 Bit Rate 109 bits/sec Samples number 32768 BE 2 R  (7) Transmitter P 2 Frequency 193.1 THz 4.2 Q- factor analysis for RZ and NRZ Laser CWL Spacing Frequency 100GHz Figure 12 and 13 shows the Q factor modulation for- Power 15dBm mats in RZ and NRZ system. Based on the eye diagram Extinction ratio 20dB the channel 2, 4, 6 and 8 results in RZ with NRZ formats Line width 0.01MHz are shown below. The “eye” of the digi-tal signal of a hu- Type of Modulation RZ/NRZ man eye is formed on an oscillo-scope, which displays Optical Wireless Channel the transmission system out-put, and the eye reflects Range 2500km the consistency of SNR at the “eye” of a digital signal. The sampling process with the largest “eye-opening” Wavelength 1550nm is the best place to determine if a given bit is “1” or a Transmitter Aperture diameter 15cm “0.” The largest “eye-opening,” is higher variance among Receiver Aperture diameter 15cm the mean values of signal levels for “1” and “0.”The eye Attenuation 0 dia-gram for channels 2, 4, and 6 furnishes a small eye- Optical Efficiency 1 opening, which means that the Inter Symbol Inter-fer- Transmitter &Receiver Pointing Error 1.1μrad ence (ISI) is high. The eye diagram of channel 8 provides Additional losses 1.5dB a big opening that refers to the ISI being low. Propagation delay 0 Receiver 4.2.1 2 and 4 channel Q factor analysis Responsively of APD 1 A/W Cut off frequency (LPBF) 0.72*Symbol Rate in Hz 4.1.2 Pointing loss The transmitter side of pointing loss [2] is calculated by given equation, K  eHWWW (4) The receiver side of pointing loss [2] is calculated by given equation, K  eHSSS (5) Where, implicates the pointing errors of transmitter and receiver respectively. 4.1.3 Bit Error Rate (BER) System representation indicates the BER and com-mu- nication quality in the OWC system. The accepta-ble signal level includes below 10-9 of the BER value. Sub- sequently, the BER value for the optical system is less; Figure 11: Maximum Q-factor of 2 and 4 Channel for RZ it is hard for analyzing the Q factor employed as the with NRZ modulation performance of the system. Q factor contains a straight relationship through the BER signal [15]. Figure 11 shows the maximum Q factor in the second channel and fourth channel for RZ and NRZ modula- tion. Q factor and BER denote the converse relationship 9 R. Fathima et al.; Informacije Midem, Vol. 53, No. 1(2023), 3 – 13 in the 2 and 4-channel modulation schemes. Here, two eight chan-nels in return to zero modulation produce channels in re-turn to zero modulation produce the the max-imum Q factor value is 17.1281, similarly, eight maximum Q factor value is 7.65005, two channels in channels in non-return to zero modulation produces non-return to zero modulation produce the maximum the maximum Q factor value is 19.2058. Q factor value is 9.20808, four channels in return to zero modulation produce the maximum Q factor value is 4.3 BER Analysis for RZ & NRZ 11.3824, simi-larly, four channels in non-return to zero mod-ulation produces the maximum Q factor value is Figures 13 and 14 show the minimum BER of channels 13.8417. 2, 4, 6 and 8 of RZ and NRZ modu-lation format. The BER specifies the possibility of wrong bit identifica- 4.2.2 6 and 8 channel Q factor analysis tion through the decision circuit. The design of OWC systems depends on performance estimation like BER figures. The channel comparison 2, 4, 6 and 8 suggests that the bit error rate is within the minimum target. 4.3.1 2 and 4 channel BER analysis Figure 12: Maximum Q-factor of 6 and 8 Channel for RZ with NRZ modulation Figure 12 depicts the maximum Q factor in the sixth channel and eighth channel for RZ and NRZ modula- tion. Q factor with BER denotes the converse rela-tion- Figure 13: Minimum BER of 2 and 4 channel for RZ and ship in the 6 and 8-channel modulation scheme. Here, NRZ modulation the sixth channels in return to zero modulation pro- duce the maximum Q factor value is 15.0768, the sixth Figure 13 shows the minimum BER in the second chan- channels in non-return to zero modulation pro-duce nel and fourth channels for RZ and NRZ modula-tion. the maximum Q factor value is 17.2058, eight channels Q factor and BER are the converse relationship in the 2 in return to zero modulation produce the maximum Q and 4-channel modulation scheme. Here, two channels factor value is 17.1281, similarly, eight channels in non- in RZ modulation produce that mini-mum BER value return to zero modulation produces the maximum Q indicates, two channels in NRZ mod-ulation produce factor value is 19.2058.Figure 12 depicts the maximum that minimum BER value indicates, four channels in RZ Q factor in the sixth channel and eighth channel for RZ modulation produce that mini-mum BER value is, simi- and NRZ modulation. Q factor with BER denotes the larly four channels in NRZ modulation produces that converse relationship in the 6 and 8-channel modula- minimum BER value indi-cates. 5053168*10-044 tion scheme. Here, the sixth channels in return to zero modulation produce the maximum Q factor value is 4.3.2 6 and 8 channel BER analysis 15.0768, the sixth channels in non-return to zero modu- Figure 14 shows the minimum BER in the sixth chan- lation produce the maximum Q factor value is 17.2058, nel and eighth channel for RZ and NRZ modulation. Q 10 R. Fathima et al.; Informacije Midem, Vol. 53, No. 1(2023), 3 – 13 6 Conflict of Interest The author of this document does not have any Con- flict of Interest (COI) in publishing this paper. 7 References 1. Kaur, R. and Kaur, H., 2018. Comparative analy- sis of chirped, AMI and DPSK modulation tech- niques in IS-OWC system Optik, 154, pp.755-762 https://doi.org/10.1016/j.ijleo.2017.10.108 2. Kumari, G. and Selwal, C., 2018. Performance Im- provement in Inter-satellite Optical Wireless Com- munication to Mitigate Losses Using Diver-sity Technique, In Optical and Wireless Technol-ogies (pp. 375-382).Springer, Singapore https://doi.org/10.1007/978-981-10-7395-3_42 3. Karpagarajesh, G., Blessie, A. and Krishnan, S., 2021. Performance Assessment of Dispersion Compensation Using Fiber Bragg Grating and Figure 14: Minimum BER of 6 and 8 channel for RZ and Dispersion Compensation Fiber Tech-niques. In- NRZ modulation formacije MIDEM, 51(4), pp.215-223. https://doi.org/10.33180/InfMIDEM2021.402 factor and BER indicate the converse relationship in the 4. Kong, M., Kang, C.H., Alkhazragi, O., Sun, X., Guo, 6 and 8-channel modulation schemes. Here, the sixth Y., Sait, M., Holguin-Lerma, J.A., Ng, T.K. and Ooi, channels in RZ modulation produce that minimum B.S., 2020. Survey of energy-autonomous solar BER value indicates, sixth channels in NRZ modulation cell receivers for satellite–air–ground–ocean op- produce that minimum BER value is, eight channels in tical wireless communication, Progress in Quan- RZ modulation produce that mini-mum BER value indi- tum Electronics, p.100300 cates, similarly eight channels in NRZ modulation pro- https://doi.org/10.1016/j.pquantelec.2020.100300 duce that minimum BER value indicates 1.67178-10-88. 5. Escribano, F.J., Wagemakers, A., Kaddoum, G. and Evangelista, J.V., 2020. A spatial time-frequency hopping index modulated scheme in turbulence- 5 Conclusions free optical wireless communication channels. IEEE Transactions on Communications, 68(7), pp.4437-4450 In this manuscript, the compara-tive analysis of IS-OWC https://doi.org/10.1109/TCOMM.2020.2987561 for RZ and NRZ modulation formats for chan-nel di- 6. Kaur, R. and Kaler, R.S., 2020. Performance of zero versity techniques is pro-posed. An n-channel IS-OWC cross correlation resultant weight spectral am- sys-tem functioning at 10 Gbit/s with several inflexions plitude codes in lower Earth orbit‐based op-tical setups for 15 dBm instance of power is encom-passed in wireless channel system. International Journal of the modeled scheme. The Inter-satellite signals are im- Communication Systems, 33(12), p.e4456 proved and let they portable to extended distances are https://doi.org/10.1002/dac.4456 done by connecting several OWC channels between the 7. Zhu, C., Li, Y., Zhang, M., Wang, Q. and Zhou, W., transmitter and the receiver. In an optical wireless com- 2020, May. An Optimization Method for the Gate- munication device, the NRZ modulation is much supe- way Station Deployment in LEO Satellite Systems. rior to RZ modulation since in this BER is lesser and the Q In 2020 IEEE 91st Vehicular Technology Confer- factor is higher. The Q factor at-tained a maximum in the ence (VTC2020-Spring) (pp. 1-7).IEEE. 8-channel OWC is 19.4385, whereas the Q factor attained https://doi.org/10.1109/VTC2020-Spring48590.2020.9128413 minimum in the 2-channel OWC is 19.4385. Moreo-ver, 8. Dubey, D., Prajapati, Y.K. and Tripathi, R., 2020. Error the increased number of chan-nels may develop the performance analysis of PPM-and FSK-based hy- proficiency of low power inter-satellite relations. brid modulation scheme for FSO satel-lite down- 11 R. Fathima et al.; Informacije Midem, Vol. 53, No. 1(2023), 3 – 13 link. Optical and Quantum Electron-ics, 52, pp.1-16 erational Concept for a Multi-orbit Disaggregat- Optical and Quantum Electron-ics (2020) 52:286 ed System. IEEE Journal on Miniaturization for Air https://doi.org/10.1007/s11082-020-02404-7 and Space Systems,1(1),pp.56-65 9. Garg, N. and Ahmad, A., 2021. Performance of Out- https://doi.org/10.1109/JMASS.2020.3001080 door Optical Wireless Communication Chan-nel 19. Kumar, N., 2014. Enhanced performance analysis through Gamma-Gamma Turbulence at Dif-fer- of inter-satellite optical-wireless communi-cation ent Frequency In Advances in Electromechan-ical (IsOWC)system. Optik, 125(8),pp.1945-1949. Technologies (pp. 551-560).Springer, Singa-pore https://doi.org/10.1016/j.ijleo.2013.11.008 https://doi.org/10.1007/978-981-15-5463-6_49 20. Lu, K., Xiao, G. and Chlamtac, I., 2005. Anal-ysis of 10. Karpagarajesh, Guruviah, and Helen Anita. “Max- blocking probability for distributed lightpath es- imizing the area spanned by the optical SNR of tablishment in WDM optical net-works. IEEE/ACM the 5G using digital modulators and filters.” Int. Transactions on networking, 13(1), pp.187197. Arab J. Inf. Technol. 18, no. 3 (2021): 306-311 https://doi.org/10.1109/TNET.2004.842233 https://doi.org/10.34028/iajit/18/3/6 21. Ravishankar, C., Gopal, R., BenAmmar, N., Zakaria, 11. Malik, S. and Sahu, P.K., 2020. Performance analy- G. and Huang, X., 2021. Next‐generation global sis of free space optical communication system satellite system with mega‐constellations.Inter- using different modulation schemes over weak national Journal of Satellite Communications and to strong atmospheric turbulence channels. In Networking, 39(1), pp.6-28 Optical and Wireless Technologies (pp. 387-399). https://doi.org/10.1002/sat.1351 Springer, Singapore 22. Singh, M., Malhotra, J. and Dhasarathan, V., 2020. https://doi.org/10.1007/978-981-13-6159-3_41 A high‐speed single‐channel inter‐satellite op- 11. Elsayed, E.E. and Yousif, B.B., 2020. Performance tical wireless communication link incorporat- enhancement of M-ary pulse-position modula- ing spectrum‐efficient orthogonal modulation tion for a wavelength division multiplexing free- scheme. Microwave and Optical Technology Let- space optical systems impaired by interchannel ters, 62(12),pp.4007-4014 crosstalk, pointing error, and ASE noise. Optics https://doi.org/10.1002/mop.32511 Communications, 475, p.126219. 23. Sharma, P. and Meena, S., 2018, July. Per-formance https://doi.org/10.1016/j.optcom.2020.126219 Analysis of Inter-Satellite Optical Wireless Commu- 12. Andreje, Ž. and Batagelj, B., 2021. Vrednotenje nication (Is-OWC) System by using Channel Diver- tehničnih lastnosti komunikacijskega omrežja sity Technique.In 2018 Inventive Research in Com- majhnih satelitov Starlink. Elektrotehniski vest- puting Applications (ICIRCA) (pp. 477-480).IEEE nik, 88(1/2), pp.1-7. https://doi.org/10.1109/ICIRCA.2018.8597344 https://ev.fe.uni-lj.si/1-2-2021/Andrejc.pdf 24. Khichar, S. and Inaniya, P.K., 2018, April. Inter-Sat- 14. Sri, I.K. and Srinivasulu, A., 2019. Perfor-mance ellite Optical Wireless Communication Sys-tem Analysis of Inter-satellite Optical Wireless Com- Design Using Diversity Technique with Filter and munication Using 12 and 24 Transponders. In Pro- Amplifier.In 2018 Communication and Signal Pro- ceeding of the Second Microelectronics, Compu- cessing (ICCSP)(pp.0481-0484).IEEE. ting & Communication Systems (MCCS 2017) (pp. https://doi.org/10.1109/ICCSP.2018.8524366 811-821). Springer, Singapore 25. Chaudhary, S., Sharma, A. and Singh, V., 2019. Opti- https://doi.org/10.1007/978-981-10-8234-4_66 mization of high speed and long haul inter-satellite 15. Patnaik, B. and Sahu, P.K., 2012. Inter-satellite optical communication link by incorporating differential wireless communication system design and simula- phase shift key and orthogonal frequen-cy division tion. IET Communications, 6(16), pp.2561-2567. multiplexing scheme. Optik, 176, pp.185-190. https://doi.org/10.1049/iet-com.2012.0044 https://doi.org/10.1016/j.ijleo.2018.09.037 16. Sharma, V. and Kumar, N., 2013. Improved analysis 26. Karpagarajesh, G., Krishnan, R.S., Robinson, Y.H., of 2.5 Gbps-inter-satellite link (ISL) in inter-satel- Vimal, S., Kadry, S. and Nam, Y., 2021. Investiga- lite optical-wireless communication (IsOWC) sys- tion of digital video broadcasting application em- tem. Optics Communications, 286, pp.99-102. ploying the modulation formats like QAM and PSK https://doi.org/10.1016/j.optcom.2012.08.055 using OWC, FSO, and LOS-FSO channels. Alexan- 17. Pandi, V.S., Singh, M., Malhotra, J. and Grover, A., 2021. dria Engineering Journal. https://www.sciencedi- A spectral-efficient 1 Tbps terres-trial free-space op- rect.com/science/article/pii/S1110016821004014 tics link based on super-channel transmission. Opti- 27. Viswanath, A., Jain, V.K. and Kar, S., 2015. Analysis cal and Quantum Electronics, 53(5), pp.1-12. of earth‐to‐satellite free‐space optical link perfor- https://doi.org/10.1007/s11082-021-02931-x mance in the presence of turbulence, beam‐wan- 18. Anyanhun, A.I., Anzagira, A. and Edmonson, W.W., der induced pointing error and weather condi- 2020. Intersatellite Communication: An MBSE Op- 12 R. Fathima et al.; Informacije Midem, Vol. 53, No. 1(2023), 3 – 13 tions for different intensity modulation schemes. IET Communications, 9(18), pp.2253-2258. https://doi.org/10.1049/iet-com.2015.0379 28. Pradhan, S., Sahu, P.K., Giri, R.K. and Patnaik, B., 2017, December. Inter-satellite optical wireless communication system design using diversity tech-niques. In 2017, Optical and Communication Engi-neering (ICMOCE)(pp.250-253).IEEE https://doi.org/10.1109/ICMOCE.2015.7489738 Copyright © 2023 by the Authors. This is an open access article dis- tributed under the Creative Com- mons Attribution (CC BY) License (https://creativecom- mons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Arrived: 09. 09. 2022 Accepted: 23. 12. 2022 13 14 Original scientific paper https://doi.org/10.33180/InfMIDEM2023.102 Journal of Microelectronics, Electronic Components and Materials Vol. 53, No. 1(2023), 15 – 30 Loss reduction and reliability improvement in distributed network using HF-SOA based optimal installation of DG, SCs and STF Ithaya Rajagopalan1, Jagatheeswari Ponnuswamy2 1Department of Electrical and Electronics Engineering, Arunachala college of engineering for women, Manavilai Road, Manavilai, Nagercoil, Tamil Nadu, India 2Department of Electronics and Communication Engineering, Ponjesly college of Engineering, College Road, Parvathipuram, Nagercoil, Tamil Nadu, India Abstract: In recent years, most of the research works related to Distributed Generation (DG), targeted on the loss minimization and reliability enhancement due to the existence of intermittent Renewable Energy Sources (RES). In this research work, this target is attained by an optimal installation of DG, shunt capacitors (SCs) and single tuned filter (STF) through a novel hybrid fuzzy based seagull optimization algorithm (HF-SOA) in a distributed power network. Compared to the literatures better harmonics mitigation is achieved in this research work due to the presence of STF. The proposed research problem is considered as multi-objective and a novel objective function that incorporates, minimization of power loss, harmonics and enhancement of voltage profile (VP) as well as system reliability is introduced in this research article. The fuzzy membership function is framed for each objective function parameter and the fuzzified membership functions are considered as an objective function for the SOA approach. Three case studies are conducted in both IEEE 33 and 69 radial networks to examine the influence of the HF-SOA algorithm in satisfying the proposed multi-objective function (MOF). In the case studies, the percentage loss, THD reduction, VP enhancement, cost reduction of DG and reliability improvement measured through expected interruption cost (ECOST) are analyses in detail. The coding of HF-SOA and analysis of the proposed work are resolved in the MATLAB R2022a Editor Software. The simulation results confirms that the proposed HF-SOA is superior than the with the recently published optimization approaches named genetic moth swarm algorithm (GMSA) and salp swarm optimization algorithm (SSA). Keywords: Distributed Generation (DG); hybrid fuzzy based seagull optimization algorithm (HF-SOA); Power loss minimization; reliability improvement; seagull optimization algorithm (SOA); Shunt Capacitors (SCs) Zmanjšanje izgub in izboljšanje zanesljivosti v porazdeljenem omrežju z uporabo optimalne namestitve DG, SC in STF na podlagi HF-SOA Izvleček: V zadnjih letih je večina raziskovalnih del, povezanih z razpršeno proizvodnjo električne energije (DG), usmerjena v zmanjševanje izgub in povečanje zanesljivosti zaradi obstoja nestalnih obnovljivih virov energije (RES). V članku je ta cilj dosežen z optimalno namestitvijo DG, vzporenih kondenzatorjev (SC) in enojnega uglašenega filtra (STF) s pomočjo novega hibridnega optimizacijskega algoritma na osnovi fuzzije (HF-SOA) v distribuiranem elektroenergetskem omrežju. V primerjavi z literaturo je v tem raziskovalnem delu zaradi prisotnosti STF dosežena boljša ublažitev harmonikov. Predlagan problem je obravnavan kot večpredmeten, predstavljena je nova ciljna funkcija, ki vključuje minimizacijo izgube moči, harmonikov in izboljšanje napetostnega profila (VP) ter zanesljivosti sistema. Za vsak parameter ciljne funkcije je oblikovana mehka funkcija pripadnosti, mehke funkcije pripadnosti pa se obravnavajo kot ciljna funkcija za pristop SOA. Da bi preverili vpliv algoritma HF-SOA pri izpolnjevanju predlagane večobjektivne funkcije (MOF), so izvedene tri študije primerov v radialnih omrežjih IEEE 33 in 69. V študijah primerov so podrobno analizirani odstotki izgube, zmanjšanje THD, izboljšanje VP, zmanjšanje GD in izboljšanje zanesljivosti, merjeno s pričakovanimi stroški prekinitve (ECOST). Kodiranje HF-SOA in analiza predlaganega dela sta rešena v programski opremi MATLAB R2022a Editor. Rezultati simulacije potrjujejo, da je predlagani HF-SOA boljši od nedavno objavljenih optimizacijskih pristopov, imenovanih algoritem genetskega roja (GMSA) in algoritem optimizacije roja Salp (SSA). Ključne besede: razpršena proizvodnja (DG); hibridni mehki optimizacijski algoritem (HF-SOA); minimizacija izgube moči; izboljšanje zanesljivosti; optimizacijski algoritem (SOA); vzporedni kondenzatorji (SC) * Corresponding Author’s e-mail: ithayaithaya777@gmail.com How to cite: I. Rajagopalan et al., “Loss reduction and reliability improvement in distributed network using HF-SOA based optimal installation of DG, SCs and STF", Inf. Midem-J. Microelectron. Electron. Compon. Mater., Vol. 53, No. 1(2023), pp. 15–30 15 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30 1 Introduction pensator (SVC) in RDN to reduce the cost and loss func- tions [7]. Nowadays, the conventional energy sources are re- placed with the RES due to the increasing energy cri- A fuzzy based optimal installation of DG using Genetic sis and lack of availability of fossil fuels. Hence, the re- Algorithm (GA) to reduce real and reactive power sup- search platform widens by virtue of power quality (PQ) ply as well as losses, stability index and enhancement issues occurring in Radial Distributed Network (RDN) minimum bus voltage was introduced by Srinivasa Rao including intermittent RES. The crucial factors com- Gampa et.al [8]. In this approach the fuzzy membership monly considered in most of the earlier research works functions are framed based on the proposed objective to mitigate PQ issues in RDN are power loss minimiza- function and the selection of nodes for distributed tion, VP improvement and reliability improvement. To generation and SCs with different power factors were effectively meet these factors, the distributed genera- optimized with GA approach. The experimental analy- tion system is suggested as an alternate solution in- sis was performed in both 51 and 69 node RDN and stead of expanding the network infrastructure since, successfully obtained the proposed objective function. its limited commissioning time and its effectiveness in minimizing network losses. In recent research works, In the same manner, optimal integration of capacitor the SCs are installed along with distributed generation and DG using Bat Algorithm (BA) was proposed by as it provides reactive power compensation for enhanc- Thangaraj Yuvaraj et.al which was aimed to reduce the ing the VP. Hence, both real as well as reactive powers power loss and stability improvement of the RDN. The could be improved with the combination of distributed load variations such as constant, industrial, residential generation and SCs in RDN. However, the optimal se- and commercial were also considered as one of the lection and installation of DG and SCs is a challenging main factors in that research work [9]. The experimen- task since the cost of installation of distributed genera- tal study in that article was performed in 33 and 69 tion should overcome the energy losses. So many opti- node network and the superiority of the Bat Algorithm mization approaches are being developed to optimize was proved over conventional methodologies. the siting and sizing of distributed generation and SCs and some of them are presented below: The ant lion optimization (ALO) was suggested by Ahmed R. and Abul’Wafa to select the optimal node for Most of the earlier research articles reported so far the installation of DG and SCs based on the proposed aimed to optimize the sitting and sizing of DG only [1- multi-objective function [10]. The ALO algorithm was 3]. The optimal siting and sizing are commonly termed examined on the IEEE 118 node network and the in- as optimal installation in this research article. The ar- stallation of various RES and the SCs were optimized in tificial intelligence (AI) based optimization algorithms that article. Amirreza Naderipour et.al have presented which mimic the social behavior of living organisms the spotted hyena optimizer (SHO) approach for opti- were suggested to foreseeing the optimization instal- mizing the DG and SCs in a cost-effective manner. The lation of distributed generation. The particle swarm SHO approach was tested in both island and grid con- optimization (PSO) algorithm was proposed to predict nected mode with different modes of operation and the cost-effective installation of DG to reduce the pow- power factors and the superiority was proved over grey er losses, and harmonics [4]. The artificial bee colony wolf optimization (GWO) algorithm [11]. (ABC) algorithm was reported to optimize the size and installation node of DG by minimizing the multi-objec- In similar manner, many research articles were pub- tive function such as cost, voltage drop and power loss lished based on the optimal sizing and installation in the network [5]. The Manta Ray Foraging optimiza- nodes of DG and SCs combination so as to reduce both tion algorithm (MRFO) was suggested to diminish the real and reactive power losses [7, 12-14]. Commonly, network loss in RDN by optimal installation of DG in in most of the article, the power loss minimization was power network. The result analysis of MRFO was per- considered as the fundamental objective function re- formed with 3, 69 and 85 bus system [6]. lated to the proposed research arena. Furthermore, the minimization of installation and operating cost and en- However, the presence of inductive components in hancement of VP, system stability and reliability were power network induces lagging power factor which also reported as effective objective functions in some minimize the VP and increases the network losses. of the research articles [9, 10, 15, 16]. Hence, it is necessary to improve the power factor and VP by installing the SCs thereby reduce the network Optimization techniques like Constriction Coefficient losses. Belkacem Mahdad and K. Srairi presented the Particle Swarm Optimization (CPSO) are used to reduce adaptive differential algorithm-based optimization of the loss of renewable energy resources (RES) in distrib- siting and sizing of DG in presence of static VAR com- uted generation systems. By used this method, losses 16 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30 are reduced by 63.90 % [27]. To reduce voltage regula- ii. Most of the articled focused only on optimal DG tion issues, non-sorting dominated genetic algorithms installation and very few articles include the pres- (NSGAs) were employed to allocate battery energy ence of shunt capacitors. storage systems and distributed energy resources op- iii. The multi-objective functions proposed in earlier timally [28]. The Modified Shuffled Frog Leaping Algo- literatures employed mainly with loss reduction rithm (MSFLA) was proposed to minimize energy loss, and VP improvement and little effort has been operational costs, and energy not supplied by DG sys- expended on reliability enhancement and har- tems [29]. Using an AC optimal power flow (OPF) and monic mitigation. genetic algorithm (GA) for 24 hours, [30] minimize in- vestment and operation costs in renewable energy sys- By reviewing the literatures related to the proposed tems integrated with battery banks. In [31] an optimal research problem, and to overcome the research gap allocation of DG and D-STATCOM within a distribution stated above the following augmentation are present- system using the Bat Algorithm, in which the loss sen- ed in the proposed research work: sitivity factor (LSF) is utilized to find the optimum loca- tion for distribution generation systems. The minimum According to the research gap, the contributions of the loss obtained is 31.94 kW, which is 89.88 %. proposed work is described as follows; - (i) A novel HF-SOA optimization algorithm is pro- In some of the recent articles the single tuned filters posed in this research article to optimize the sit- (STF) were also installed in the distribution network ing and sizing of GD, SCs and STF and better re- which consist of a series connected resistor, inductor sults are produced compared to GMSA and SSA and a capacitor. The number of harmonics presented in approaches. the network is measured with Total harmonics distor- - (ii) The SCs also included in the power network tion (THD) and it is added as one of the prime factors in and its influence on enhancing the VP and dimin- multi-objective function of the optimization problem. ishing of power losses are proved in this article. The comparative THD values of the network in the pres- - (iii) The STF is installed in the power network and ence and absence of STF undoubtedly confirms that a novel multi-objective function that incorporates the THD is much reduced in the presence of STF [17]. network loss minimization, harmonics mitigation, VP enhancement, reliability improvement and to- A technical research growth was observed from the tal cost minimization of DG is developed in this literature survey. Most of the earlier research works proposed research problem. The effectiveness presented so far have initially targeted only on DG in- of the proposed algorithm is also experimented stallation to minimize the power loss with analytical in standard IEEE 33 and 69 node networks with calculations. Subsequently, intelligent algorithms were three case studies and its dominance is confirmed proposed to optimize the installation of DG with a sin- over genetic moth swarm algorithm (GMSA) [18] gle objective function. Afterwards, the multi-objective and salp swarm optimization algorithm (SSA) functions were introduced which incorporated mini- [19]. mization of losses, harmonics, voltage drop, cost also to improve VP, stability as well as reliability. To provide The content presented in this research work are struc- the reactive power compensation and to enhance the tured as follows: Section 2 describes the distributed VP the compensating devices such as static VAR com- power network with 33 and 69 node RDN. Section 3 pensator, shunt capacitor and DSTATCOM were then in- explains the system description of DG network. The stalled and their location are optimized along with DG. problem formulation and the (HF-SOA) optimization approach proposed in this research work is apprized in section 4. The section 5 acquainted with detailed 2 Materials and methods computational analysis performed in this paper out and their results and discussions. Finally, section 6 de- From the observations made from the literature study, scribes the conclusion of proposed work together with the following research gaps are identified: the future scope. i. However, enormous optimization methodologies have been proposed so far to predict a suitable sizing and location of DG so as to compensate 3 Distributed power network the suggested objective functions, the research works are still in progress on developing the 3.1 System description novel metaheuristic approaches to reimburse the objective functions in a better way. While integrating the intermittent renewable energy system in DG the distributed power network faces the 17 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30 problems such as voltage fluctuations, intensifies the the minimization function in the optimization problem network losses and increases the THD. Hence, the fu- to mitigate the harmonics. ture development of power systems is mainly targeted to develop smart distributed power networks which at- tracts the focus of so many researchers to work towards it [20]. One of such attempts is made in this proposed 4 Optimizing the size and allocation of research article by developing an optimization algo- DG with proposed AO algorithm rithm to foreseeing the siting and sizing of DG, SCs and STF. In this proposed research work, the solar photo 4.1 Problem Formulation voltaic (PV) and wind turbines are considered to be in- tegrated in DG system and the SCs and STFs are shunt- The optimal installation of DG, SCs and STF is the pro- ed with the buses. The rating of SCs might be carefully posed research problem and the results are optimized chosen that the voltage rise problem not occurring in with the HF-SOA approach. The multi-objective function power system. In general, the optimization problems is developed for this optimization problem that incor- are experimented in reconfigured IEEE systems. In the porates loss minimization, harmonics mitigation and similar manner, the proposed HF-SOA based installa- VP improvement. The development of multi-objective tion of DG, SCs and STF is examined in 33 and 69 RDN function parameters and the constraints are described which are illustrated in Fig 1 and Fig 2. Prepare the fig- as below. The overall objective function for the proposed ures and tables according to these instructions research problem is developed as stated below (0a,0b): Obj fn  min Obj1 Obj 3Obj 5  max Obj 2Obj 4 (0a) N Obj fn wn min PTL THD CostDG  max Rindex VP  (0b) i1 Where, Obj1 represents the power loss minimization (PTL), Obj2 represents the reliability index (Rindex), Obj3 represents the Harmonics minimization (THD), Obj4 rep- resents the Voltage profile improvement (VP), and Obj5 represents the overall cost of the DG system. wn indicates the weighting factor used to emphasize the importance of lowering each term of the objective function, w1 = 0.4, Figure 1: IEEE 33 node reconfigured RDN w2 = 0.1, w3 = 0.1, w4 = 0.4, and w5 = 0.1 attributed to power loss, reliability index, THD minimization, Voltage profile and overall cost of the DG system. 4.1.1 Power loss minimization (Obj 1): The power loss (PL) minimization is the parameter is commonly considered in most of research work related to DG installation. The optimal sizing and allocation of DG alone supports only for the real power losses and the installation of SCs along with DG assist for reactive power compensation in addition. The active and reac- Figure 2: IEEE 69 node reconfigured RDN tive power losses between i and i + 1 are denoted in (1) and (2) moreover, the voltage and power losses for real When the RES is connected with the DG, the intermit- power and reactive power could be evaluated as illus- tent nature of RES and the power converter arrange- trates in equations (3) and (4), respectively. ments equipped with the utility grid makes nonlinear DG (NLDG) and generates harmonics in the power net- P2 2 P Q i  P i i i  PL i  R ( ) 1 1 i 1 work system. Very few recent research works related   V 2 i to DG allocation are focused on harmonics mitigation 2 using either active or passive filters [21, 22]. In the re- search work the single tuned filter (STF) is preferred Q Q P Q2 Q  X i i i Li i (2) 1 i 1 V 2 which consist of series connected resistor, inductor and i a capacitor. The THD is considered as the measuring Whereas, Pi and Pi+1 are the real power at the sending as factor for the harmonics analysis and it is included as well as receiving end, Qi and Qi+1 are the reactive power at the sending as well as receiving end, PL(i+1) and QL(i+1) 18 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30 are the real and reactive power losses at the receiving rent induces due to the capacitor placement reduces end and Vi is the voltage at the sending end. the temperature as well as the loss of transmission line. Active and reactive power requirements in distri- 2    2 V 2 V 2 2 2 Pi Qi bution networks can be met by the utilization of DGs i  i  2 Ri Pi  Xi Qi  (R ) 1 i  Xi (3) V 2 i and capacitor allocation. The losses efficiently decrease P2 2 as a result of a reduction in current magnitude. Due to P i Qi Li ,i R   (4) 1 i * V 2 this, higher temperatures have a less destructive effect i on the reliability indices of both overhead and under- Whereas, Ri and Xi are the resistance and reactance of ground lines. Distribution feeder components will have the transmission line. The total loss of the distributed a lower failure rate as a result of these impacts. If the power network is represented in equation (5). The n ith feeder is not equipped with DGs and capacitors, it represents the number of nodes in the network. has an uncompensated failure rate of λuncomp i .In feeder n1 laterals with fully compensated active and reactive cur- PTL PLi i ,  (5) 1 rent components, failure rate drops to λ comp i i . It can be 1 observed that if the active and reactive elements of the 4.1.2 Reliability Index (Obj 2): current are not compensated, the failure rate will be lin- The reliability index is one of the essential terms that early related to the level of compensation. The change might be considered during the development of objec- in real and reactive currents due to the capacitor place- tive function. In this research work, the economical ori- ment is measured by the compensation coefficient ∝i ented reliability parameter which describes the ratio be- (10). The Ireal, and Ireac in equation (10) represents the real tween the expected interruption cost in the presence and and reactive current components respectively. absence of DG, SCs and STF respectively (6) is considered as one of the parameters in the objective function. I new new   reac Ireal i I old * old (10) ECOST reac I R  DG , SCs , STF index (6) real ECOST Moreover, the modified rate of failure after the capacitor Where, expected interruption cost (ECOST) is an ef- placement (λi ) could be derived as shown below (11). new fective tool in system planning which decides the ad-  uncom equate level of reliability for users. Accordingly, the c m i   ew i  p comp o  p i i  i (11) n ECOST without the installation of any DGs and capaci- tors ( ECOSTwithout DG , Cap i ) is evaluated as follows: 4.1.3 Harmonics minimization (Obj 3): The harmonics minimization is computed through To- ECOSTwithout DG , Cap = tal Harmonic Distortion (THD) presented in the signal i N (12) [23]. busLoad C uncomp (7) i1 avgi i i N  H  THD 1  V 2  V hi  (12) ECOSTwithout DG , Cap = i 1  1i h2 i  NbusLoad C uncomp (8) Whereas, V1i is the fundamental bus voltage, Vhi i s the i1 avgi i i order of harmonics at ith bus. Where, Loadavg is the average load in KW, Ci is the inter- i 4.1.4 Voltage profile improvement (Obj 4): ruption cost and λi is the modified rate of failure after new In DG system the power generated from the various the capacitor placement, λuncomp i is the without DG and distributed energy sources are linked to the common capacitor installation case. distributer. The sudden change in load conditions af- fects the VP. Hence, the improvement of voltage profile Following optimal DG and capacitor installation, the is considered as one on the objective in this research cost benefit from reduced ECOST can be expressed as work which is stated in equation (13). following Eq. (9): N 2  V V spec  Voltage p i e  i i _ rof l max min (13 (9) i Vi V  ) 1 i  Whereas, Vi is the ith bus voltage, V spec i is the specified The installation of capacitor has a direct influence on voltage magnitude (1.0 p. u) and V max i , V min i a re the ith the reliability as it moderates the rate of failure. The cur- bus minimum and maximum voltage respectively. 19 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30 4.1.5 Cost minimization for the DGs (Obj 5): (a) Different types of DG have different costs. There will be a variation in the total cost of the DGs as the size and number of the DGs change. To minimize the cost of the DG, the objective can be formulated as follows: (14) (b) Where, CostDG is the DGs’ overall cost, Capcis the capital cost of DG system, LF is the load factor KDG is the size of i the ith DG, CFueli s the size of the ith DG system, and CO&M is the operation and maintenance cost of DG system. (c) 4.2 Hybrid Seagull Optimization Algorithm (HF-SOA) 4.2.1 Fuzzification of objective function parameters This research work is targeted to foreseeing the optimal installation of DG, SCs and STF by satisfying the objec- tive function stated above Eq. (0). In HF-SOA approach initially the fuzzification of this multi- objective function is performed corresponding to the membership func- (d) tions selected. In this proposed problem the trapezoidal membership function is commonly chosen for fuzzifica- tion of all the fitness function parameters such as loss, the VP, ECOST and THD. The membership functions for the proposed research problem are illustrated in Fig3. (i) Fuzzification of Power loss: (e) The power loss is stated in equation (5). The active pow- er loss index (APLI) is evaluated as the ratio between the active power loss in the presence (APLDG , SCs, STF ) and absence (APL) of DG, SCs and STF (15). APL APLI DG , SCs, STF index  (15) APL Figure 3: Fuzzy membership function for the objective Using the trapezoidal membership function depicted function parameters in Fig 3a, the APLI is fuzzified as described in the fuzzy set (16) given below. (ii) Fuzzification of minimum and maximum voltage profiles (VPs): The VP is stated in equation (13). Using the trapezoidal (16) membership function the VP is fuzzified as described in the fuzzy set (17) given below.  0 forV P V PL1   VP VP   L1 Where, APLImin and APLImax a re the minimum maxi- forV PL V P V P  VP VP 1 min min L  mum limits of power loss index. This minimum limit de-  1 VP   1 forV Pmin V P V  (17) pends on the utility requirement and for the proposed  Pmax  problem the maximum value is chosen as 1.  VPL VP   2 forV Pmax V P VPVP VP L   L  2 2 max  20 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30  1 CMDG  CMDG Where, VP min L1 and VPL2 are the minimum and maximum  VP limits of the membership function and VP   CMDGmax CMDG  CMDG =   CMDGmin  CMDG  CMDG (20) min and CMDGmax CMDG max min  VP  max are the primary and secondary limits represented  0 CMDG CMDGmax in Fig 3b. These limits are chosen as VPL1 =0.94, VPmin =0.95, VPmax =1.05 and VPL2 =1.06 respectively. Where, CMDGmin and CMDGmax are the minimum and maximum DG cost. (iii) Fuzzification of Reliability Index (RI): The RI describes the ratio between the expected inter- 4.2.2 Seagull Optimization Algorithm (SOA) ruption cost in the presence and absence of DG, SCs SOA is the most persuasive recently developed me- and STF respectively which is stated in equation (6). Us- taheuristic algorithm which is framed from the inspira- ing the trapezoidal membership function depicted in tion of hunting behavior of the bird Seagulls which are Fig 3c, the RI is fuzzified as described in the fuzzy set technically known as Laridae. It follows two significant (18) given below. strategies called exploration and exploitation to catch the prey. In this research, the proposed SOA algorithm is used to tune the membership function of fuzzy rules. The effectiveness of SOA algorithm has been proved in various engineering studies [23, 24]. The elucidation of (18) hunting strategies of this algorithm with its mathemat- ical modelling is presented in this section. Migration (exploration): Where, RImin and RImax a re the minimum maximum In migration, the mathematical modelling of the parti- limits of reliability index and their values are chosen as cle movement is developed in a grouping pattern with 0.6 and 1.0 for the proposed research problem. respect to each other by satisfy the following three steps. (iv) Fuzzification of total harmonics distortion (THD): - Avoiding collision- To avoid collisions, while moving The THD describes the harmonics presented in the bus  voltage and it is stated in equation (12). This research to the new position Cs ) the current position ( Ps (x)) work aimed to mitigate the harmonics to a minimum of particles are multiplied with the variable A (21). value Using the trapezoidal membership function de- The variable A is represented in equation (10). The picted in Fig 3d, the THD is fuzzified as described in the fc represented in equation (22) regulate the fre- fuzzy set (19) given below. quency of retaining A which is decreased from fc to 0.   Cs  APs x (21) (19)   f  A  f  x* c c  (22)  Max  iteration  Where, THDmin and THDmax are the minimum maximum - Move towards optimal solution- In this step, the limits of harmonic distortion and their values are cho- particles are moving towards its neighbor with sen as 0.25 and 1.0 for the proposed research problem.  best fitness value (23). Where, the M s represent new particle position after moving towards best (v) Fuzzification of Cost minimization for the DGs  (CMDG) solution, Pbs (x) is the particle best solution and As shown in Fig 3(e), the DG cost function has been  modeled as a fuzzy function. The cost minimization Ps (x) is the current position of particle. The B is presented in the bus voltage and it is stated in equation represented in equation (16). The rd in equation (14). Eq. (20) represents the fuzzy membership function (24) states the random number between 0-1. of the cost that is under or equal to the permissible cost as follows:   M  s  B (Pbs x - Ps x) (23) 21 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30 B  2 2 A  rd (24) ing and sizing of energy components presented in this research work. - Stay nearer to the particles with best fitness-In this step the position of particles is updated ac- 5.1 Analysis of HF-SOA in IEEE 33 RDN cording to the best position (25). The computational analysis of RDN is normally tested    Ds  Cs M in reconfigured IEEE test systems. In this research study, s (25) two IEEE test systems with 33 and 69 nodes are con- sidered for analysis purpose. The schematic layout of Attacking (exploitation): standard IEEE 33 bus considered in this research study In this process the Seagulls makes a spiral movement is shown in Fig 1. The DG system considered in this re- in the air to reach the prey. The best positions with op- search work includes 25MW of grid capacity with the timal fitness values are finally achieved in this position fuel cost of 0.044 $/ /kWh, 1 MW of solar PV with the in- with respect to the distance of each particle from the stallation as well as operation and maintenance cost of best fitness position and their movement in x, y and z 3985.0121 $/ /kWh and the 5 MW of wind turbine (WT) plane (26). with the installation as well as operation and mainte- nance cost of 1822.0095 $/ /kWh [19, 26]. The optimal    Ps x  Ds  x y z  Pbs x (26) installation of DG, SCs and STF are analyzed with three case studies as follows: Finally, the optimal solutions are reached in SOA with Case 1: Optimal installation of DG alone a smooth transition between exploration and exploita- Case 2: Optimal installation of DG and SCs tion moves. Case 3: Optimal installation of DG, SCs and STF (i) Case 1 5 Results and discussions The optimal DG installation alone is performed in this case using the HF-SOA algorithm, by averting harmon- ic parameter (Obj 3) in objective function (1) without The optimal installation of DG, SCs and STF is the re- violating the constraints. The harmonic load flow was search problem and the solution is optimized using simulated using MATLAB as a statistical method to an effective proposed HF-SOA algorithm to predict account for uncertainty of input values. This problem the proposed fitness functions (objective function). represents the input data as random values from some The novel fitness function suggested in section 3.1 (1) specified, measured range. To work with and create a is used for this optimization problem. With determin- database for the proposed algorithm, the load flow istic demand of load and DG system, the present arti- method has been modified. A classical harmonic flow cle presents optimum location and size of DG in radial also takes into account the effect of background har- distribution network (RDN) to reduce network loss. The monics in the network. For HF-SOA, SSA and GMSA DG in this study is selected by Locating the bus in the approaches, harmonic load flows are applied to sys- network with the highest level of sensitivity. The study tems with DG in three cases. To estimate different DG examines the placement of three DGs in both net- production losses, voltage drops, and THD online, the works. Loss sensitivity analysis is used to determine the proposed HF-SOA is used. The optimized solutions are location of DG, one bus at a time, by selecting the most denoted in Table 1. These solutions confirms that the sensitive. The next sensitive bus is determined after loss reduction obtained with the proposed HF-SOA is employing DG at the selected bus. In order to find the 68.31% with reference to the base case (without DG in- next sensitive bus, the process is repeated once again. stallation), which is higher than the results of SSA and Therefore, DGs are placed in the respective RDNs based GMSA approaches. Similarly, the maximum and mini- on the loss-sensitive buses in the network. For both 33- mum VP of the network is obtained as 0.9991 p.u and bus and 69-bus RDNs, the procedure identifies the loss- 0.9742 p.u correspondingly which is also higher than sensitive buses. A detail performance study of locating the results of the literature [18, 19]. Hence, the results the DG, SCs and STF in standard IEEE 33 and 69 RDN confirms that the suggested HF-SOA algorithm outper- which are depicted in Fig 1 and Fig 2 is conferred in this forms SSA and GMSA approaches with, higher VP and section. In each IEEE RDN under consideration, three less losses while installing the DG in bus numbers 13, case studies are acquired to perform the proposed op- 17, 24, 30 and 32 of an IEEE 33 bus networks, with the timization problem such as installation of DG alone, in- corresponding sizes denoted in Table 1. stallation of both DG and SCs and installation of DG SCs and STF using the software MATLAB R2022a. Whereas, the term installation in this section represents both sit- 22 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30 Table 1: Case 1: Comparative analysis of optimal DG installation - IEEE 33 bus network Optimization Vmin (p.u) Vmax (p.u) DG size (kW) and Placement Power Loss (kW) Loss of reduction techniques (Bus Number) (%) Base case 0.903 0.997 - 210.98 - Proposed 0.9732 0.9991 541.32 (13), 301.20 (17), 66.84 68.31 HF-SOA 978.74 (24), 505.54 (30), 407.61 (32) SSA [19] 0.9686 0.9988 753.6 (13), 1100.4 (23), 71.456 66.12 1070.6 (29) GMSA [18] 0.9725 0.9988 445.4 (29), 399.1 (10), 439.4 67.97 67.78 (15), 495.3 (25), 495.3 (26), 461.8 (32) ii) Case 2 The SCs are also installed along with DG system in the same 33 node system and their installation criterions are optimized with the proposed HF-SOA algorithm in this case. The objective function preferred in case 1 is used for this case also. The optimal siting and sizing of DG as well as SCs are foreseeing by the HF-SOA and the resultant solutions are illustrated in Table 2. This table reveals that the network losses are considerably dimin- ished to 7.56 kW (96.41%) with the proposed HF-SOA algorithm, which is precisely lesser than the power loss- es developed by SSA and GMSA approaches. Moreover, the total operating cost of the RDN with the proposed algorithm is evaluated as 236.51 $/h which is less than the operating cost of SSA (238.8 $/h). The minimum and maximum VPs are also enhanced to 0.9946 p.u and 1.0012 p.u which are comparatively higher than the al- ternate approaches depicted in Table 2 [18, 19]. Hence, it is confirmed that the HF-SOA could effectively reduce the power losses, operating cost, and maximize the VP compared to SSA and GMSA approaches while analyz- Figure 4: Source voltage waveform and its harmonic ing the DG and SCs installation in IEEE 33 network. spectrum when passive filter connected (iii) Case 3 search articles [17] such as Cm=29.8572µF, Lm=7.7793mH, The STF is installed along with DG and SCs in this case to Rm=0.2056Ω respectively. Fig 4 shows the waveform and support for harmonics mitigation. Whereas, the STF con- harmonic spectrum of the source voltage waveform with tains series connected resistor, inductor and capacitor a THD of 15.2314%, dominated by the 8th, 10th, and 12th and the configurations are directly referred from earlier re- harmonics. The THD of the source voltage and the perfor- Table 2: Case 2: Comparative analysis of optimal DG and SCs installation - IEEE 33 bus network Optimization Vmin (p.u) Vmax Optimal DG installation Optimal SCs installation Power Loss Loss of techniques (p.u) (kW) (kVAr) (kW) reduction (%) Base case 0.903 0.997 - 210.98 - Proposed 0.9946 1.0012 392.14 (13), 738.12 (24), 350 (13), 300 (23), 7.5651 96.41 HF-SOA 999.87 (29), 450 (29), 500 (32) 451.43 (32) SSA [19] 0.9918 1.0010 746.6 (13), 1078.9 (23), 300 (13), 600 (23), 11.8 94.41 1049.2 (29) 1050 (29) GMSA [18] 0.9938 1.0010 418.2 (24), 474.8 (28), 600 (30), 350 (11), 450 7.94 96.24 478.6 (32), 448.7 (11) (31), 150 (14) 23 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30 mance can be improved by appending the passive filter into the existing passive filter system. The optimal installation of filter in IEEE 33 RDN to mini- mize the objective function (1) that includes THD com- ponent is performed with the HF-SOA algorithm in this study. The solutions are presented in Table 3. This table reveals that the optimal location of STF is identified as bus 8 and the minimum and maximum THD values are obtained as 2.3221 and 4.3287 respectively. The graphi- cal bar chart representation of THD level in each bus is Figure 5: Percentage THD of bus voltage in IEEE 33 RDN depicted in Fig 5. The minimum and maximum THDs in the presence and absence of STF in bar chart are high- (a) lighted in brown and blue color in Fig 5. Similarly, the power loss also reduced by 96.48% with respect to the base case after the STF installation, which is compara- tively higher than the power loss in base case and IEEE 33 network with only DG and SCs. In similar manner the minimum and maximum VPs are also enhanced after the installation of STF. The ECOST is also reduced from 316.58 $/h to 252.23 $/h which is nearly reduced to 20.64% with reference to the base case (without DG, SCs and STF) which ensures the reliability of operation after the installation of DG, SCs and STF. All the results (b) clearly reveal that the proposed HF-SOA approach could effectively mitigate the harmonics after the installation of filter. The comparative VPs of Case 1 Case 2 and Case 3 op- timized with HF-SOA are shown in Fig 6(a) which evi- dently illustrates that the VP has been considerably im- proved in Case 3 and Case 2 related to Case 1. Hence, it is confirmed that, the VP is considerably enhanced after the installation of SCs as well as STF. The fig in 6(b) illus- Figure 6: Comparative analysis of different cases in trates the reduction in the cost of DG system based on IEEE 33 RDN of (a) VP (b) DG cost the analysis of the HF-SOA, the GMSA, the SSA and the base case of different cases in IEEE 33 RDN. As a result, is evident from this figure that, fast convergence could the proposed method has a better performance than be obtained with the proposed HF-SOA approach with the existing. minimum power loss in all the cases. The power loss at Case 3 is comparatively lower than other two cases. The convergence of the proposed HF-SOA is examined Hence, it is confirmed that fast convergence with least with the five samples of power loss convergence char- fitness values could successfully obtained with the pro- acteristics for all the three cases. Here, the sample is posed algorithm. considered as DG size and bus number in a test system. The power loss plot over consecutive iterations for five The Power losses, VP as well as the THD values for the samples for Case 1, 2 and 3 are illustrated in Fig 7. It three cases with and without filter are compared using Table 3: Case 3: Comparative analysis of optimal DG, SCs, STF installation - IEEE 33 bus network Condition Vmin (p.u) Vmax (p.u) Filter THDmax THDmin Power Loss Loss of Location (kW) reduction (%) Base case 0.9036 0.9971 - 15.2314 10.5269 210.98 - With DG and SCs 0.9946 1.0010 - 9.6967 7.2924 7.5651 96.41 With DG, SCs and 0.9987 1.0052 8 4.3287 2.3221 7.4278 96.48 STF 24 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30 (a) (a) (b) (b) (c) (c) Figure 7: Power losses convergence - IEEE 33 bus sys- tem (a) Case 1(b) Case 2 (c) Case 3 (d) bar chart analysis as illustrated in Fig 8. It is evidently clear from Fig 8(a) and Fig 8(b) that the minimum and maximum VPs are enhanced in Case 3 after the installa- tion of DG, SCs and STF in IEEE 33 bus network related to Case 1 and 2. In similar manner, the Fig 8(c) reveals that the power loss is also considerably reduced in Case 3 compared to Case 1 and 2. Meanwhile, the com- parative analysis of VP illustrated in Fig 8(b) evidently reveals that the presence of both SCs and STF could Figure 8: Comparative analysis of objective function effectively minimize the THD values. Hence, the Fig 8 parameters in three cases of IEEE 33 bus (a) Minimum ensures that the presence of SCs and filters in power voltage profile (b) Maximum voltage Profile (c) Power network could diminish the harmonics, losses and im- loss (d) THD values prove the VP. In similar manner, maximum THD values with and without filter is presented in Fig 8(c) clearly Case 1: Optimal installation of DG alone confirms the need of filter installation in power net- Case 2: Optimal installation of DG and SCs work for the purpose of harmonic mitigation. Case 3: Optimal installation of DG, SCs and STF 5.2 Performance analysis on IEEE 69 RDN The objective functions considered in the three cases of IEEE 33 bus study is considered in IEEE 69 bus also. The installation of DG, SCs and STF in 69 RDN shown in Fig 2 are optimized in this section using HF-SOA. Simi- (i) Case 1 lar to the section 4.1 three cases are framed to exam- The installation of DG system in a 69 RDN is optimized ine the effectiveness of proposed HF-SOA algorithm in in this case using the proposed HF-SOA approach and foreseeing the installation of DG, SCs and STF integrat- the solutions are represented in Table 4 which, reveals ed in an IEEE 69 network. that the optimized siting of DG in IEEE 69 RDN using proposed HF-SOA approach are 17, 21, 42, 61 and 62 respectively. Meanwhile, the percentage reduction in 25 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30 power losses is also stated as 70.25% which is higher also. The bar chart representation of THD in each bus than the SSA and GMSA algorithms. In the same way in this optimization process is depicted in Fig 9 which the minimum and maximum VPs are also improved to demonstrates that the optimal location of STF is identi- 0.9732 and 0.9991 respectively which is higher than fied as bus 37 and the minimum and maximum VP val- the results of GMSA and SSA algorithms [18, 19]. Table ues are obtained as 0.9987 and 1.0032 respectively. 4 ensures that the proposed HF-SOA outperforms the The minimum THDs with and without STF in the bar earlier published articles. Hence, the dominance of the chart representation are highlighted in brown and blue proposed HF-SOA is confirmed in this case study. color in Fig 9. For detailed analysis the minimum and maximum THD values are from this figure and listed in (ii) Case 2 Table 6. This table clearly describes that the harmon- Both DG and SCs are integrated in IEEE 69 bus system ics are mitigated after the installation of STF in the bus in this case, and their size as well as locations are opti- number 37 which is optimized with the proposed intel- mized with HF-SOA algorithm. The resultant solutions ligent algorithm. It is also revealed from the table that, are shown in Table 5. This table reveals that the optimal the minimum and maximum THD values of the pow- locations of DG in IEEE 69 bus system are 17, 23, 61 and er network without DG, SCs and STF are 9.5126 and 63 and the locations of SCs are 16, 35, 23 and 61, re- 17.2562 and this value has been reduced by 77.22 % spectively. It is also revealed that the percentage power and 49.92 % after the installation of DG and SCs which loss reduction (97.79%) is high in proposed approach is further reduced to 84.49 % and 73.32 % after the in- compared to GMSA approach. The minimum and maxi- stallation of STF. In addition, the losses of the network mum VPs are also enhanced to 0.9986 and 1.0021using also minimized by 97.83 % compared to base case. The the proposed hybrid algorithm, which is higher than ECOST is also reduced from 421.65 $/h to 314.23 $/h the results documented earlier in the literature [18, 19]. which is nearly reduced to 25.47% related to the base From this case study, the dominance of HF-SOA over case. Thereby, the reliability is also ensured. This case SSA and GMSA in optimizing the siting and sizing of DG study confirms the dominance of proposed HF-SOA ap- and SCs in IEEE 69 bus system is evidently proved. proach over SSA and GMSA in terms of power loss, har- monics minimization and VP and reliability improve- (iii) Case 3 ment. The installation of STF along with DG and SCs in IEEE 69 RDN is discussed in this case for effective mitigation The comparative voltage profiles of the three cases op- of harmonics. The STF with same configurations as dis- timized using proposed HF-SOA approach is depicted cussed in IEEE 33 case study is considered in this case in Fig 10 (a). This figure clearly describes that the VP is Table 4: Case 1: Comparative analysis of optimal DG installation - IEEE 69 bus network Optimization Vmin (p.u) Vmax (p.u) DG size (kW) and Placement Power Loss Loss of techniques (Bus Number) (kW) reduction (%) Base case 0.909 0.999 - 224.98 - Proposed 0.9802 1.0010 341.32 (17), 524.65 (21), 387.23 66.9314 70.25 HF-SOA (42), 542.34 (61), 687.32 (62) SSA [19] 0.9789 1.0003 380 (17), 527 (10), 1718 (60) 69.41 69.14 GMSA [18] 0.9725 0.9988 359.8 (53), 282.2 (67), 100.1 (42), 67.79 69.87 281.07 (62), 307.04 (60) Table 5: Case 2: Comparative analysis of optimal DG and SCs installation - IEEE 69 bus network Optimization Vmin Vmax Optimal DG installation (kW) Optimal SCs installation Power Loss Loss of techniques (p.u) (p.u) (kVAr) (kW) reduction (%) Base case 0.909 0.999 - - 224.98 - Proposed 0.9986 1.0021 592.14 (17), 320.89 (23), 420 (16), 490 (35), 650 4.9623 97.79 HF-SOA 951.77 (61), 851.65 (63) (23), 750(61) SSA [19] 0.9971 1.0010 358 (19), 518 (10), 600 (11), 600 (48), 4.837 97.85 1673.5 (60) 1200 (60) GMSA [18] 0.9976 1.000 346.5 (69), 383.2 (18), 446.4 450 (50), 150 (48), 450 5.093 97.74 (62), 360.7 (58) (61), 1200 (23), 150 (10) 26 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30 Table 6: Case 3: Comparative analysis of optimal DG, SCs, STF installation - IEEE 69 bus network Condition Vmin (p.u) Vmax (p.u) Filter THDmax THDmin Power Loss Loss of Location (kW) reduction (%) Base case 0.909 0.999 - 17.2562 9.5126 224.98 - With DG and SCs 0.9986 1.0021 - 8.6426 2.1669 4.9623 97.79 With DG, SCs and STF 0.9987 1.0032 37 4.6032 1.4748 4.8656 97.83 considerably increased in third case compared to other algorithm with minimum fitness value. In addition, it two cases. Further, the convergence analysis of HF-SOA is also observed that the fitness value of power loss is also examined with IEEE 69 bus network by examin- in Case 3 is less when compared to other two cases. ing the most fundamental parameter of objective func- Hence, the convergence ability of the HF-SOA ap- tion called power loss for the three cases. The figure in proach is proved while optimizing the installation of 10(b) illustrates the reduction in the cost of DG system DG, SCs and STF in the 69 RDN. based on the analysis of the HF-SOA, GMSA, SSA and the base case of different cases in IEEE 69 RDN. As a re- The comparative analysis of minimum and maximum sult, the proposed HF-SOA method has a better perfor- VPs and power losses for the three cases as well as the mance than the existing. THD values with and without filter are presented with the bar chart representation as depicted in Fig 12. Fig The power loss plot over consecutive iterations for five 12(a) and Fig 12(b) describes that the minimum and samples for Case 1, 2 and 3 are illustrated in Fig 11. This maximum VPs are enhanced in Case 3 after the instal- figure ensures the earliest convergence of proposed lation of DG, SCs and STF in IEEE 69 bus network relat- ed to Case 1 and 2. In similar manner, it is understood from Fig 12(c) that the power loss is also considerably reduced in Case 3 compared to Case 1 and 2. In the same way, the Fig 12(d) confirms the mitigation of har- (a) Figure 9: Percentage THD of bus voltage in IEEE 69 RDN (a) (b) (b) (c) Figure 10: Comparative analysis of different cases in Figure 11: Convergence of power losses in IEEE 69 bus IEEE 69 RDN of (a) VP (b) DG cost system (a) Case 1(b) Case 2 (c) Case 3 27 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30 (a) a distributed network. The reduction of power losses, harmonics, and enhancement of VP and reliability are considered as the fundamental parameters in the pro- posed novel multi-objective function of this proposed research problem. Three case studies are conducted in standard IEEE network with 33 and 69 nodes, to exam- ine the effectiveness of proposed intelligent algorithm. The results optimized with the HF-SOA in IEEE 33 bus system ensures that the power losses in Case 1, Case 2 and Case 3 are reduced by 68.31%, 96.41% and 96.48% (b) respectively when compared to the base case. In the same way, the power losses are minimized by 70.25%, 97.79% and 97.83% compared to base case while op- timizing with proposed intelligent algorithm. In addi- tion, the minimum and maximum THD values are also reduced by 77.94% and 71.58% in IEEE 33 bus system and 84.49 % and 73.32 % in IEEE 69 bus system. The re- liability enhancement is also proved with ECOST mini- mization which is reduced by 20.64% in 33 node and 25.47% in 69 node system. The effective VP enhance- ment is also achieved through the proposed intelligent (c) algorithm. The results of convergence analysis revealed that earlier convergence could be obtained by the pro- posed HF-SOA approach with minimum power loss. Hence, with the proposed novel HF-SOA algorithm based on the novel fitness function, the optimal size and allocation of DG, SCs and STF are obtained suc- cessfully. This research work could be extended in near future with the following suggestions: - In this research article, only the ECOST is consid- ered in the objective function to improve the reli- (d) ability however, the other cost parameters such as energy not supplied (ENS) could also be con- sidered. - In this article the filter parameters are referred from literature instead, they could be optimized along with siting of filters. 7 Acknowledgments Figure 12: Comparative analysis of objective function The Author with a deep sense of gratitude would thank parameters in three cases of IEEE 69 bus (a) Minimum the supervisor for his guidance and constant support voltage profile (b) Maximum voltage Profile (c) Power rendered during this research. loss (d) THD values monics after the installation of single tuned filter in Case 3. Hence the effectiveness of proposed HF-SOA in 8 Conflict of Interest optimizing the siting and sizing of DG, SCs and STF is confirmed by this analysis. Authors do not have any conflicts. 6 Conclusion 9 References A novel HF-SOA algorithm is proposed in this arti- 1. T. Yuvaraj, K. Devabalaji, and K. Ravi, “Optimal al- cle to optimize the installation of DG, SCs and STF in location of DG in the radial distribution network 28 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30 using bat optimization algorithm,” in Advances in densers in distribution networks,” International power systems and energy management: Spring- Transactions on Electrical Energy Systems., vol. er., pp. 563-569, 2018, 29, no. 3, p. e2755, 2019, https://doi.org/10.1007/978-981-10-4394-9_55. https://doi.org/10.1002/etep.2755. 2. K. Varesi, “Optimal allocation of dg units for power 11. A. Naderipour et al., “Spotted hyena optimizer loss reduction and voltage profile improvement algorithm for capacitor allocation in radial distri- of distribution networks using PSO algorithm,” bution system with distributed generation and International Journal of Computer and Systems microgrid operation considering different load Engineering., vol. 5, no. 12, pp. 1847-1851, 2011. types,” Scientific reports., vol. 11, no. 1, pp. 1-15, 3. S. R. Ghatak and P. Acharjee, “Optimal allocation 2021, of DG using exponentential PSO with reduced https://doi.org/10.1038/s41598-021-82440-9. search space,” in 2016 second international con- 12. S. C. Reddy, P. Prasad, and A. J. Laxmi, “Placement ference on computational intelligence & commu- of distributed generator, capacitor and DG and nication technology (CICT)., 2016, pp. 489-494: capacitor in distribution system for loss reduction IEEE, and reliability improvement,” Editors-in-Chief., p. https://doi.org/10.1109/CICT.2016.103. 198, 2013. 4. H. HassanzadehFard and A. Jalilian, “Optimal siz- 13. R. Baghipour and S. M. Hosseini, “Placement of DG ing and location of renewable energy-based DG and capacitor for loss reduction, reliability and units in distribution systems considering load voltage improvement in distribution networks growth,” International Journal of Electrical Power using BPSO,” International Journal of Intelligent & Energy Systems., vol. 101, pp. 356-370, 2018 Systems and Applications., vol. 4, no. 12, p. 57, https://doi.org/10.1016/j.ijepes.2018.03.038. 2012, 5. E. A. Al-Ammar et al., “ABC algorithm based opti- https://doi.org/10.5815/ijisa.2012.12.08. mal sizing and placement of DGs in distribution 14. R. Baghipour and S. M. Hosseini, “Optimal place- networks considering multiple objectives,” Ain ment and sizing of distributed generation and Shams Engineering Journal., vol. 12, no. 1, pp. capacitor bank for loss reduction and reliabil- 697-708, 2021, ity improvement in distribution systems,” Majlesi https://doi.org/10.1016/j.asej.2020.05.002. Journal of Electrical Engineering, vol. 7, no. 3, pp. 6. M. G. Hemeida, A. A. Ibrahim, A.-A. A. Mohamed, 59-66, 2013. S. Alkhalaf, and A. M. B. El-Dine, “Optimal alloca- 15. M. Salari and F. Haghighatdar-Fesharaki, “Opti- tion of distributed generators DG based Manta mal placement and sizing of distributed genera- Ray Foraging Optimization algorithm (MRFO),” tions and capacitors for reliability improvement Ain Shams Engineering Journal, vol. 12, no. 1, pp. and power loss minimization in distribution net- 609-619, 2021, works,” Journal of Intelligent Procedures in Electri- https://doi.org/10.1016/j.asej.2020.07.009. cal Technology., vol. 11, no. 43, pp. 83-94, 2020. 7. S. G. Naik, D. Khatod, and M. Sharma, “Optimal 16. P. Reddy, V. Reddy, and T. G. Manohar, “Whale op- allocation of combined DG and capacitor for timization algorithm for optimal sizing of renew- real power loss minimization in distribution net- able resources for loss reduction in distribution works,” International Journal of Electrical Power & systems,” Renewables: wind, water, and solar., vol. Energy Systems., vol. 53, pp. 967-973, 2013, 4, no. 1, pp. 1-13, 2017, https://doi.org/10.1016/j.ijepes.2013.06.008. https://doi.org/10.1186/s40807-017-0040-1. 8. S. R. Gampa and D. Das, “Simultaneous optimal 17. V. R. Pandi, H. Zeineldin, and W. Xiao, “Passive har- allocation and sizing of distributed generations monic filter planning to overcome power quality and shunt capacitors in distribution networks us- issues in radial distribution systems,” in 2012 IEEE ing fuzzy GA methodology,” Journal of Electrical Power and Energy Society General Meeting, 2012, Systems and Information Technology., vol. 6, no. pp. 1-6: IEEE, 1, pp. 1-18, 2019, https://doi.org/10.1109/PESGM.2012.6345247. https://doi.org/10.1186/s43067-019-0003-2. 18. E. A. Mohamed, A.-A. A. Mohamed, and Y. Mitani, 9. T. Yuvaraj et al., “Optimal integration of capacitor “Hybrid GMSA for Optimal Placement and Sizing and distributed generation in distribution system of Distributed Generation and Shunt Capacitors,” considering load variation using bat optimization Journal of Engineering Science & Technology Re- algorithm,” Energies., vol. 14, no. 12, p. 3548, 2021, view., vol. 11, no. 1, 2018, https://doi.org/10.3390/en14123548. https://doi.org/10.25103/jestr.111.07. 10. A. R. Abul’Wafa, “Ant‐lion optimizer‐based mul- 19. K. S. Sambaiah and T. Jayabarathi, “Optimal allo- ti‐objective optimal simultaneous allocation of cation of renewable distributed generation and distributed generations and synchronous con- capacitor banks in distribution systems using salp 29 I. Rajagopalan et al.; Informacije Midem, Vol. 53, No. 1(2023), 15 – 30 swarm algorithm,” International journal of renew- 29. H. Lotfi, “Optimal sizing of distributed genera- able energy research., vol. 9, no. 1, pp. 96-107, tion units and shunt capacitors in the distribution 2019. system considering uncertainty resources by the 20. J. Liu et al., “Modelling and analysis of radial modified evolutionary algorithm,” Journal of Am- distribution network with high penetration of bient Intelligence and Humanized Computing., renewable energy considering the time series vol. 13, no. 10, pp. 4739-4758, 2022. characteristics,” IET Generation, Transmission & https://doi.org/10.1007/s12652-021-03194-w Distribution., vol. 14, no. 14, pp. 2800-2809, 2020, 30. F. García-Muñoz, F. Díaz-González, and C. Corche- https://doi.org/10.1049/iet-gtd.2019.1874. ro, 2021. “A novel algorithm based on the combi- 21. A. Lakum and V. Mahajan, “Optimal placement nation of AC-OPF and GA for the optimal sizing and sizing of multiple active power filters in radial and location of DERs into distribution networks,” distribution system using grey wolf optimizer in Sustainable Energy, Grids and Networks., vol. 27, presence of nonlinear distributed generation,” p.100497.s, Electric Power Systems Research., vol. 173, pp. https://doi.org/10.1016/j.segan.2021.100497 281-290, 2019, 31. S.R. Salkuti, “Optimal allocation of DG and D-STAT- https://doi.org/10.1016/j.epsr.2019.04.001. COM in a distribution system using evolutionary 22. B. Singh, K. Al-Haddad, and A. Chandra, “A review based Bat algorithm,” International Journal of Ad- of active filters for power quality improvement,” vanced Computer Science and Applications., vol. IEEE transactions on industrial electronics., vol. 12, no. 4, 2021, 46, no. 5, pp. 960-971, 1999, https://doi.org/10.1109/41.793345. Copyright © 2023 by the Authors. 23. A. S. Abbas et al., “Optimal harmonic mitigation in This is an open access article dis- distribution systems with inverter based distrib- tributed under the Creative Com- uted generation,” Applied Sciences., vol. 11, no. 2, mons Attribution (CC BY) License (https://creativecom- p. 774, 2021, mons.org/licenses/by/4.0/), which permits unrestricted https://doi.org/10.3390/app11020774. use, distribution, and reproduction in any medium, 24. A. Appathurai, et al., “An efficient optimal neural provided the original work is properly cited. network-based moving vehicle detection in traffic video surveillance system,” Circuits, Systems, and Signal Processing., vol.39, no.2, pp.734-756, 2020, Arrived: 15. 08. 2022 https://doi.org/10.1007/s00034-019-01224-9. Accepted: 30. 12. 2022 25. P. Samal, “Optimal DSTATCOM Allocation Using a Hybrid Seagull-Differential Evolution Algorithm,” in 2020 IEEE 17th India Council International Con- ference (INDICON)., 2020, pp. 1-6: IEEE, https://doi.org/10.1109/INDICON49873.2020.9342524. 26. A. A. Abou El-Ela, R. A. El-Sehiemy, and A. S. Ab- bas, “Optimal placement and sizing of distributed generation and capacitor banks in distribution systems using water cycle algorithm,” IEEE Sys- tems Journal., vol. 12, no. 4, pp. 3629-3636, 2018, https://doi.org/10.1109/JSYST.2018.2796847. 27. A. Rathore, and N.P. Patidar, “Optimal sizing and allocation of renewable based distribution gen- eration with gravity energy storage considering stochastic nature using particle swarm optimiza- tion in radial distribution network,” Journal of En- ergy Storage., vol. 35, pp. 102282, 2021, https://doi.org/10.1016/j.est.2021.102282. 28. R. Siddique, S. Raza, A. Mannan, L. Khalil, N. Alwaz, and M. Riaz, “A modified NSGA approach for opti- mal sizing and allocation of distributed resources and battery energy storage system in distribution network,” Materials Today: Proceedings., vol. 47, pp. S102-S109, 2021. https://doi.org/10.1016/j.matpr.2020.05.669. 30 Original scientific paper https://doi.org/10.33180/InfMIDEM2023.103 Journal of Microelectronics, Electronic Components and Materials Vol. 53, No. 1(2023), 31 – 37 Liquid Metal Droplet Tunable RF MEMS Inductor Issam El Gmati1,2, Ridha Ghayoula3 1College of Engeneering Al Qunfudha Umm al Qura University, KSA 2Higher school of sciences and technology of Hammam Sousse, Tunisia 3Department of Electrical and Computer Engineering, Laval University, Quebec City, Canada Abstract: A new variable inductor has been simulated, manufactured and tested. The idea is based on changing the morphology of a Galinstain droplet by electrostatic actuation. A drop of 200 µm diameter is used and the applied voltage is limited to 100 V. We demonstrate a tunable inductor that simultaneously achieves wide tuning range of 400 % with high inductance from 1.8 nH to 9.1 nH with a measured quality factor of 26 at 2 GHz and the self-resonance frequency is 4 GHz. Their results were compared. The conclusion showed that simulation results matched well with measurement. The Comparison between our work and other published works show excellent performance. Keywords: Tunable inductor, radiofrequency, MEMS, droplet, Galinstain Nastavljiva mikrofluidna RF MEMS tuljava Izvleček: Simuliran, izdelan in preizkušen je bil nova spremenljiva tuljava. Zamisel temelji na spreminjanju morfologije kapljice Galinstain z elektrostatičnim vzbujanjem. Uporabljena je kapljica s premerom 200 µm, napetost pa je omejena na 100 V. Prikazali smo nastavljivo tuljavo, ki hkrati dosega široko območje nastavitve 400 % z visoko induktivnostjo od 1,8 nH do 9,1 nH z izmerjenim faktor- jem kakovosti 26 pri 2 GHz, samorezonančna frekvenca pa je 4 GHz. Njihovi rezultati so bili primerjani. Zaključek je pokazal, da se rezul- tati simulacije dobro ujemajo z meritvami. Primerjava med našim delom in drugimi objavljenimi deli je pokazala odlično učinkovitost. Ključne besede: nastavljiva tuljava, radio frekvenca , MEMS, kaplica, Galinstain * Corresponding Author’s e-mail: iagmati@uqu.edu.sa 1 Introduction goals, it is important to replace fixed components with Passive RF MEMS (Radiofrequency Micro-Electro-Me- variable components. [17]. Variable inductors have the chanical Systems) components play a primary role in same performance as fixed inductors; to vary induct- modern transceiver and receiver architectures. They ance. We can play on several geometric or technologi- provide significant gains in terms of miniaturization, cal parameters as well as on the variation techniques motivating performance in the gigahertz band as well [18]. In the work [19] and in order to vary the induct- as low power consumption. Many MEMS components ance having discrete values, the researchers used have been developed for radio frequency application; micro-relays. Other techniques have been invented in particular inductors. Micro inductors are used in, by modifying the magnetic flux [20-21], or by playing for example, RF MEMS [1–8], micro-actuators [9-11], on mutual inductance [22-23]. Other works have pro- bio-sensors [12], micro-actuators [13], power MEMS posed the micro-fluidic action by using liquids [24-26]. [14], energy harvesters [15], transformers and elec- The common objective of all this work is to achieve a tromagnetic motors [16]. Most inductors have fixed good variation of the inductance, a good quality factor inductance value. The new generations of communi- as well as a high resonant frequency in the gigahertz cations systems, on the other hand, aim to use very band allowing the integration of these components in wide frequency bands to simplify these systems, re- the application devices. Until today, none of the pub- duce the number of transmission / reception channels, lished structures have corresponded to our objectives, and consequently reduce their costs. To achieve these namely a high variation ratio, a good quality factor, a How to cite: I. El Gmati et al., “Liquid Metal Droplet Tunable RF MEMS Inductor", Inf. Midem-J. Microelectron. Electron. Compon. Mater., Vol. 53, No. 1(2023), pp. 31–37 31 I. El Gmati et al.; Informacije Midem, Vol. 53, No. 1(2023), 31 – 37 variable inductance working in the gigahertz bands. In order to achieve our goals, we were able to design a new method to vary the inductance. The idea is to slide a droplet of Galinstain horizontally above the inductor and control it by electrostatic actuation. The structure is manufactured and tested. Good performances have been shown. 2 Design of inductance Figure 1 shows the model of the designed inductance and the principle used to vary the inductance. The structure consists of a planar spiral inductor and a liq- uid metal such as Galinstain. The metal drop is placed above the spiral inductor. Figure 2: Schematic cross section of the fabricated in- ductor. Figure 1: Model of the designed inductance and prin- ciple used to vary the inductance. The planar inductor fabricated on a glass substrate; de- Figure 3: Photography of one inductor under fabrica- signed in double circular form with 3 turns with a coil tion. 20 μm wide, spaced 20 μm. The length of the internal diameter is 600 μm so the external one is 1200 μm. The Table 1 shows physical properties of materials used in basic idea is to place a drop above the metal coils and fabrication process. thus modify the geometric parameters of the induc- tance by short-circuiting a portion of the coil. This al- Table 1: Physical properties of materials used in fabri- lows a reduction in the current path and consequently cation process. a change in the inductance value. Figure 2 shows the steps followed in a clean room to fabricate the induc- tor. The gold coils were deposited on a glass substrate. This manufacturing process requires 3 masks. A Ti/Au (500 Å/500 Å) seed layer was regularly popped on highest side of a 500 μm thick glass substrate. The Ti layer was placed to improve the bond, and a posi- Glass 5.5 0.0037 1 0 tive photoresist (AZ 4562) was next spotted in order to Gold 1 0 0.99996 41E6 form the electroplating mould (thickness ≈ 5.5 μm). A 2 μm thick gold were electroplated into the resist mould. The photoresist mould was then removed and the seed layer was chemically etched. Finaly, a droplet of 200 μm 3 Results and discussion is used and the applied bias voltage is limited to 100 V. Figure 3 shows a top view of the circular shaped induc- This section may be divided by subheadings. It should tor made of gold on a glass substrate. provide a concise and precise description of the experi- 32 Materials Relative Permittivity Dielectric Loss Relative Permeability Bulk Conductivity (S/m) I. El Gmati et al.; Informacije Midem, Vol. 53, No. 1(2023), 31 – 37 mental results, their interpretation, as well as the exper- As shown in figure 6 (a) the inductor has a peak of 25 at imental conclusions that can be drawn. The measure- 2 GHz, and the self-resonance frequency is 4 GHz. The ments shown in figure 4 were carried out under spikes quality factor decreases from Qmax = 26 when the ap- with a vector network analyzer (VNA ‘Vector Network plied voltage is 100 V to Qmin = 12 when it is equal to Analyzer’) HP 8510 which has a frequency range of 50 0 V. As demonstrated in figure 6(b), the measured in- MHz-13 GHz using GSG ‘Ground Signal Ground’ micro- ductance at high frequency increases continuously be- probe RF tips. The inductance L and the quality factor Q tween Lmin =1.8 nH and Lmax = 9.1 nH at 2 GHz by apply- were calculated by using the following equations [27] ing voltage. A 400 % tuning range is achieved at 2 GHz. L 1  Im 1   ( 2 f Y  1) (a)  11    Im 1 Y  Q   11  (2)   Re 1 Y   11  Yind, Ymeas and Yopen are the two-port admittance matrix the admittance matrices of the measured inductor and the open pattern, respectively. (b) Figure 4: Test bench for S-parameters characterization. During the measurements droplets of Galinstain [28] Figure 6: Measured results a Quality factor by frequen- with conductivity of the order than 3.46 106 S/m at 20º cy (a) and inductance (b) of a fabricated tunable RF C were used. The inductance varies by touching the MEMS inductor by applied voltage. liquid metal droplet upon the spiral inductor shown in figure 5. The Galinstain droplet is flattening out as HFSS (3D Electromagnetic Field Simulator for RF and applying a voltage. Measurements were taken for 6 Wireless Design) software was used to simulate an in- voltage levels applied to the liquid metal droplet for a ductor 3D model structure. Figures 6 show the varia- frequency span of 100 MHz to 10 GHz. tions of the measured results as a function of the volt- age at a frequency of 2 GHz. When the Galinstain droplet is present on the metal coils, it short-circuits a portion of the coil allowing a re- duction in the number of coils of the inductance and subsequently a decrease in the value of the induct- ance. On the other hand, this reduction in the number of turns allows an increase in the quality factor shown in figure 7(a). The inductance L varies proportionally as a function of the number of coils N. The variation in the quality factor is inversely proportional to N. Figure 5: Galinstain droplet placed above the spiral in- ductor for different applied voltage. 33 I. El Gmati et al.; Informacije Midem, Vol. 53, No. 1(2023), 31 – 37 Table 2: Performance comparison published works (a) and our work. 6 1.1 46 47.5 [29] 1 8.86 25.5 27.7 [30] 2.5 3.3 >20 230 [31] 2 0.93 0.96 10 [32] 1.5 1.2 2.87 191 [33] 4 0.3 8 60 [34] 5 1.2-2.3 38.2 90 [35] (b) 4 2.8 18 380 [36] 2 1.8-9.1 26 400 This work We can notice that our inductor showed good perfor- mance in terms of variation, quality factor as well as reasoning frequency by comparing it with other works. The inductance produced is an ambitious solution in terms of small size in micrometers, energy consump- tion, low cost meeting the objectives set at the begin- ning such as a good inductance variation of the order of Nano-henry, a good quality factor in the Gigahertz frequency band and a resonance frequency of the or- der of Giga-hertz. The works cited in Table 1; not been able to group all these constraints together. To imple- Figure 7: Comparison between simulated and meas- ment this inductor in real applications, you would need ured inductance (a) and Quality Factor (b) of the tun- a wrapper or something to keep the liquid metal from able inductance at 2 GHz by applied voltage. moving around. Variations in inductance shown in figure 7 (b) values The inductance is modelled by Ls, and the finite con- have been visualized for different Galinstain contact ductivity of the metal is represented with the series pins on metal coils. You can notice a slight difference of resistance Rs. The capacitive coupling of the windings around 5% between the measurement results and the between the input and output is modelled by Cs shown simulation results. This can be attributed to other fac- in Figure 8. tors due to the existence of an external potential. The position of the liquid droplet above the metal coil may cause this slight difference. In fact, the drop can either be in direct contact with the gold turns, or it can form a capacitive contact through a thin layer. In the presence of the Galinstain droplet, the mag- Figure 8: Electric Equivalent model for tunable induc- netic flux of the inductance infiltrates the Galinstain tor. droplet causing the creation of an eddy current induc- ing a counteractive magnetic field according to Lenz’s The compact model was fitted to the measured S-pa- law causing magnetic loss. When the droplet is moved rameters. The fitted model parameters for different ap- away from the metal spires, the inductance exhibits plied voltage are listed in Table 3. very low loss. A comparison of the measurement results of this vari- able inductance with others published and presented in Table 2. 34 Operating Frequency GHz Inductance Quality fac tor Tunability% Ref I. El Gmati et al.; Informacije Midem, Vol. 53, No. 1(2023), 31 – 37 Table 3: Fitted compact model parameters from Meas- 6 Conflict of Interest ured. The authors declare no conflict of interest Frequency Applied Ls (nH) Rs (Ω) Cs (fF) GHz voltage 2 0 9.1 10 0.83 2 20 8 13 0.85 7 References 2 40 5.25 15 0.87 1. Yao JJ. RF MEMS from a device perspective. Jour- 2 60 4.3 17 0.91 nal of Micromechanics and Microengineer- ing 2 80 3.1 20.4 0.98 2000; 10: 9–38. 2 100 1.8 24 1.01 2. Yoon JB, Kim BK, Han CH et al. Surface microma- chined solenoid on-Si and on glass inductors for The fitted Ls correspond to the measured results in Fig- RF applications. IEEE Electron De- vice Letters ure 6 (b). The Rs does fluctuate with injected fluid, in- 1999; 20: 487–489. dicating that the proximity effect of neighboring con- 3. Kral A, Behbahani F, Abidi AA RF-CMOS oscillators ductors induces some losses. The compact model of with switched tuning. Proceedings of the IEEE in the inductor serves as a reasonable approximation for Custom Integrated Cir-cuits Con- ference; Santa low frequency performance and gives some insight on Clara, CA; 1998: 555-558. the effects of the parasitic. The Rs does fluctuate with 4. Young DH, Malba V, Ou JJ et al. A low-noise RF applied voltage, indicating that the proximity effect of voltage-controlled oscillator using on-chip high neighboring conductors induces some losses. The Cs is Q three dimensional coil inductor and micro ma- very low since there is no coupling with an underlying chined variable capacitor. Proceedings of the Sol- metal layer, and the crosstalk between adjacent turns is id-State Sensor and Actuator Work- shop; Cleve- not significant. land, OH, USA; 1998: 128-131. 5. J. J. Yao, “RF MEMS from a device perspective,” Journal of Micromechanics and Micro engineer- 4 Conclusion ing, vol. 10, no. 4, pp. 9–38, Dec. 2000. https://doi.org/10.1088/0960 -1317/10/4/201. In this work a new variable inductor was designed, 6. J.-B. Yoon, B.-K. Kim, C.-H. Han, E. Yoon, and C.-K. manufactured and tested. A liquid metal was used to Kim, “Surface micro machined solenoid on-Si and vary the inductance. The fabricated tunable RF induc- on-glass inductors for RF applications,” IEEE Elec- tor demonstrates a wide tuning range of 400%. Induct- tron Device Letters., vol. 20, no. 9, pp. 487–489, ance varied between 1.8 nH and 9.1 nH with a meas- Sep. 1999. ured quality factor of 26 at 2 GHz, and the resonant https://doi.org/10.1109/55.784461. frequency is 4 GHz. Variable inductance can be exploit- 7. Yasser Mafinejad, Abbas Kouzani, Khalil Mafine- ed in the design of mobile communication systems zhad, «Review of low actuation voltage RF MEMS for new compatibility and improved electrical perfor- electrostatic switches based on metallic and mance over a wide range and frequency. The Inductor carbon alloys», Journal of Microelectronics, Elec- considered ambitious in terms of small size, power con- tronic Components and Materials, Vol. 43, No. 2 sumption and low cost. To implement this inductor in (2013), 85 – 96. real applications, you would need some packaging or 8. Alireza Ardehshiri, Gholamreza Karimi, Ramin something to prevent liquid metal from moving. Dehdasht-Heydari, «Optimization of shunt capacitive RF MEMS switch by using NSGA-II algorithm and utility algorithm» Journal of Micro- electronics, Electronic Components and Materi- 5 Acknowledgments als, Vol. 49, No. 1(2019), 43 – 50,  9. Ahn CH, Allen MG. A planar micro machined spiral The authors would like to thank the Deanship of Scien- inductor for integrated magnetic micro actuator tific Research at Umm Al-Qura University. applications. Journal of Micromechanics and Mi- cro engineering 1999; 3: 37–44. Funding: The authors would like to thank the Dean- 10. Fulcrand R, Bancaud A, Escriba C et al. On chip ship of Scientific Research at Umm Al-Qura Uni- magnetic actuator for batch-mode dynamic ma- versity for supporting this work by GRANT Code nipulation of magnetic particles in compact lab- 23UQU4361156DSR02 on-chip. Sensors and Actuators B: Chemical 2011; 160: 1520–1528. 35 I. El Gmati et al.; Informacije Midem, Vol. 53, No. 1(2023), 31 – 37 11. Olivo J, Carrara S, De Micheli G. Micro-fabrication 23. J. M. Dell, K. Winchester, C. A. Musca, J. Antosze- of high-thickness spiral inductors for the remote wski, and L. Faraone, “Variable MEMS based in- powering of implantable biosensors. Microelec- ductors fabricated from PECVD silicon nitride,” in tronic Engineering 2014; 113: 130–135. Proc. Conf. Optoelectron. Microelectron. Mater. 12. S. S. Bedair, J. S. Pulskamp, C. D. Meyer, M. Mira- Devices, Dec. 2002, pp. 567–570. belli, R. G. Polcawich, and B. Morgan, “High-perfor- 24. V. M. Lubecke, B. Barber, E. Chan, D. Lopez,M. E. mance micro machined inductors tunable by lead Gross, and P. Gammel, “Self-assembling MEMS zirconate titanate actuators,” IEEE Electron Device variable and fixed RF inductors,” IEEE Trans. Mi- Lett., vol. 33, no. 10, pp. 1483–1485, Oct. 2012. crow. Theory Techn., vol. 49, no. 11, pp. 2093– https://doi.org/10.1109/LED.2012.2207700. 2098, Nov. 2001. 13. M. B. Coskun, K. Thotahewa, Y.-S. Ying, M. Yuce, 25. B. Assadsangabi, M. S. Mohamed Ali, and K. Taka- A. Neild, and T. Alan, “Nanoscale dis- placement hata, “Ferrofluidbased variable inductor,” in Proc. sensing using micro fabricated variable induct- IEEE 25th MEMS Conf., Jan./Feb. 2012, pp. 1121– ance planar coils,” Appl. Phys. Lett., vol. 103, Oct. 1124. 2013, Art. no. 143501. 26. I. E. Gmati et al., “Fabrication and evaluation of an https://doi.org/10.1063/1.4823828. on-chip liquid micro-variable inductor,” J. Micro- 14. Araghchini M, Member S, Chen J et al. A technol- mech. Microeng., vol. 21, no. 2, p. 025018, 2011. ogy overview of the power chip development 27. Jyh-Chyurn Guo and Teng-Yang Tan, “A Broad- program. IEEE Transactions on Power Electronics band and Scalable Modelfor On-Chip Inductors 2013; 28: 4182–4201 Incorporating Substrate and Conductor Loss Ef- 15. R. Wu, J. Chen, and X. Fang, “A novel on-chip trans- fects”. IEEE TRANSACTIONS ON ELECTRON DEVIC- former with patterned ground shield for high ES, VOL. 53, NO. 3, MARCH 2006. common-mode transient immunity isolated sig- 28. Ahlberg, P. et al. Graphene as a Diffusion Barrier in nal transfer,” IEEE Electron Device Lett., vol. 39, no. Galinstain-Solid Metal Contacts. IEEE Trans. Elec- 11, pp. 1712–1715, Nov. 2018. tron Devices 1–5 (2014). https://doi.org/10.1109/LED.2018.2871049. 29. F. Khan, Y. Zhu, J. Lu, and J. Pal, “MEMS- based tun- 16. Y. Wang, Q. Zhang, L. Zhao, Y. Tang, A. Shkel, and E. able meander inductor,” Electron. Letters vol. 51, S. Kim, “Vibration energy harvester with low reso- no. 20, pp. 1582–1583, Oct. 2015. nant frequency based on flexible coil and liquid https://doi.org/10.1049/el.2015.2495 spring,” Appl. Phys. Lett., vol. 109, no. 20, 2016, Art. 30. A. Bhattacharya, D. Mandal, and T. K. Bhattacha- no. 203901. ryya, “A 1.3–2.4-GHz 3.1-mW VCO using electro- https://doi.org/10.1063/1.4967498. thermo-mechanically tunable self-assembled 17. J. Y. Park and J. U. Bu, “Packaging compatible mi- MEMS inductor on HR substrate,” IEEE Trans. Mi- cro transformers on a silicon substrate,” IEEE Trans. crow. Theory Techn., vol. 63, no. 2, pp. 459–469, Adv. Packag., vol. 26, no. 2, pp. 160– 164, May Feb. 2015 2003. https://doi.org/10.1109/TMTT.2014.2380357 https://doi.org/10.1109/tadvp.2003.817341. 31. A. Shirane, H. Ito, N. Ishihara, and K. Masu, “Pla- 18. Yen-Chung Chiang and al, “A Study on the Variable nar solenoid inductor in radio frequency micro- Inductor Design by Switching the Main Paths and electro-mechanical systems technology for vari- the Coupling Coils” Electronics 2021, 10, 1856. able inductor with wide tunable range and high https://doi.org/10.3390/electronics10151856 quality factor,” Jpn. J. Appl. Phys., vol. 51, no. 5S, p. 19. Y. Yokoyama, T. Fukushige, S. Hata, K. Masu, and A. 05EE02, May 2012. Shimokohbe,“On-chip variable inductor using mi- https://doi.org/10.1143/JJAP.50.05EE01 cro electromechanical systems technology,” Jpn. 32. Tengxing W; Wei J; Ralu D; Daniel R; Leonidas E. J. Appl. Phys., vol. 42, no. 4B, p. 2190, 2003. O; Yujia P;Guoan W “Novel electrically tunable mi- 20. S. Zhou, X.-Q. Sun, and W. N. Carr, “A micro variable crowave solenoid inductor and com- pact phase inductor chip using MEMS relays,” in Proc. Chicago shifter utilizing permaloy and PZT thin films,” IEEE Int. Conf. Solid State Sens. Actuators TRANSDUC- Trans. Microw. Theory Techn., vol. 65, no. 10, pp. ERS, vol. 2. Jun. 1997, pp. 1137–1140. 3569–3577, Oct. 2017. doi 21. H. Sugawara, Y. Yoshihara, H. Ito, K. Okada, and https://doi.org/10.1109/TMTT.2017.2731765 K. Masu, “Wide range RF variable inductor on Si 33. H.Chen,X. Wang, Y. Gao, X. Shi, Z. Wang, N. Sun, CMOS chip with MEMS actuator,” in Proc. 34th Eur. M. Zaeimbashi, X. Liang, Y. He, C. Dong,Y. Wei, J.G. Microw. Conf., Oct. 2004, pp. 701–704. Jones, M.E. McConney, M.R. Page, B.M. Howe, G.J. 22. H. Sugawara and al, “Variable RF inductor on Si Brown, N.-X. Sun “Integrated Tunable Magneto- CMOS chip,” Jpn. J. Appl. Phys., vol. 43, no. 4B, pp. electric RF Inductors”. IEEE Trans. Microw. Theory 2293–2296, 2004. Tech. 2020 36 I. El Gmati et al.; Informacije Midem, Vol. 53, No. 1(2023), 31 – 37 https://doi.org/10.1109/TMTT.2019.2957472 34. Yen-Chung Chiang, Juo-Chen Chen and Yu- Hsin Chang « A Study on the Variable Inductor Design by Switching the Main Paths and the Coupling Coils» Electronics 2021, 10, 1856. https://doi.org/10.3390/electron- ics10151856 35. Alexander M. Watson; Thomas F. Leary; Jonathan Itokazu; Aji G. Mattamana; Tony Quach; Aaron T. Ohta; Wayne A. Shiroma; Christopher E. Tabor “Tunable Microwave Inductor Using Liquid-Metal Microfluidics” 2021 IEEE Texas Symposium on Wireless and Microwave Circuits and Systems (WMCS), May 2021. 36. N. Lazarus and S. S. Bedair, “Bubble inductors: pneumatic tuning of a stretchable inductor,” AIP Advances, vol. 8, p. 056601, 2018. Copyright © 2023 by the Authors. This is an open access article dis- tributed under the Creative Com- mons Attribution (CC BY) License (https://creativecom- mons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Arrived: 21. 09. 2022 Accepted: 23. 02. 2023 37 38 Original scientific paper https://doi.org/10.33180/InfMIDEM2023.104 Journal of Microelectronics, Electronic Components and Materials Vol. 53, No. 1(2023), 39 – 48 Power and Area Efficient Sense Amplifier Based Flip Flop with Wide Voltage and Temperature Upholding for Portable IoT Applications Prashant Teotia, Owais Ahmad Shah Department of Electronics & Communication Engineering, Noida International University, Gautam Budh Nagar, Uttar Pradesh, India Abstract: A sense amplifier based flip-flop (SAFF) capable of operating unfailingly at wide voltage and temperature ranges is proposed in this work. The proposed flip-flop (FF) has a single ended latch design which results in a significant improvement in power and area requirements. The modified sense amplifier along with the single ended latch design enables robust and low power operation at all variations in the input data activity. The proposed SAFF is developed in 32nm CMOS technology, and a thorough and conclusive investigation with corner case simulation for wide process, voltage and temperature (PVT) variations is carried out in order to verify the design utilization. Comprehensive comparison and analysis with previously available state-of-the-art SAFFs validate that the proposed SAFF is functional at wide voltage ranges for temperature changes of 120 o to -40 o while upholding better and optimal power and power delay product (PDP) results. The proposed FF, because of its power efficiency, is best suited for portable Internet of Things (IoT) devices. Keywords: low power design; CMOS digital circuit; sense amplifier based flip-flop; single ended; IoT Energetsko in prostorsko učinkovit senzorski ojačevalnik na osnovi flip flopa s širokim razponom napetosti in temperature za prenosne aplikacije interneta stvari Izvleček: V članku je predlagan flip-flop, ki temelji na senzorskem ojačevalniku (SAFF) in lahko deluje brezhibno v širokem razponu napetosti in temperature. Predlagan flip-flop (FF) ima enonivojski zapah, kar znatno izboljša zahteve po moči in površini. Spremenjen ojačevalnik zaznavanja skupaj z enonivojskim zapahom omogoča zanesljivo delovanje z nizko porabo energije pri vseh spremembah vhodne podatkovne aktivnosti. Predlagani SAFF je razvit v 32 nm tehnologiji CMOS, za preverjanje uporabe zasnove pa je opravljena temeljita in prepričljiva raziskava s simulacijo robnih primerov za velike spremembe procesa, napetosti in temperature (PVT). Obsežna primerjava in analiza s predhodno razpoložljivimi najsodobnejšimi SAFF potrjujeta, da je predlagani SAFF funkcionalen v širokem razponu napetosti za temperaturne spremembe od 120 o do -40 o, hkrati pa zagotavlja boljše in optimalne rezultate glede moči in zakasnitve moči (PDP). Predlagani FF je zaradi svoje energetske učinkovitosti najprimernejši za prenosne naprave interneta stvari (IoT). Ključne besede: zasnova z nizko porabo energije; digitalno vezje CMOS; flip-flop, ki temelji na senzorskem ojačevalniku; enonivojsko vezje; internet stvari * Corresponding Author’s e-mail: mail_owais@yahoo.co.in How to cite: P. Teotia et al., “Power and Area Efficient Sense Amplifier Based Flip Flop with Wide Voltage and Temperature Upholding for Portable IoT Applications", Inf. Midem-J. Microelectron. Electron. Compon. Mater., Vol. 53, No. 1(2023), pp. 39–48 39 P. Teotia et al.; Informacije Midem, Vol. 53, No. 1(2023), 39 – 48 1 Introduction requirements because it has a shorter set-up time and greater power output than standard FF. But there are Smart and IoT based technologies have advanced two real-world issues with this FF architecture. First, swiftly, creating a wide range of prospects for tech- switching power reduction becomes less effective at nological advances in many different areas of life. The reduced or no data activity and power dissipation may main goals of IoT technologies are to improve quality be more at sense amplifier’s pre-charge processes. Sec- of life, ensure improved system (or process) efficiency, ond, a longer clock-to-output delay results from the and streamline processes across a range of industries. NAND latching stage receiving the data from the sense IoT technologies are rapidly evolving and providing a amplifier stage. Additionally, it has been shown that number of beneficial outcomes, but in order to prevent complementary FF outputs are not always required in negative environmental effects and ensure the sensi- applications. ble use of limited global resources, this rapid develop- ment must be closely monitored and analyzed from an A sense amplifier based flip flop suited for the low environmental point of view [1]. In the former sense, power domain is proposed in this study. The structure significant research is required to thoroughly examine of the paper is as follows: The prior-art SAFFs are ex- the benefits and drawbacks of IoT technologies. For IoT amined in Section 2, along with their advantages and applications like wearable technology and portable disadvantages. The proposed SAFF, which is intended medical equipment, on-chip security mechanisms are for reliable and low power operations, is presented in essential [2]. Memory offers a physically unclonable Section 3. The simulation findings and evaluations with function that can be employed in security implementa- existing SAFFs are presented in Section 4. Finally in sec- tion. Problems with resistive RAM memory applications tion 5 the proposed SAFF is experimented in an actual stem from its stochastic switching process and the re- digital circuit along with power measurements to verify sistance’s inherent unpredictability. For use in smart its worthiness. mote devices, biomedical implants and wireless sensor nodes, the present IoT era requires ultra-low power Sys- tem on Chip (SoC) designs and architectures [3]. 2 Related work In modern digital circuit design, flip-flop is the critical A basic sense amplifier based flip-flop consists of two part of most essential circuit components since they stages: a sensing stage and a latching stage. The sens- synchronize data flow and allow for local data storage ing stage uses a fast differential sense amplifier having [4]. A typical processor requires a lot of flip-flops, often two outputs (SB and RB), which is followed by a slave hundreds of thousands, because this synchronization latch. The principle of operation of this flip-flop may be must take place across the whole clock domain [5]. Due described as follows: to their high density, flip-flops consume a substantial The precharge state: When the clock is low, both the amount of both space and power in the overall circuit sensing stage’s outputs, SB and RB are high. The latch architecture [6]. In light of this, minimizing the power stage retains the prior value. dissipation of flip-flops has a major effect on system lev- The evaluation state: When clock is high, one of the el power efficacy, particularly for IoT applications with outputs of sensing stage is low. If the input is high, the constrained energy resources [7]. For IoT Integrated Cir- sensing stage’s SB output terminal will be low, and if cuits (IC) to maximize battery life, power consumption the input is low, the sensing stage’s RB output terminal must be kept to a minimum. However, a significant per- will be low. At this point, the state of SB and RB deter- centage of power, specifically the dynamic power dis- mines how the output of the latch stage is driven. On sipation, in a synchronous system is used by flip-flops, the rising edges of the clock, for instance, if SB is low, which can change its state at every clock pulse. Due to the output is driven to a high state logic, and if RB is this, numerous research projects have been carried out low, the output is driven to a low state logic. in an effort to create flip-flops that use less energy and are more efficient. Fig. 1 shows the prior-art SAFF, known as CBSAFF, pro- posed by [14]. The sensing stage in this FF uses a ca- The overall system architecture greatly benefits from pacitive boosting approach. The capacitive boosters the chip area and power consumption that FFs pro- which are actually MOS transistors that are being used vide. In order to satisfy varied application require- as capacitors amplify the data signal and provide a big ments, many FF designs have been presented [8]–[12]. enough voltage swing to trigger the latch stage from One of them is the sense amplifier-based FF architec- −β.VDD to β.2VDD. Here, the value of “β” (the boost- ture, which comprises of a latch and a dynamic logic ing efficiency) is <1 due to parasitic capacitances. The sense amplifier [13]. It is thought to be a superior de- buffer turns on the pre-charge transistor to charge the sign choice to meet both low power and high speed 40 P. Teotia et al.; Informacije Midem, Vol. 53, No. 1(2023), 39 – 48 boost capacitor to provide voltage when the clock in- Vdd put is ‘0’. This buffer is utilized to help boost the volt- Vdd age to twice the value when the clock input becomes ‘1’ and to stop the clock signal from being loaded with CLK CLK Vdd CLK Vdd capacitor. SB RB SB RB In Fig. 2, Strollo’s SAFF [15] is schematically depicted. The output stage can be thought of as a combination Q QB of the N-C MOS circuit and the typical NAND based SR CLK latch. The high-to-low output transition is accelerated using the additional transistors. In this design, remov- D DB DB ing the speed up network and condensing the sizes of few transistors significantly minimize power dissipa- RB SB tion. As a result, the capacitive load is reduced, which CLK permits a reduction in the size of the driving transis- tors at the sensing stage. In order to improve latency and power dissipation during the “1” to “0” transition a Figure 2: Architecture of Strollo’s FF [15] PMOS is added to the output stage, this further reduces the current (crow-bar). conflict between latching pMOS transistors and output or compliment output pull-down pathways. Vdd Vdd Vdd Vdd Vdd Vdd CLK CLK SB RB SB RB CLK SB RB Q QB Q QB CLK CLK TC TC DB D CLK CLK SB RB SB RB TC SB RB D DB D DB CLK CLK CLK Figure 3: Architecture of Jeong’s FF [16] Another example of SAFF is the Nikolic’s SAFF [17] de- picted in Fig. 4. The sense amplifier is the same as that Figure 1: Architecture of CBSAFF [14] in Fig. 2. Latching stage allows for small keeper transis- tors because only one transistor is active in each branch A SAFF based on a transition complete (TC) signal was when the state is changed. The slave latch has symmet- introduced by Jeong in [16] in order to address the is- ric true and complementary trees, which causes the sues with earlier SAFFs at low voltages. This SAFF-struc- delays at both outputs to be the same. The small size ture is depicted in Fig. 3. The sensing stage’s outputs of the keeper transistors causes them to quickly turn are the two inputs of a NAND gate, which produces off during the transition. This enables the load to be the TC signal. Only after SET output of sensing stage is driven and the latch’s status to be changed by exterior charged up, the signal TC discharges. As a result, at the driver transistors. This characteristic makes the proce- clocks falling edge, both TC and SET signal are briefly dure of output transistor size optimization simple. The high, which could result in an output glitch. However, resistance of the output stage to crosstalk when there compared to the SAFF-related issue, this hiccup is mi- is a low clock pulse is the only restriction on minimiz- nor and far smaller. This kind of FF prevents the current ing the keeper transistors. It enables reduced clock- swing operation as well as logic integration inside the 41 P. Teotia et al.; Informacije Midem, Vol. 53, No. 1(2023), 39 – 48 flip-flop. Additionally, just one transistor being active 3 Proposed SAFF during the changeover improves the output stage’s ca- pacity to drive other transistors and prevents crowbar Fig. 6 shows the architecture of proposed SAFF. An in- current, which lowers power consumption. put/output structure with a novel sense amplifier de- sign and a single-ended pass transistor logic-based Vdd Vdd latch is used. This design optimization technique helps the leading sense-amplifier stage’s transistor size re- CLK CLK quirement be reduced. The resulting design is more SB RB efficient than traditional designs in terms of layout SB RB and performance. A real single-ended latch is used to R overcome the shortcoming of the dynamic latch. This S Q improves the speed and power performances by sig- QB nificantly reducing the loading effect on the sensing SB amplifier. To create signal S from the output signal SB RB D DB R S of the sense amplifier, an additional inverter INV1 is introduced. Signal RB is in the interim removed from the latch design’s list of inputs. This results in a genuine CLK single-ended latch circuit as depicted in Fig. 6. The operation of the FF can be realized as when clock signal is low, both the outputs of sensing stage RB and SB are precharged to logic ‘1’, transistors N1 and N2 are Figure 4: Architecture of Nikolic’s FF [17] turned ON, X stays low during the ‘0-1’ transition of clock. During  this, the latching  stage maintains the Kim proposed a SAFF in [18] with updated N-C2MOS state of the flip-flop. At the time when clock changes latches shown in Fig. 5. The sense amplifier is the same to ‘1’, the sense amplifier stage starts the transition. as it was in the conventional SAFF. Since the coupled One of RB or SB will drop if the shift is successful, rais- inverters are constructed using transistors of the small- ing X. Because of the connection between X and the est possible size, the additional capacitive loadings transistor associated with it remains OFF during the they cause at the output nodes are insignificant (less sensing stage transition and is only turned ON when than 10% of the total capacitance value). The differen- the sensing stage transition is over (i.e one of the pre- tial output nodes are completely separated, therefore charged node is either brought down to low logic via the load capacitance at the other output node does not by transistor  N1  or N2  whereas  the other precharged affect the transition speed of one output node. node continues to be at logic high). Therefore both the operational yield and speed loss of this transistor is not Vdd Vdd Vdd Vdd CLK CLK CLK P1 P2 P3 P4 CLK Vdd SB RB SB RB INV3 Vdd Q SB P5 Q QB N7 QB C CLK CLK N1 N2 X INV2 N8 D DB SB X S SB RB INV1 A N5 B Latching Stage CLK D N3 N4 DB CLK N6 Figure 5: Architecture of Kim’s FF [18] Sensing Stage Figure 6: Proposed SAFF 42 P. Teotia et al.; Informacije Midem, Vol. 53, No. 1(2023), 39 – 48 present as compared in the sense amplifiers discussed in previous sections. Remember that as soon as the sens- ing stage swings, either N1 or N2 will switch off, and as a result, any further change in input will not be able to affect the state of the reset and set terminals. The transis- tor N7 adds better power efficiency to the FF at the cost of small delay. Nevertheless this small delay is compen- sated by the latching stage’s single-ended structure. When the clock is low (precharge state), the fact that two nodes A and B are not equalized is another no- table aspect of this design. Also at the rising edge of clock, the voltages at the nodes A and B are the same which lessen the effect of mismatch. Conversely, the proposed design has the potential for nodes A and B to diverge at the rising edge of the clock, in turn may reduce the stability of the sensing stage. Neverthe- Figure 7: Transient waveform of proposed FF less, the beneficial impact obtained by turning OFF the stabilize the pull down operations on SB. This leads in a transistor while high clock outweighs the unfavorable shorter hold and setup time for the proposed SAFF. The impact of node “A” equal to “B”. standard sizes of the transistors are maintained at the latching stage for P5 and N8.  Since the inverter’s driv- For the latch of the proposed design the signal X pro- ing capabilities are unnecessary, the size of the invertor duced in the sensing stage is applied in place of the transistors can  be minimized, resulting in significant global clock. The signal RB is not present in the latch- power savings. ing stage since it has a single-ended form. If the input signal data D is low at the rising edges of the clock, SB Table 1: Transistor size of proposed SAFF will remain at operating voltage, and node C will be discharged to low through transistor N8. The transistor P5 will switch ON if the input data D is high, in a simi- Component W (nm) Component W (nm) lar manner to how node SB will discharge to low if D is P1-P5 320 N6 800 high. Now, transistors P5 and N8 together are pulling N1, N2, N4, N5, N7 & N8 160 P (Inv1-Inv3) 320 Node C to operating voltage. The outputs QB and Q are driven by two inverters that are coupled at node C. The N3 320 N (Inv1-Inv3) 160 operational waveforms in Fig. 7 corroborate and make clear this entire process. 4 Simulation results & discussions The transistor size of the proposed SAFF is listed in Ta- ble 1. The sizes of N3 and N6 are selected to be larger The proposed architecture is tested with available de- because it controls the performance of the FF since signs to show its competitiveness and performance. it directly influences the pull down speed of SB. The The simulations are carried out at 32 nm CMOS tech- single-ended structure of the SAFF latch stage allows nology node in SPICE using PTM models [19]. The nom- transistors N2, N4, and N7 to be of minimal size. In ad- inal working parameters are 25 ℃ temperature, 200 dition, since the load on RB is “0” than that on SB, reduc- MHz frequency and 0.9 V operating voltage. A 16-bit ing the widths of the transistors N2, N4, and N7 may input data is used with data activity of 50%. Table 2: Performance comparison of flip-flops. Flip Flop Layout area No. of Transitors PDP @0.7 V PDP @0.8 V PDP @0.9 V C-Q delay @0.9 (um2) (uW) (uW) (uW) V (ps) Kim’s 0.352 26 62.971 66.685 72.239 30.065 Nikolic’s 0.262 28 34.061 26.155 31.431 26.408 Strollo’s 0.279 23 26.673 24.898 24.693 16.948 Jeong’s 0.247 26 33.916 27.874 17.84 13.424 CBSAFF 0.316 38 159.698 70.594 127.457 56.22 Proposed 0.209 25 23.325 21.134 18.666 19.266 43 P. Teotia et al.; Informacije Midem, Vol. 53, No. 1(2023), 39 – 48 The simulation results for power consumption at sup- ply voltage variation between 0.5 V to 1.1 V and at vari- ations in temperature from 0 ℃ to 100 ℃ are shown in Fig. 8. This figure makes it very clear that, for all the variations considered in this paper, the proposed de- sign consumes the least amount of power at various voltage levels and temperature changes. (a) (b) Figure 8: Average power at variations in (a) Voltage (b) Temperature (a) (b) Figure 10: Corner case analysis Table 2 is the detailed comparison of proposed sense amplifier based flip flop with existing FFs. It includes the layout area, the clock to Q delay and the power de- lay product at variations in supply voltage. It was ob- served that at nominal operating conditions, Jeong’s FF was the fastest followed by Strollo’s FF. The proposed flip flop was third fastest but nearly comparable to its Figure 9: Area requirements (a) Sum of width (b) Num- rivals. When power and delay both were taken into ac- ber of transistors count, the proposed flip flop showed better results. The PDP at three different voltage variations of 0.7 V, 0.8 V 44 P. Teotia et al.; Informacije Midem, Vol. 53, No. 1(2023), 39 – 48 (a) and 0.9 V showcases the worthiness of proposed de- sign for portable power efficient devices. Although in terms of number of transistors the pro- posed design has second least count but the overall layout area of the proposed design is smaller compared to any other design. This difference and area advantage of proposed design can be seen in Fig. 9. In any VLSI design, it is important to know the effect (b) of process, voltage and temperature variations at ex- treme corners. This test was conducted with corner cas- es of FS 0.9 V /25 °C, SF 0.9 V /25 °C, TT 0.9 V /25 °C, SS 0.8 V /125 °C and FF 1.1 V /-40 °C. The power results of proposed design at all extreme corners were far better than its counterparts. These corner results are shown in Fig. 10. Fig. 11 is the power results at variations in data activ- ity. Wide variation in input data from 0% to 100% was taken into account. For 0% both all input high and all Figure 11: Average power at variation in data activity input low were obtained. This test was conducted at (a) at 100 MHz (b) at 200 MHz. two different frequencies i.e. at 100 MHz and 200 MHz. Figure 12: Results of aggressive temperature and voltage scaling (Green: Pass-case; Red: Fail-case) 45 P. Teotia et al.; Informacije Midem, Vol. 53, No. 1(2023), 39 – 48 Results in Fig. 11(a) and 11(b) showcases the proposed 5 Experimental verification design’s superiority over other designs at all trails. In order to illustrate and validate the correct logical Simulations so far have shown that the proposed de- operation and power performance of the proposed de- sign is resilient, however, additional testing is conduct- sign, the proposed SAFF is tested for use in complex ed against changes in voltages for the near-threshold digital circuitries. The flip-flop under test (FUT) was to sub-threshold range along with the effects of a wide evaluated as a four-bit shift register and a three-bit range of temperatures. To test this, Monte Carlo simu- counter in simulation settings under the nominal con- lations are conducted with aggressive voltage scaling ditions of 25 ℃ temperature, 0.9 V supply voltage, and across the range of 0.3 V to 0.5 V with 0.02 V increments 200 MHz frequency. Figure 13 are the transient wave- (20 mV) and over the range of temperature with 10 °C forms of these tests thereby demonstrating the correct steps from -40 °C to 120 °C. The results of the tests per- circuit functionality of the proposed design. formed on all of the flip-flops are shown in Figure 12. A ‘pass’ case test is represented by a green box, while a The test was also conducted on existing FFs to deter- ‘fail’ case test is represented by a red box. Kim’s FF was mine and compare the power performance of all the the only one to pass all of the tests, followed by CBSAFF designs. The average power was determined at volt- with 2 fail instances and Strollo’s FF with 6 fail cases. ages of 0.8 V, 0.9 V and 1 V and at variations in tempera- Notably, neither Kim’s FF nor the CBSAFF performed ture of -40 ℃, 25 ℃ and 125 ℃. Figure 14 and 15 shows well throughout the analysis. In addition, the proposed the results obtained from these tests, the proposed SAFF has also restricted functioning when subjected to FUT demonstrated better power performance at all sub-threshold voltage levels and temperature ranges conducted tests and therefore is worthy of considera- and is therefore recommended for use at voltage levels tion in designs requiring low power operations. of more than 0.34 V. (a) (a) (b) (b) Figure 14: Power results of shift register at variation in (a) voltage (b) temperature 6 Conclusion A modified sensing stage is used to create a sense am- plifier based flip flop. The major goal was to reduce Figure 13: Transient waveform of proposed FUT as (a) power consumption so that the proposed flip flop will 4-bit shift register (b) 3-bit counter be a viable option for digital circuits and IoT applica- tions. After comprehensive simulations, the proposed design excelled practically all other designs in metrics 46 P. Teotia et al.; Informacije Midem, Vol. 53, No. 1(2023), 39 – 48 (a) 4. O. A. Shah, I. Ahmed Khan, G. Nijhawan, and I. Garg, “Low Transistor Count Storage Elements and their Performance Comparison,” in 2018 In- ternational Conference on Advances in Comput- ing, Communication Control and Networking (ICACCCN), Greater Noida (UP), India, Oct. 2018, pp. 801–805. https://doi.org/10.1109/ICACCCN.2018.8748364. 5. H. You, J. Yuan, W. Tang, Z. Yu, and S. Qiao, “A Low- Power High-Speed Sense-Amplifier-Based Flip- (b) Flop in 55 nm MTCMOS,” Electronics, vol. 9, no. 5, p. 802, May 2020, https://doi.org/10.3390/electronics9050802. 6. I. A. Khan, O. A. Shah, and M. T. Beg, “Analysis of different techniques for low power Single Edge Triggered Flip Flops,” in 2011 World Congress on Information and Communication Technologies, Mumbai, India, Dec. 2011, pp. 1363–1367. https://doi.org/10.1109/WICT.2011.6141447. 7. Y. Lee, G. Shin, and Y. Lee, “A Fully Static True-Single- Figure 15: Power results of counter at variation in (a) Phase-Clocked Dual-Edge-Triggered Flip-Flop for voltage (b) temperature Near-Threshold Voltage Operation in IoT Applica- tions,” IEEE Access, vol. 8, pp. 40232–40245, 2020, like as average power, PDP, data activity, and in espe- https://doi.org/10.1109/ACCESS.2020.2976773. cially layout area. The proposed design also upheld the 8. D. Pan, C. Ma, L. Cheng, and H. Min, “A Highly Ef- voltage variation of 0.5 V to 1.1 V and corner case tem- ficient Conditional Feedthrough Pulsed Flip-Flop perature variation of -40 ℃ to 120 ℃. Further advan- for High-Speed Applications,” IEEE Trans. VLSI tage of the novel design includes the power efficiency Syst., vol. 28, no. 1, pp. 243–251, Jan. 2020, at input data activity variation. In addition, the pro- https://doi.org/10.1109/TVLSI.2019.2934899. posed FF is recommended for designs with minimum 9. A. Karimi, A. Rezai, and M. M. Hajhashemkhani, “A voltage of 0.34 V. novel design for ultra-low power pulse-triggered D-Flip-Flop with optimized leakage power,” Inte- gration, vol. 60, pp. 160–166, Jan. 2018, https://doi.org/10.1016/j.vlsi.2017.09.002. 7 Conflict of interest 10. O. A. Shah, G. Nijhawan, and I. A. Khan, “Low pow- er area efficient self-gated flip flop: Design, im- The authors declare no conflict of interest. plementation and analysis in emerging devices,” Engineering and Applied Science Research, vol. 49, no. 6, pp. 744-752, 2022, 8 References https://doi.org/10.14456/EASR.2022.72. 11. J. Lin, Y. Hwang, C. Wong, and M. Sheu, “Single‐ 1. M. A. S. Bhuiyan, “CMOS series-shunt single-pole ended structure sense‐amplifier‐based flip‐flop double-throw transmit/receive switch and low for low‐power systems,” Electron. lett., vol. 51, no. noise amplifier design for internet of things based 1, pp. 20–21, Jan. 2015, radio frequency identification devices,” Informaci- https://doi.org/10.1049/el.2014.3922. je MIDEM, vol. 50, no. 2, pp. 105–114, Sep. 2020, 12. S. Lapshev and S. M. R. Hasan, “New Low Glitch https://doi.org/10.33180/InfMIDEM2020.203. and Low Power DET Flip-Flops Using Multiple C- 2. M. Arunraja, V. Malathi, and E. Sakthivel, “Distrib- Elements,” IEEE Trans. Circuits Syst. I, vol. 63, no. uted Energy Efficient Clustering Algorithm for 10, pp. 1673–1681, Oct. 2016, Wireless Sensor Networks,” Informacije MIDEM, https://doi.org/10.1109/TCSI.2016.2587282. vol. 45, no. 3, pp. 180–187, 2015.. 13. O. A. Shah, G. Nijhawan, and I. A. Khan, “Improved 3. S. Raj and S. S. Ramapackiam, “Energy Efficient Sense Amplifier Based Flip Flop Design For Low and Low dynamic power Consumption TCAM Power And High Data Activity Circuits,” Journal of on FPGA,” Informacije MIDEM, vol. 51, no. 3, Nov. Applied Science and Engineering, vol. 26, no. 7, 2022, pp. 1047–1053, Oct. 2022, https://doi.org/10.33180/InfMIDEM2022.304. https://doi.org/10.6180/jase.202307_26(7).0015. 47 P. Teotia et al.; Informacije Midem, Vol. 53, No. 1(2023), 39 – 48 14. A. Ramaswami Palaniappan and L. Siek, “Wide‐in- put dynamic range 1 MHz clock ultra‐low supply flip‐flop,” Electron. lett., vol. 54, no. 15, pp. 938– 939, Jul. 2018, https://doi.org/10.1049/el.2018.1134. 15. A. G. M. Strollo, D. De Caro, E. Napoli, and N. Pe- tra, “A novel high-speed sense-amplifier-based flip-flop,” IEEE Trans. VLSI Syst., vol. 13, no. 11, pp. 1266–1274, Nov. 2005, https://doi.org/10.1109/TVLSI.2005.859586. 16. H. Jeong, T. W. Oh, S. C. Song, and S.-O. Jung, “Sense-Amplifier-Based Flip-Flop With Transition Completion Detection for Low-Voltage Opera- tion,” IEEE Trans. VLSI Syst., vol. 26, no. 4, pp. 609– 620, Apr. 2018, https://doi.org/10.1109/TVLSI.2017.2777788. 17. B. Nikolic, V. G. Oklobdzija, V. Stojanovic, Wenyan Jia, James Kar-Shing Chiu, and M. Ming-Tak Le- ung, “Improved sense-amplifier-based flip-flop: design and measurements,” IEEE J. Solid-State Cir- cuits, vol. 35, no. 6, pp. 876–884, Jun. 2000, https://doi.org/10.1109/4.845191. 18. J.-C. Kim, Y.-C. Jang, and H.-J. Park, “CMOS sense amplifier-based flip-flop with two N-C2MOS out- put latches,” Electron. Lett., vol. 36, no. 6, p. 498, 2000, https://doi.org/10.1049/el:20000409. 19. “Predictive Technology Model,” PTM. https://ptm. asu.edu (accessed Oct. 10, 2022). Copyright © 2023 by the Authors. This is an open access article dis- tributed under the Creative Com- mons Attribution (CC BY) License (https://creativecom- mons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Arrived: 29. 01. 2023 Accepted: 20. 04. 2023 48 Original scientific paper https://doi.org/10.33180/InfMIDEM2023.105 Journal of Microelectronics, Electronic Components and Materials Vol. 53, No. 1(2023), 49 – 53 The Design of Frequency-tunable Mechanical Tuning Coupler Based on Coupled Line Structure Jiayi Wang1,2, Yuepeng Yan1,3 1Institute of Microelectronics of the Chinese Academy of Sciences, Beijing, China 2University of Chinese Academy of Sciences, Beijing, Beijing, China 3Beijing Key Lab of New Generation Communication RF Chip Technology, Beijing, China Abstract: In the current study, a frequency-tunable mechanical tuning coupler based on coupled line structure is proposed, which shall have high stability and high flexibility of tuning. This mechanical tuning coupler is based on coupled line structure of microstrip technology. This coupler is composed by baseplate for signal transmission and mobile plate for mechanical tuning. By mechanically tuning the rotation angle, the coupling length is changed and tunability is achieved. The operating frequency can be tuned from 1.02 GHz to 1.27 GHz while the rotation angle is tuned from 180° to 135° approximately. The design principle is based on quarter-wave resonance phenomenon of coupled line. Keywords: frequency-tunable, coupled line, mechanical tuning, microstrip Načrtovanje mehanskega spojnika z možnostjo nastavitve frekvence na podlagi strukture sklopljene linije Izvleček: V pričujoči študiji je predlagan frekvenčno nastavljiv mehanski sklopnik, ki temelji na strukturi sklopljene linije in ima visoko stabilnost in prilagodljivost uglaševanja. Ta mehanski nastavljiv sklopnik temelji na strukturi sklopljene linije mikropasovne tehnologije. Sestavljata ga osnovna plošča za prenos signala in premična plošča za mehansko nastavljanje. Z mehanskim nastavljanjem kota vrtenja se spreminja dolžina sklopke in doseže nastavljivost. Delovno frekvenco je mogoče nastaviti od 1,02 GHz do 1,27 GHz, kot vrtenja pa približno od 180° do 135°. Načelo zasnove temelji na resonančnem pojavu četrtvalovne resonance sklopljene linije. Ključne besede: frekvenčno uglaševanje, sklopljena linija, mehansko uglaševanje, mikropasec * Corresponding Author’s e-mail: safe_iluosi@163.com 1 Introduction 10]. Besides, some couplers also use magnetic material Coupler, known as an important communication elec- to realize tunablility [11]. tronics, plays an essential role in optical and microwave system [1-4]. Couplers are usually used for power dis- Coupled line is a popular structure in microstrip indus- tribution, phase shift and so on [5]. Among them, tun- try. When two waveguides get close to each other, the able couplers can realize further functionality for its function of power redistribution can be realized. This tunability of coupling coefficients or frequency. Most structure is popular in the application of couplers [12- tunable couplers realize tunablility by using recon- 13]. The principle of coupled line coupler is based on figurable components, such as tunable capacitors or quarter-wave resonance phenomenon. Their operating Micro Electro Mechanical Systems (MEMS) switches [6- frequencies depend on coupling length mostly. How to cite: J. Wang et al., “The Design of Frequency-tunable Mechanical Tuning Coupler Based on Coupled Line Structure", Inf. Midem-J. Microelec- tron. Electron. Compon. Mater., Vol. 53, No. 1(2023), pp. 49–53 49 J. Wang et al.; Informacije Midem, Vol. 53, No. 1(2023), 49 – 53 In this study, a frequency-tunable mechanical tuning Fig. 2 shows the perspective model of frequency-tun- coupler based on coupled line structure is proposed. able mechanical tuning coupler with a rotation angle This tunable coupler uses two-layer structure. Using at 135°. two-layer structure can realize much functions like re- fining smoothness and reducing size [14]. In current study, the tunability of component is achieved by using this two-layer structure. Mechanical tuning is a popular tuning way in the de- sign of tuning resonators and filters [15-17]. Tuning screw structure is a popular structure in mechanical tuning. This coupler is made by printed circuit board (PCB) technology, which should have high stability and low manufacturing cost. This mechanical tuning cou- pler can be driven by a stepper micromotor [18]. By changing rotation angle of mobile plate, the coupling length of coupler can be changed leading to the shift- ing of operating frequency. The coupling lines is de- formed into semicircle-ring type. It still obeys the law Figure 2: perspective model of frequency-tunable me- of quarter-wave resonance phenomenon, but some chanical tuning coupler differences exist. The coupler includes baseplate for signal transmis- sion and mobile plate for mechanical tuning. The inner 2 The simulation of miniature tunable metal ring is printed on mobile plate. By rotating the mobile plate, the coupling length C2 is changed lead- inductor ing to the tunability of frequency. The coupling signal is conducted to Port 3 by the coupled lines. The end of The principle diagram of this frequency-tunable me- inner metal ring is open circuit. C1 will lead to the de- chanical tuning coupler can be given by Fig. 1. terioration of reflection coefficient S11, so the ratio of C2/C1 shall be controlled. The mathematic relation of operating frequency can be written as Eq. 1 [19] f c  0 (1) L 0.5 4 r where f0 is operating frequency, c is light speed, α is shape correction factor, L is arc length of coupled lines, εr is relative permittivity of dielectric material. The structure of frequency-tunable mechanical tun- ing coupler is shown as Fig. 1. The substrate material is FR4 with a relative permittivity εr about 4.2-4.7 with a Figure 1: principle diagram of frequency-tunable me- height h of 1 mm. The metal line is made by gold with chanical tuning coupler a height of 35 μm. The total size of the coupler is 40 mm×30 mm. Fig. 3 is the structure of frequency-tuna- C1 means the length of microstrip which does not ble mechanical tuning coupler. participate in coupling in signal transmission end. C2 means the length of coupling part. C3 means the The parameter r1, r2, s and w in Fig. 3 mean external length of microstrip which does not participate in cou- radium of inner ring, external radium of outer ring, in- pling in coupling end. The coupling length C2 decides terval between coupled lines and width of metal line the operating frequency. By mechanically tuning the which equal to 11.45 mm, 12.5 mm, 0.05 mm, 1 mm length of C2, tunability of frequency can be realized. respectively. The mobile plate is connected to a step- According to the simulation of Advance Design Sys- per micromotor, which drive the mobile plate rotating tem (ADS), the ratio of C2/C1 and C2/C3 is better to be from the bottom. The micromotor has a small volumn larger than 300%, or the reflection coefficients S11 and of Φ3.4 mm ×10.75 mm, which can ensure the com- transmission coefficient S21 will deteriorate gradually. pactness of this component. Besides, stepper motor 50 J. Wang et al.; Informacije Midem, Vol. 53, No. 1(2023), 49 – 53 The 1 dB-passband of Fig. 4(e) is from 0.97 GHz to 1.69 GHz with a coupling coefficient S31 around -8.66 dB at a center frequency of 1.27 GHz. While the 1 dB-pass- band of Fig. 4(f ) is from 0.71 GHz to 1.58 GHz with a coupling coefficient S31 around -8.75 dB at a center fre- quency of 1.02 GHz. The transmission coefficient S21 of 135° deteriorates a bit than that of 180° in passband, but it is still in reasonable range. Figure 3: (a) top view of frequency-tunable mechanical 3 Result and discussion tuning coupler (b) bottom view of frequency-tunable mechanical tuning coupler (c) top view of baseplate (d) This coupler is based on quarter-wave resonance phe- bottom view of baseplate (e) top view of mobile plate nomenon, so the tuning range appears inapparently (f ) bottom view of mobile plate for the limited change of coupling length in lower fre- quency. In theory, the tuning range is 24.5% with the means a motor driven by electric pulse signal. Each rotation angle tuned from 180° to 135°. So the tun- electric pulse signal will make the motor rotate for a ing range can be more apparent in higher operating certain degree, ensuring the controllability and stabil- frequency. Table 1 shows the comparison with other ity of tuning. The operating frequency is determined works. by its rotation angle. The simulation and measurement result is shown in Fig. 4. Table 1: Comparison of measurement result. Tuning range (GHz) Coupling (dB) NNC This work 1.02-1.27 (24.5%) -8.66 1 [5] 1.6-2.3 (36%) -3~-4 4 [6] 6.7-7.1 (5.7%) 3~-6 4 [7] 1.3-1.9 (46%) -3~-10 4 NNC means number of necessary tuning components. As a passive component, the properties of current coupler is based on the structure of coupled line it- self. Compared with active tunable couplers, current couplers cannot realize the tunability of S31 or signal amplification. However, the tuning of former works need to adjust their four tuning components (variable capacitors or RF switches) into respective certain value. While the tuning of this work relies on one micromotor, which has more flexibility of tuning. The coupled lines of this coupler is semicircle-ring type, whose properties are similar to common linear coupled line coupler. So when this coupler is designed, classical Figure 4: (a) (b) tunable coupler with a rotation angle equations of common linear coupled line coupler can about 135° and 180° (c) (d) simulation and (e) (f ) meas- be used as reference to choose parameters as shown in urement result of tunable coupler with a rotation angle Eq. 2 and Eq. 3 [19] about 135° and 180° Z 2  ) eZ ( e oZ 2 S31 in measurement is weaker than that in simulation for 0 0 0o a distance. The metal line is set as perfect conductor in K eZ  Z simulation, while real gold line has a weaker conductiv-  0e o 0o (3) ity than perfect conductor. It causes some difference be- eZ e 0 oZ0o tween simulation and measurement. The height of metal lines can be improved to reduce this difference. The tight- where Z0 is characteristic impedance of microstrip, αe ness between two plates is another important factor. is shape correction factor of even mode, αo is shape 51 J. Wang et al.; Informacije Midem, Vol. 53, No. 1(2023), 49 – 53 correction factor of odd mode, Z0e is even mode char- The funding sponsor of this work is Institute of Micro- acteristic impedance and Z0o is odd mode character- electronics of the Chinese Academy of Sciences. istic impedance. The width of microstrip w determines Z0 mostly, the relation is shown as Eq. 4 approximately. w shall be calculated to make impedance matched [19]. 6 References Z 60  8h w    log 0    (4) 1. A.M. Abbosh. A Compact UWB Three-Way Power Divider[J]. IEEE Microwave and Wireless Compo- r  w 4h  nents Letters, 2007, 17 (8): pp.598-600. Besides, the interval between coupled lines s deter- https://doi.org/10.1109/LMWC.2007.901777) mine coupling degree K and S31. In general, a larger 2. H. L. Ting, S. K. Hsu, T. L. Wu et al. A Novel and Com- interval between coupled lines s gets a smaller S31. S31 pact Eight-Port Forward-Wave Directional Coupler goes from -8.6 dB to -12.4 dB when s goes from 0.05mm With Arbitrary Coupling Level Design Using Four- to 0.5mm in measurement of this study. Therefore, s Mode Control Technology[J]. IEEE Transactions on shall be chosen according to desired S31. The height of Microwave Theory & Techniques, 2016, 1 (1): pp.1-9. metal lines can be improved to intensity the coupling https://doi.org/10.1109/TMTT.2016. 2623709) degree K in some degree as well. The tunability of S31 3. C. Caloz, A. Sanada, T. Itoh et al. A Novel Compos- can be realized if the tuning of s is achieved. ite Right-/Left-Handed Coupled-Line Directional Coupler with Arbitrary Coupling Level and Broad Some differences caused by shape difference can be Bandwidth[J]. IEEE Transactions on Microwave adjusted by simulation software subsequently. Theory & Techniques, 2004, 52 (3): pp.980-992. https://doi.org/10.1109/TMTT.2004.823579) Port 1 and Port 2 are connected by the mobile plate. 4. M. Kaneta, K. Yashiro, S. Ohkawa et al. A Magneto- If the metal lines do not contact tightly, the signal will static Forward Volume Wave Directional Coupler not transmit normally. The height of metal line can be with a Guiding Slot Structure[C]. International Mi- thicked for strengthening tightness. Besides, the equiv- crowave Symposium Digest IEEE. 1988. alent impedance of interlamellar connection ring shall https://doi.org/10.1109/MWSYM.1988.22174) be well designed and impedance matched in case of re- 5. A. Babnik, A. Suhadolnik, J. Mozina, “Single-Mode turn loss. The interlamellar connection ring is better to Optical Fiber Directional Coupler,” Informacije be a small circle whose diameter is similar to the width MIDEM, vol. 21 (1991), no. 3, pp. 145-150, http:// of metal line. Besides, the ratio of C2/C1 and C2/C3 will www.midem-drustvo.si/Journal%20papers/ influence the reflection coefficient S11, especially C2/ MIDEM_21(1991)3p145.pdf C1. Namely, the C2 cannot be close to C1. If this ratio is 6. O. D. Gurbuz, G. M. Rebeiz. A 1.6–2.3-GHz RF too small, reflection coefficient S11 will improve largely MEMS Reconfigurable Quadrature Coupler and and deteriorates transmission coefficient S21. So the Its Application to a 360 Reflective-Type Phase rotation angle shall be controlled in reasonable range. Shifter[J]. IEEE Transactions on Microwave Theory & Techniques, 2015, 63 (2): pp.414-421. https://doi.org/10.1109/TMTT.2014.2379258) 7. B. Hur, W. R. Eisenstadt. Tunable Broadband MMIC Ac- 4 Conclusions tive Directional Coupler[J]. IEEE Transactions on Micro- wave Theory & Techniques, 2013, 61 (1): pp.168-176. In this study, a frequency-tunable mechanical tuning https://doi.org/10.1109/TMTT.2012.2228218) coupler is proposed. By changing rotation angle of 8. Y. F. Pan, W. S. Chan, S. Zheng et al. A Highly Re- mobile plate, the coupling length of coupled line is configurable Coupler with Tunable Frequency, changed leading to tunability of operating frequency. Phase Difference and Coupling Coefficient Based The operating frequency of this coupler can be tuned on Circular Patch[C]. 2020 IEEE Asia-Pacific Micro- from 1.27 GHz to 1.02 GHz while the rotation angle is wave Conference. 2020: pp.1033-1035. tuned from 180° to 135°, the tuning range is 24.5%. https://doi.org/10.1109/APMC47863.2020.9331573) 9. S. M. Wang, C. Y. Chang, J. S. Lin et al. A Software Configurable Coupler with Programmable Cou- 5 Conflict of Interest pling Coefficient[C]. 2007 IEEE/MTT-S Internation- al Microwave Symposium. 2007: pp.185-188. The authors declare that they have no known competing https://doi.org/10.1109/MWSYM.2007.380321) financial interests or personal relationships that could have 10. S. Toyoda. Variable Coupling Directional Couplers Us- appeared to influence the work reported in this paper. ing Varactor Diodes[C]. 1982 IEEE MTT-S International 52 J. Wang et al.; Informacije Midem, Vol. 53, No. 1(2023), 49 – 53 Microwave Symposium Digest. 1982: pp.419-421. https://doi.org/10.1109/MWSYM.1982.1130742) 11. M. Kaneta, K. Yashiro, S. Ohkawa et al. A Magneto- static Forward Volume Wave Directional Coupler with a Guiding Slot Structure[C]. International Mi- crowave Symposium Digest IEEE. 1988. https://doi.org/10.1109/MWSYM.1988.22174) 12. X. Li, Z. H. Shao, M. K. Shen et al. High Selectivity Tunable Filtering Power Divider Based on Liquid Crystal Technology for Microwave Applications[J]. Journal of Electromagnetic Waves and Applica- tions, 2016, 1 (1): pp.1-10. https://doi.org/10.1080/09205071.2016.1155500) 13. L. Guo, H. Zhu, A. Abbosh et al. Phase Reconfig- urable Microwave Power Divider[J]. IEEE Transac- tions on Circuits and Systems II: Express Briefs, 2018, 1 (1): pp.1-5. https://doi.org/10.1109/TCSII.2018.2829200) 14. J.Y. Wang, Y. P. Yan et al. Coupled line coupler with smooth passband based on two-layer structure [J]. IEICE Electronics Express, 2022, 19 (17): pp.1-4. https://doi.org/10.1587/elex.19.20220292) 15. F. Mira, J. Mateu, C. Collado et al. Mechani- cal Tuning of Substrate Integrated Waveguide Resonators[J]. IEEE Microwave and Wireless Com- ponents Letters, 2012, 22 (9): pp.447-449. https://doi.org/10.1109/LMWC.2012.2208735) 16. F. Mira, J. Mateu, C. Collado et al. Mechanical Tun- ing of Substrate Integrated Waveguide Filters[J]. IEEE Transactions on Microwave Theory and Tech- niques, 2015, 63 (12): pp.3939-3946. https://doi.org/10.1109/TMTT.2015.2490144) 17. S. Kurudere, V. B. Ertürk. Novel Microstrip Fed Me- chanically Tunable Combline Cavity Filter[J]. IEEE Microwave and Wireless Components Letters, 2013, 23 (11): pp.578-580. https://doi.org/10.1109/LMWC.2013.2281432) 18. Zhaowei Machinery & Electronics Co., Ltd, 3.4mm stepper motor. Available from: https://www.zw- gearbox.com/stepper-motor/3-4mm-stepper- motor 19. Microstrip circuit compilation Group of Tsinghua University. Microstrip Circuit[M]. Tsinghua Univer- sity, 2017. Copyright © 2023 by the Authors. This is an open access article dis- tributed under the Creative Com- mons Attribution (CC BY) License (https://creativecom- mons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Arrived: 6.2.2023 Accepted: 22.6.2023 53 54 Boards of MIDEM Society | Organi društva MIDEM MIDEM Executive Board | Izvršilni odbor MIDEM President of the MIDEM Society | Predsednik društva MIDEM Prof. Dr. Barbara Malič, Jožef Stefan Institute, Ljubljana, Slovenia Vice-presidents | Podpredsednika Prof. Dr. Janez Krč, UL, Faculty of Electrical Engineering, Ljubljana, Slovenia Dr. Iztok Šorli, Mikroiks d.o.o., Ljubljana, Slovenia Secretary | Tajnik Olga Zakrajšek, UL, Faculty of Electrical Engineering, Ljubljana, Slovenia MIDEM Executive Board Members | Člani izvršilnega odbora MIDEM Prof. Dr. Slavko Bernik, Jožef Stefan Institute, Slovenia Assoc. Prof. Dr. Miha Čekada, Jožef Stefan Institute, Ljubljana, Slovenia Prof. DDr. Denis Đonlagić, UM, Faculty of Electrical Engineering and Computer Science, Maribor, Slovenia Prof. Dr. Leszek J. Golonka, Technical University, Wroclaw, Poljska Prof. Dr. Vera Gradišnik, Tehnički fakultet Sveučilišta u Rijeci, Rijeka, Croatia Mag. Leopold Knez, Iskra TELA, d.d., Ljubljana, Slovenia Mag. Mitja Koprivšek, ETI Elektroelementi, Izlake, Slovenia Asst. Prof. Dr. Gregor Primc, Jožef Stefan Institute, Ljubljana, Slovenia Prof. Dr. Janez Trontelj, UL, Faculty of Electrical Engineering, Ljubljana, Slovenia Asst. Prof. Dr. Hana Uršič Nemevšek, Jožef Stefan Institute, Ljubljana, Slovenia Dr. Danilo Vrtačnik, UL, Faculty of Electrical Engineering, Ljubljana, Slovenia Supervisory Board | Nadzorni odbor Prof. Dr. Franc Smole, UL, Faculty of Electrical Engineering, Ljubljana, Slovenia Prof. Dr. Drago Strle, UL, Faculty of Electrical Engineering, Ljubljana, Slovenia Igor Pompe, retired Court of honour | Častno razsodišče Darko Belavič, Jožef Stefan Institute, Ljubljana, Slovenia Dr. Miloš Komac, retired Dr. Hana Uršič Nemevšek, Jožef Stefan Institute, Ljubljana, Slovenia Informacije MIDEM Journal of Microelectronics, Electronic Components and Materials ISSN 0352-9045 Publisher / Založnik: MIDEM Society / Društvo MIDEM Society for Microelectronics, Electronic Components and Materials, Ljubljana, Slovenia Strokovno društvo za mikroelektroniko, elektronske sestavne dele in materiale, Ljubljana, Slovenija www.midem-drustvo.si