Localization system improvement using a special designed sectorised antenna

Localization System Improvement using a Special Designed Sectorised Antenna Luis Brás, Marco Oliveira, Ludimar Guenda, Nuno Borges Carvalho

Pedro Pinho Instituto Superior de Engenharia de Lisboa, Lisboa, Portugal

Departamento de ETI, Telecomunicações e Informática, Instituto de Telecomunicações, Universidade de Aveiro Aveiro, Portugal Abstract— This paper presents the impact of different antenna strategies for localization scenarios evaluation, based on ZigBee wireless sensor networks (WSN). On our first strategy the Received Signal Strength (RSS) is measured and evaluated using a chip antenna on the mobile unit, and a patch antenna on the WSN fixed nodes. Then, we present another strategy where all fixed nodes are exchanged by a new type of antenna called Hive5, a sectorised antenna that combines RSS and Angle of Arrival (AoA) in the same node. This antenna will improve significantly the localization system, either in resolution, and/or cost. Keywords- localization systems; sectorised antenna; horn antenna; RSSI; AoA.

I.

switch

beam

INTRODUCTION

Localization or estimation of spatial distance between objects has been a classical problem of last decades. These systems integrated with wireless sensor networks (WSN) led to the development of an huge open market of applications such as: people/asset management and tracking, specially applied to children, hospital patients and consumers tracking; health care and mobile patient monitoring; emergencies support; games interaction, autonomous tour guides. Outdoor localization systems are commonly based on Global Navigation Satellite Systems (GNSS)[1], such as Global Positioning System (GPS) Global Orbiting Navigation Satellite System (GLONASS) and Galileo or techniques based on cellular network, have been explored and well standardized. Although these systems provide global coverage they do not provide high positioning accuracy, do not work in indoor environments and present a high power consumption solution. When we speak about indoor localization, a well accepted standard by the community is not yet reach, being the technology and localization technique chosen according with a tradeoff of desire characteristics: cost, accuracy, power consumption, size, availability, mobility and interoperability. As localization systems we could for instance, refer several interesting implementations such as: Smart Floor [2] based on physical contact, Easy Living [3] based on image processing, Cricket [4], LANDMARC [5], Ekahau [6], ZigBee localization engine, based on RF and ultra-sounds signals.

Previous systems performance is highly related to implemented positioning technique which can be divided as Received Signal Strength Indication (RSSI), Time of Arrival (ToA), Time difference of Arrival (TDoA), Angle of Arrival (AoA) or either hybrid solutions [7]. As a brief description of each, systems based on time of flight (ToA and TDoA) of electromagnetic waves provide a high accuracy solution, although, they require an infrastructure extremely well synchronized leading to higher system costs; AoA demand extra hardware and is highly dependent of antennas accuracy on the reference nodes; finally RSSI systems are the simpler ones and consequently the less expensive. The main drawback of these systems is that RF signal is sensitive to various phenomena such as multipath fading and shadowing, as a result, received signal strength is not very accurate and can vary a lot over a short distance. These systems typically use channel propagation models or RF fingerprinting to estimate the user positioning. Especially for localization systems based on RSSI and AoA the choice of a proper antenna is a crucial issue to have in consideration for the good performance of the system. There are several antennas types used for localization system, each of them with specific characteristics according to size, cost, radiation pattern, polarization and gain. We could for instance refer several simple antennas types such as: dipoles, monopoles, helical, IFA (Inverted F Antennas), chip, microstrip or even antennas arrays [8]. Among the referred antennas, microstrip antennas show to be very versatile, being light and inexpensive, presenting formats and polarization diversity, and easy to use for antennas arrays [9]. Other group of antennas, although more complex, are the “smart antennas” [8]. These antennas can be divided as switched beam (sectorised) – providing a finite number of fixed predefined radiation patterns; or adaptive array – providing an infinite number of radiation patterns adjustable in real time. Some examples of sectorised antennas arrays are presented in [10] and [11].

In this paper we present the impact of different antennas strategies for the localization scenario evaluation, based on a ZigBee WSN. The first implemented strategy was performed using a WSN mounted with six reference nodes under an area or 7000m2 with LP patch antenna. The second strategy is based on a replacement of entire infra-structure by a new type of antenna which provides higher resolution, the Hive5. This antenna is based on six patch antennas independent orientated, separated by a metallic structure allowing the combining of RSSI with AoA measurements. This antenna was specially designed to reduce coupling between the neighborhood antennas, increase multipath rejection, gain and directivity.

For the set up of the localization system we defined a scenario of 7000m2 covered by six reference nodes. These sensors, presented on the map as RNx were integrated on light posts around the localization scenario at a height of 3 meters. A mapping of the measured scenario with a grid of 88 points spaced 10 meters apart was superimposed to our map as presented in Fig. 3.

Based on previous introduction this paper is organized as follows: section II presents the set up of the implemented localization system. In section III we describe the measured scenario propagation model. After, in section IV we present the localization system with the use of the Hive5 antenna. Finally some conclusions and future work are presented in section V. II.

SETTING UP THE LOCALIZATION SYSTEM

As previous referred the implemented localization system is based on a ZigBee WSN. The reference nodes were integrated with a LP patch antenna, presented in Fig.1, designed for 2.44 GHz, bandwidth around 3.3%, and a gain of 4.03 dBi.

Figure. 3: Localization Scenario Map

After this testing grid implemented, the mobile node was moved between each testing point with antenna orientation co polarized with the reference nodes antennas. III.

MEASURED RESULTS

Accomplishing the previous set up, the RSSI of the mobile node were then measured on the ground level at an average of 10 measures for each point. Figure. 1: Reference Node LP Patch Antenna

The measured propagation model of FUSCA chip antenna is presented on Fig. 4, as a 3D plot, with the same axis of the localization scenario map.

The mobile nodes, presented in Fig. 2, were integrated with a LP FUSCA chip antenna, with 0.4 dBi of gain which can be seen in Fig.2.

Figure. 2: Developped Mobile Node

Figure. 4: FUSCA Chip Antenna Graphical Propagation Model

It can be noticed that the measured RSSI values present high variance among the measurement scenario. Analyzing the

propagation model, several zones are almost, or in some cases completely not detected by the WSN. So, inevitable, some areas will be detected just by one reference node leading to high uncertain of the mobile node positioning. A solution to significantly reduce this positioning uncertain could be achieved by the use of Hive5 antenna. With this antenna, the uncertainty of the positioning would be significantly reduced, combining the sectorised detection with RSSI and AoA techniques.

#P3

#P4

#P1 #P5

#P2

The constraint of time costly set ups and big WSNs would also be significantly reduced, replacing them by a single one or few Hive5 antennas. The Hive5 is an improved sectorised antenna that cans significantly improving localization systems performance as will be described in next section. IV.

#P6

THE HIVE5 ANTENNA

The Hive5, shown in Fig. 5, is a sectorised antenna with a semi-dodecahedron format which can identify six independent areas. This antenna was based on the sectorised antenna presented in [11] and horn antennas, providing an integration of AoA technique with a RSSI based localization system.

Figure. 6: Simulated Hive5 antenna design

The simulated Hive5 antennas show a gain of 5.67 dBi. This represents an improvement of around 1.13dB compared with patch antenna out of the Hive5 structure. V.

PROPOSED LOCALIZATION SOLUTION

The desired implementation for the Hive5 antenna is presented in Fig. 7. As can be seen, the Hive5 antenna is settled on the center of the localization scenario, at a high position, being able to discriminate six independent zones. These six zones can be well seen by the diagram patterns achieved with HFSS and presented in Fig. 8. This antenna selection can be performed by a reference ZigBee module which controls a RF demux.

Figure. 5: Implemented Hive5 Antenna

With an appropriate protocol, these modules will then be able to discriminate mobiles nodes RSSI and AoA, associated to each individual antenna. This information is then sent to the main coordinator of the WSN and then processed by a localization algorithm in order to estimate mobile nodes correct positioning.

The origin of this antenna name was inspired on a beehive, but instead of a hexagonal shape we use a pentagonal, from there the Hive5 name. Actually, our antenna is formed by six simple patch antennas with ellipsoidal left hand polarization, centered on 2.45GHz. The hollow structure has impact not only on each patch antenna gain but also in his polarization. Taking into consideration this factor, the patch was dimensioned in order to compensate this factor and guarantee circular polarization when inserted into the hollow structure. The presented antenna was simulated using Ansoft HFSS (High Frequency Structure Simulator) which help us to design a good correction of the patch antenna polarization. With this, was possible to compensate the polarization changes due to the hollow pipe structure of the Hive5. Figure. 7: Localization system based on Hive5 antenna

#P1

#P4

#P2

#P3

#P5

#P6

Figure. 8: Localization system based on Hive5 antenna

VI.

REFERENCES

CONCLUSIONS

This paper presents the impact of different antennas strategies on localization systems. The first based on a WSN with six reference nodes, and the second, an alternative to substitute the reference node by a single one, the Hive5. The localization system integrated with the proposed Hive5 is a cost effective solution for reduce the number of infrastructure reference nodes and improve system resolution by the combining of RSSI with AoA techniques. ACKNOWLEDGMENT This work was supported by Fundação da Ciência e Tecnologia (FCT), reference SFRH/BD/61834/2009, Lisbon, Portugal, and implemented in Instituto de Telecomunicações Labs, Aveiro.

[1]

B. Hofmann-Wellenhoff, H. Lichtenegger Elmar Wasle, “GNSS – Global Navigation Satellite System”, Fourth Edition, Springer Verlag, 2008. [2] R.J Orr and G.D. Abowd, “The smart floor: A mechanism for natural user indetification and tracking”, Proceedings of Human Factors in Computing Systems (Hague Netherlands) pp. 275-276 April 2000. [3] J. Krumm,S. Harris, B. Meyers, B. Brumitt, M. Hale, and S. Shafer “Multi-Camera Multi-Person Tracking for EasyLiving” Third Workshop on Visual Surveillance Redmond, WA, USA June 2000. [4] N.B Priyantha “The cricket indoor localization system”, Ph.D thesis, Massachusetts Institute of Technology Massachusetts Institute of Technology June 2005. [5] L.M. Ni, Y. Liu, Y.C. Lau, and A.P Patil “Landmarc: Indoor localization sensing using active RFID Texas”, USA March 2003. [6] Heidari, M. ; Pahlavan, K. “Performance evaluation of indoor geolocalization systems using PROPSim hardware and ray tracing software” International Workshop on Wireless Ad-Hoc Networks 2004, Worcester Polytech. Inst., MA, USA, October 2005. [7] Hui Liu, Darabi H., Banerjee P., Jing Liu, “Survey of Indoor Positioning Techniques and Systems”, Part C: Applications and Reviews, IEEE Trans., vol. 37, no. 6, pp.1067-1080, November 2007. [8] Dr. John Volakis, “Antenna Engineering Handbook, Fourth Edition”, Chapter 25: “Smart Antennas”, McGraw-Hill Professional, 2007. [9] Dr. Rodney B. Waterhouse, “Microstrip Patch Antennas A designer's Guide”, Chapter 1, pag 7-10, 2003. [10] Nasimuddin Z,. N.Chen, X.Qing. T.S.P.See, “Sectorised antenna array and measurement Methodology for indoor ultra-wideband applications”, IET Microwave Antennas Propagation, vol.3, no.4, pp.621-629, 2009. [11] Cidronali, A. et al, “Analysis and Performance of a Smart Antenna for 2.45-GHz Single-Anchor Indoor Positioning”, IEEE Transactions on Microwave Theory and Techniques, vol. 58, no. 1, pp. 21-31, 2010.