ERK'2019, Portorož, 35-39 35 Device positioning and location-aware communications Klemen Bregar 1;2 , Aleˇ s ˇ Svigelj 1;2 1 Department of Communication Systems, Joˇ zef Stefan Institute, Ljubljana, Slovenia 2 Joˇ zef Stefan International Postgraduate School, Ljubljana, Slovenia E-poˇ sta: klemen.bregar@ijs.si Abstract Modern communication systems can substantially bene- fit from the availability of location information. For ex- ample tracking the user’s location can help predict future position and thus increase the resource allocation perfor- mance. Location information can also help disabled peo- ple live in their apartments in the assisted living facilities. We first introduced fundamental positioning techniques and overlooked the location awareness in the communi- cation systems. We proposed an UWB localization system for passive and active localization with sensor network definition, communication protocol overview and simple active localization technique based on least squares esti- mation. 1 Introduction Knowing the position and being aware of device’s lo- cation in modern communication systems can help im- prove user experience and help improve network opti- mization. It can also help residents in assisted living facil- ities with tracking their activity, detect exceptional events like falling and many more. The localization systems can be very diverse regarding the used technologies and loca- tion estimation principles. Based on the location in the network where the posi- tion of a device is being computed the positioning meth- ods can be divided into two categories [1]: Device-based positioning methods: The device with the unknown location calculates its own lo- cation. The device uses all the measurements to the reference points available in the network and combines them with the localization method of a choice to determine its location. Centralized positioning methods: The central de- vice in the network (e.g. a localization server) uses all available measurements from the network de- vices to the device with the unknown location and combines them in the location of the device. Position can be calculated using different radio sig- nal properties and or other network-related information. They can be classified in the following groups: Multilateration techniques: The localization sys- tem uses range-based information obtained by time of flight (ToF), time of arrival (ToA), time differ- ence of arrival (TDoA) or received signal strength (RSS) between the node with the unknown location and several reference nodes in the network. Triangulation techniques: Intersection of two (for 2D localization) or more direction of arrival (DoA) or angle of arrival (AoA) measurements of the re- ceived signals are used for position calculation. Proximity-based techniques: Location of device is determined by the location of the nearest reference device in the network. Fingerprinting techniques: Fingerprinting tech- niques are actually the pattern matching tech- niques, where a combination of measured signal features (e.g. RSS, multipath components, delay spread, etc.) is matched against the fingerprints in the fingerprint database. Network techniques: Network techniques are based on inter-node distances. Inter-node distances can be described by the connectivity, proximity or range information. Cooperative techniques: Cooperative techniques are used when the centralized localization systems are not feasible, when a scalable localization so- lution for large networks is needed or when the dense network of anchors is available [2]. The co- operation relies on direct communication between the nodes, where individual nodes use available in- formation of neighbouring devices to estimate their own position. Communication systems can benefit from location awareness at many communication protocol layers. For example as a general application of localization aware- ness, the channel database holding location-based chan- nel quality metric can be used for resource allocation at different protocol layers [3]. In the physical layer, location information can be used to reduce interference, increase capacity gains and im- prove the resource allocation performance [3]. It can be used e.g. in combination with Gaussian processes to gen- erate primary user power density maps that enable sec- ondary users in the spectrum to select the most vacant channels for communication [4]. At the MAC layer, location information can help deal- ing with scalability, efficiency and latency challenges 36 [3]. In [5], authors proposed a round-robin schedul- ing algorithm exploiting location information for frac- tional frequency reuse, achieving higher total through- put by allowing temporary sharing of resources between cell-center and cell-edge users. In [6], authors propose location-based multicast MAC protocol that increases the reliability of packet delivery. In [7] authors propose a location-aware power control and scheduling proto- col for IEEE 802.15.3 ultra-wideband (UWB) technol- ogy, allowing the pairs of devices communicate in paral- lel within a piconet, successfully lowering latencies and increasing throughput compared to a traditional round- robin scheduling. Further, location information can im- prove interference prediction and intercell interference coordination techniques in heterogeneous mobile net- works consisting of combination of small cells and macro cells [3]. There are also applications of location-aware MAC protocols in vehicular networks. For example, authors in [8] propose a family of MAC protocols for efficiently broadcasting information only at special communication areas. Omitting the CTS and RTS signals in proposed MAC schemes lowers the message delivery time and im- proves the overall information flow when the flow rate of traffic increases. In [9] a decentralized location-based channel access protocol for intervehicle communication is defined, where unique channels are allocated to differ- ent geographic locations. Finally, authors in [10] propose a geocasting MAC protocol where multicast regions are defined by geographical regions and packets are sent to the nodes in the same geocasting group. At the network layer a location-aided handover mech- anism can reduce number of handovers for 30% com- pared to the RSS-based methods [11]. Another appli- cation of location information at network layer is geo- graphic routing in wireless ad hoc networks [12]. It can provide a scalable and efficient information delivery but is sensitive to localization errors [13]. Geographic rout- ing can benefit from node mobility prediction schemes as in [14]. At higher network layers, location information sup- ports navigation and location-based services [3]. First, there come location-aware information delivery [15] en- abling location-based advertising, sending location-based notification, delivering to-do lists connected to the cur- rent location and many more. Location information can help also with multimedia streaming applications by cap- turing user mobility patterns and predicting future net- work planning events [16]. Regarding the intelligent transportation systems in- tervehicle communication protocols are being developed. The work of [17] provides a location-aware application- layer communication protocol supporting distributed ad hoc, best effort service infrastructure over vehicular ad hoc networks providing drivers time-sensitive informa- tion about traffic conditions and traffic related services on their route. Location information can also be beneficial for autonomous vehicles, tactile internet [18] and other cyber-physical systems as unmanned aerial vehicles and robots. 2 Localization for assisted living Assisted living (AL) systems can substantially benefit from the accurate and reliable location information. Lo- calization can help applications such as activity recogni- tion, mobility pattern analysis and anomaly detection im- prove their reliability while supporting decisions of care- givers and automated systems helping the elderly in their home. AL can benefit the most by exploiting the ultra- wideband (UWB) radio signals because of the high tem- poral resolution of the UWB communication equipment. High temporal resolution enables UWB communication systems to accurately timestamp the beginning of re- ceived and sent frames enabling the localization schemes based on time-of-flight (ToF) information. ToF localiza- tion systems with sub-nanosecond resolution can provide the localization accuracy in centimetre accuracy range. Other benefit of using UWB systems is also great multi- path component (MPC) resolvability, which is a key en- abler for a multipath assisted localization systems and propagation conditions analysis for reducing the impact of NLoS ranges on the localization accuracy [19]. IEEE 802.15.4-2011 standard defining low-rate per- sonal area networks (LR-WPANs) for very low-cost, low- power communications, introduces the UWB technol- ogy as an alternative physical layer (PHY) specifica- tion. UWB PHY implements communication interface based on short pulses with duration in range of 1 ns with bandwidths between 499.2 MHz and 1354.97 MHz all in frequency range between 3.1 to 10.6 GHz. FCC limits the power spectral density emission by - 41.3 dBm/MHz for UWB transmitters in this frequency range. IEEE 802.15.4-2011 UWB standard supports the ranging functionality which enables the unsynchronized localization in low-power sensor networks. 2.1 Localization system For AL applications several localization system types can be used. For the range-based localization active systems are required but for passive systems several approaches based on fingerprinting [20] and radio tomographic imag- ing [21] are suitable. To deploy a localization system that is capable of both active and passive localization we de- fined the minimalistic UWB communication system that is presented in Figure 1. Master anchor device is con- nected to computer with GNU/Linux operating system, two slave anchor devices are connected to the master an- chor device using UWB radio link and slave peripheral device is connected to anchor devices with UWB radio links. For a passive localization process, anchor devices are sending messages to all their neighbour anchor de- vices and measure channel impulse response (CIR) and RSS information for received packets. This information is used for scene analysis and thus passive localization process. First thing when the localization system is turned on is that the master anchor device periodically sends a bea- 37 Figure 1: UWB localization system scheme capable of active and passive localization in assisted living scenarios. con with device discovery request and builds the table of active devices. According to the table of active devices master anchor device then requests the ranging proce- dure for all pairs in the localization network. Based on the measured internode distances local coordinate system is defined where each anchor device gets its reference location. The reference location in the local coordinate system is crucial for locating the nodes or objects with unknown location. When the local coordinate system is defined master anchor device starts coordinating ranging process from the anchor devices to the device with the unknown location in a case of active localization and co- ordinates measurements of CIRs between anchor devices in case of passive localization. 2.2 Communication protocol Communication protocol takes care of communication between al the devices present in a localization network. Communication is organized in a superframe structure with three basic building blocks. First type of communi- cation slot is network management slot (MGMT) where master anchor device requests response from all devices in the network. According to the response the master an- chor device builds a device table that is later used to direct the communication between devices. Next slot type is ranging slot (RNG) where master anchor device master anchor device sends a ranging bea- cons containing the packet ID that represents the ranging request packet, destination and source addresses of the devices that should measure the range between them and duration of RNG slot in microseconds. The last slot type is channel sounding slot (SOUND), where CIR is measured for selected device pairs. Master anchor device at the beginning of SOUND slot sends bea- con packet with sounding request. The sounding request contains the packet ID, source and destination of device pair involved in the sounding procedure and slot duration in microseconds. The structure of a superframe is dynamic and depends on the number of devices present in the localization net- work and desired mode of operation. The network can support three localization modes: active, passive and hy- brid localization mode. The MAC superframe structure Figure 2: Localization MAC superframe structure. is depicted in Figure 2. 2.3 Active localization system We estimate the location of a node with the unknown lo- cation using multilateration approach with least squares (LS) method. As in (1) distance between a node with an unknown locationp and i-th anchorp i is defined by Eu- clidean distance. d i =kp p i k= p (x i x) 2 +(y i y) 2 ; (1) By writing the equations for all anchors and corre- sponding ranges d i we get a predefined system of equa- tions that can be written in a matrix form b = A (2, 3) by introducing a new variable R = x 2 + y 2 where : =[x;y;R] T . A= 2 6 6 6 4 2x 1 2y 1 1 2x 2 2y 2 1 . . . . . . . . . 2x N 2y N 1 3 7 7 7 5 (2) b= 2 6 6 6 4 ^ d 2 1 x 2 1 y 2 1 ^ d 2 2 x 2 2 y 2 2 . . . ^ d 2 N x 2 N y 2 N 3 7 7 7 5 ; (3) By introducing the variable R the system becomes a system of linear equations where location can be es- timated by a linear matrix equation (4) in a normal form. ^ =(A T A) 1 A T b: (4) The system of linear equations (4) can be solved if the matrixA is not singular, which is the case when the anchors are not collinear. In the case of anchor collinear- ity, the rank ofA is less than the number of anchors and therefore the matrixA cannot be inverted which is a cru- cial step in location estimation using LS method. LS-based localization method is computationally one of the simplest methods for location estimation. The drawback of this method is that it is stateless and thus it does not use any information of a previous state of the system. It also does not include any kind of kinematic system model as it is used with Kalman filters. Because of such method simplicity the method use less computa- tional resources than more complex algorithms and can be thus easier to use on small embedded devices which are limited regarding computational resources. 3 Challenges and conclusions In this paper we presented a high-level overview of lo- calization techniques based on location in the network where the location is being estimated. We also presented 38 an overview of localization techniques based on the radio signal properties that are used for localization. We looked also into aspects of location awareness in the general communication systems. Communication technologies can benefit by using the location informa- tion at many communication protocol layers. Location information can be used to reduce interference, to im- prove the resource allocation, to increase throughput, re- duce number of handovers, enable location-based ser- vices and to support navigation. We presented the UWB-based localization system for assisted living applications which supports both active and passive localization. UWB anchor nodes are wire- lessly connected to the master anchor node using UWB technology while measuring range between the nodes and measure channel impulse responses between the nodes for passive localization algorithms. We presented an active localization scheme based on multilateration and least squares location estimation. Passive localization algorithms still represents a chal- lenge for the future work where a combination of prin- ciples from fingerprinting techniques and radio tomo- graphic imaging techniques will be used as a hybrid and easy to install passive localization system. Acknowledgement This work was partly funded by the Slovenian Research Agency (Grants no. P2-0016 and L2-7664) and the Eu- ropean Community under the SAAM (Grant no. 769661) project. References [1] J. A. del Peral-Rosado, R. Raulefs, J. A. L´ opez- Salcedo, and G. Seco-Granados, “Survey of cellu- lar mobile radio localization methods: From 1G to 5G,” IEEE Communications Surveys Tutorials, vol. 20, no. 2, pp. 1124–1148, Secondquarter 2018. [2] H. Wymeersch, J. Lien, and M. Z. Win, “Coopera- tive localization in wireless networks,” Proceedings of the IEEE, vol. 97, no. 2, pp. 427–450, Feb 2009. [3] R. Di Taranto, S. Muppirisetty, R. Raulefs, D. Slock, T. Svensson, and H. Wymeersch, “Location-aware communications for 5G networks: How location information can improve scalability, latency, and robustness of 5G,” IEEE Signal Pro- cessing Magazine, vol. 31, no. 6, pp. 102–112, Nov 2014. [4] I. Nevat, G. W. Peters, and I. B. Collings, “Location-aware cooperative spectrum sensing via gaussian processes,” in 2012 Australian Communi- cations Theory Workshop (AusCTW), Jan 2012, pp. 19–24. [5] A. Dammann, G. Agapiou, J. Bastos, L. Brunelk, M. Garcia, J. Guillet, Y . Ma, J. Ma, J. J. Nielsen, L. Ping, R. Raulefs, J. Rodriguez, D. Slock, D. Yang, and N. Yi, “Where2 location aided com- munications,” in European Wireless 2013; 19th Eu- ropean Wireless Conference, April 2013, pp. 1–8. [6] Min-Te Sun, Lifei Huang, A. Arora, and Ten- Hwang Lai, “Reliable mac layer multicast in ieee 802.11 wireless networks,” in Proceedings Inter- national Conference on Parallel Processing, Aug 2002, pp. 527–536. [7] S. B. Kodeswaran and A. Joshi, “Using location information for scheduling in 802.15.3 mac,” 2nd International Conference on Broadband Networks, 2005., pp. 668–675 V ol. 1, 2005. [8] N. Wen and R. Berry, “Information propagation for location-based mac protocols in vehicular net- works,” in 2006 40th Annual Conference on In- formation Sciences and Systems, March 2006, pp. 1242–1247. [9] S. Katragadda, C. N. S. Ganesh Murthy, M. S. Ranga Rao, S. Mohan Kumar, and R. Sachin, “A decentralized location-based channel access proto- col for inter-vehicle communication,” in The 57th IEEE Semiannual Vehicular Technology Confer- ence, 2003. VTC 2003-Spring., vol. 3, April 2003, pp. 1831–1835 vol.3. [10] Y . . Ko and N. H. Vaidya, “Geocasting in mobile ad hoc networks: location-based multicast algo- rithms,” in Proceedings WMCSA’99. Second IEEE Workshop on Mobile Computing Systems and Ap- plications, Feb 1999, pp. 101–110. [11] H. Celebi and H. Arslan, “Utilization of location information in cognitive wireless networks,” IEEE Wireless Communications, vol. 14, no. 4, pp. 6–13, August 2007. [12] F. Cadger, K. Curran, J. Santos, and S. Moffett, “A survey of geographical routing in wireless ad-hoc networks,” IEEE Communications Surveys Tutori- als, vol. 15, no. 2, pp. 621–653, Second 2013. [13] S. Kwon and N. B. Shroff, “Geographic routing in the presence of location errors,” in 2nd Interna- tional Conference on Broadband Networks, 2005., Oct 2005, pp. 622–630 V ol. 1. [14] J. Li and S. M. Shatz, “Toward using node mobility to enhance greedy forwarding in geographic rout- ing for mobile ad hoc networks,” in In the Interna- tional Workshop on Mobile Device and Urban Sens- ing, 2008. [15] N. Marmasse and C. Schmandt, “Location-aware information delivery withcommotion,” in Handheld and Ubiquitous Computing, P. Thomas and H.-W. Gellersen, Eds. Berlin, Heidelberg: Springer Berlin Heidelberg, 2000, pp. 157–171. 39 [16] Baochun Li and K. H. Wang, “Nonstop: continuous multimedia streaming in wireless ad hoc networks with node mobility,” IEEE Journal on Selected Ar- eas in Communications, vol. 21, no. 10, pp. 1627– 1641, Dec 2003. [17] M. D. Dikaiakos, A. Florides, T. Nadeem, and L. Iftode, “Location-aware services over vehicular ad-hoc networks using car-to-car communication,” IEEE Journal on Selected Areas in Communica- tions, vol. 25, no. 8, pp. 1590–1602, Oct 2007. [18] G. P. Fettweis, “The tactile internet: Applications and challenges,” IEEE Vehicular Technology Mag- azine, vol. 9, no. 1, pp. 64–70, March 2014. [19] K. Bregar and M. Mohorˇ ciˇ c, “Improving indoor lo- calization using convolutional neural networks on computationally restricted devices,” IEEE Access, vol. 6, pp. 17 429–17 441, 2018. [20] S. He and S. . G. Chan, “Wi-Fi Fingerprint-Based Indoor Positioning: Recent Advances and Compar- isons,” IEEE Communications Surveys Tutorials, vol. 18, no. 1, pp. 466–490, Firstquarter 2016. [21] D. Romero, D. Lee, and G. B. Giannakis, “Blind Radio Tomography,” IEEE Transactions on Signal Processing, vol. 66, no. 8, pp. 2055–2069, April 2018.