Angle of Arrival Estimation Using RSSI and ... - IEEE Xplore

1950

IEEE SENSORS JOURNAL, VOL. 12, NO. 6, JUNE 2012

Angle of Arrival Estimation Using RSSI and Omnidirectional Rotatable Antennas Marko Malajner, Peter Planinšiˇc, and Dušan Gleich

Abstract— This paper proposes a simple novel method for angle of arrival estimation using multiple rotating omnidirectional antennas on a receiver device. The used omnidirectional microstrip antenna has a symmetrical radiation pattern with sharp minimum along the x antenna axis. An algorithm based on the fact that an angle of arrival is obtained along a direction where the measured received strength signal indicator is minimal. Our experimental results for the outdoor measurements reached a mean error of less than 4°, and an indoor of less than 6°. Index Terms— Angle of arrival (AoA), omnidirectional antenna, received signal strength indicator (RSSI), wireless sensor network (WSN).

I. I NTRODUCTION

R

ECENT advances in micro-electro-mechanical systems technology, wireless communications, and digital electronics, have enabled the development of low-cost sensor nodes capable of communicating with each other over short distances [1]. There are many applications where WSN can be used: from military applications to healthcare and environmental monitoring. These small sensor nodes consist of the radio part for spreading data, the sensor part for sensing environmental phenomena, the processing unit, and the power supply. Nodes are connected to a base station, where data from each sensor node is collected. The localization of nodes within a wireless network is usually unknown, therefore node location or position estimation has become a very interesting research topic over recent years. Localization information from sensor nodes is important for inventory tracking, intruder detection, the tracking of firefighters or miners, home automation, and patient monitoring. The time of the signal (acoustic or RF) arrival (ToA), the Angle of signal Arrival (AoA) or the received power of the signal, i.e. Received Signal Strength Indicator (RSSI), are usually the indicators used for localization [1]. The ToA is costly because the speed of light is very fast, approximately 106 times the speed of ultrasound. In order to calculate the

Manuscript received October 19, 2011; revised December 6, 2011; accepted December 24, 2011. Date of publication December 28, 2011; date of current version April 25, 2012. The associate editor coordinating the review of this paper and approving it for publication was Prof. Gerald Gerlach. The authors are with the Faculty of Electrical Engineering and Computer Science, University of Maribor, Maribor 2000, Slovenia (e-mail: [email protected]; [email protected]; [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JSEN.2011.2182046

ToA parameter, the nodes must have a common clock, or exchange timing information using certain protocols as, for example, two-way ranging protocol [2, 3]. Most 802.11 and 802.15.4 radio modules support the RSSI, which enables the calculation of received power for each received packet. The power or energy of a signal travelling between two nodes (transmitting and receiving) is a signal parameter which can be used for distance estimation, together with a path-loss and shadowing model. RSSI measurements are very unpredictable because there are several error sources such as the many delayed multipath signals arriving at the receiver [4, 5]. Instead of ToA or distance, the AoA can be measured and used for localization. Directions to neighboring sensors can be obtained using this parameter. There are two common ways that sensors measure the AoA. The most common method is to use antenna array, where the array geometry is known, and the measuring of different signal arrival times is done at different antennas [6, 7]. The second approach to AoA estimation uses the RSSI ratio between two or more directional antennas on the sensor node [8]. Most researchers use directional rotatable antennas and measure the peak power of the received signal [7, 9]. The authors in [10] reported the measurement of AoA using two rotatable omnidirectional antennas. They used transmitters with horizontally polarized omnidirectional antenna, and a receiver with two vertically polarized antennas. Using a complex algorithm they could obtain AoA from the received power of an RF signal. This paper presents a simple method for AoA measurement that proposes an omnidirectional rotatable monopole antenna, using RSSI measurements. The radiation shape of the used monopole antenna is not ideally isotropic and has a minimum in a direction of antenna axis. The idea presented in this paper is to find a minimal measured RSSI by rotating this antenna. Searching for the minimum was chosen because better selectivity is obtained due to the radiation pattern of the antenna. A minimal of two antennas placed on the circle are needed in order to detect direction of arrivals from two opposite sides. Four antennas are used in the proposed approach, in order to improve the accuracy of estimation the AoA, as a compromise between accuracy and cost. The main advantage of the proposed method is that a commercial radio module with already-equipped microstrip omnidirectional antenna can be used, instead of using antennas with massive directional reflectors or special antenna arrays, and specially shaped antennas. The rest of this paper is organized as follows: Section II describes the idea of our approach, based on radiation pattern

1530–437X/$26.00 © 2011 IEEE

MALAJNER et al.: AoA ESTIMATION USING RSSI AND OMNIDIRECTIONAL ROTATABLE ANTENNAS

dB (Gain total) 2.0921e + 000 1.1359e − 001 −1.8649e + 000 −3.8435e + 000 −5.8220e + 000 −7.8005e + 000 −9.7791e + 000 −1.1758e − 001 −1.3736e + 001 −1.5715e + 001 −1.7693e + 001 −1.9672e + 001 −2.1650e + 001 −2.3629e − 001 −2.5607e + 001 −2.7586e + 001 −2.9564e + 001

Fig. 1. 3-D radiation pattern of MRF24J40MA’s antenna (Phi = ϕ, Theta = θ ) [13].

and RSSI. Section III presents the used hardware, Section IV presents the experimental results, and Section V concludes the paper.

1951

Radiation pattern 1 0 −30

Final comb 50 ohms analyze Curve info dB(Gain Total) Setup 1: Last adaptive Phi = ‘0 deg’

30

m1 −2.00 −9.00

−60

dB(Gain total) Setup 1: Last adaptive Phi = ‘90 deg’

60

−16.00

m6

−23.00

m2

−90

90

m5

−120

m3

m4

−150

120

150

180 Fig. 2.

H-plane (violet curve) and E-plane (red curve) [13].

z Antenna

II. BACKGROUND AND BASIC I DEA OF THE P ROPOSED A PPROACH WSN is composed of sensor nodes distributed randomly within a sensor field. The objective of this paper is to provide a novel simple method for the accurate estimation of AoA for localization of the sensor nodes. In literature, many methods based on antenna arrays and directional antenna are used for AoA estimation. Advantage was taken of the radiation pattern shape of the monopole antenna, as used in almost all commercially manufactured radio modules. In [7] the commercial XBee radio module with omnidirectional antenna and additional mounted reflector is used for AoA estimation. The maximal RSSI is used for the directional detection of an arrived signal. In the proposed method, four rotatable omnidirectional antennas are used and the direction of the received signal is determined by measuring the minimum of RSSI. Let’s imagine that an infinitely small antenna has an ideal radiation pattern represented by spherical shape, which radiates isotropic. This means that the antenna radiates equallywell in all directions. But the real antennas have some degree of directivity, which mean for some directions the radiation is stronger, than in others. The radiation pattern can be found by measuring the strength of the field at every point on the surface of a sphere. The performance of a linearlypolarized antenna, which is almost an approximation of an ideal antenna, is described with E- and H-plane patterns. E- and H-planes represent the fields in which the electric (electric-field intensity) and the magnetic (magnetic-field intensity) lines of force lie, respectively [11]. The H [A/m]and E[V/m]- planes are perpendicular to each other. The cross product of E and H* (complex conjugate) divided by 2, gives the time-average Poynting vector of the radiated field: 1 (1) Wav (x, y, z) = Re E × H∗ (W/m2 ) 2 where r is the radial distance. The E and H planes represent the peak values, and should be divided by 2 for the RMS

y

Ground plane

x Fig. 3. Microstrip feeding monopole as part of MRF24J40MA module [13].

values. Based on (1) the average power radiated by a general antenna can be written as: (2) Prad = Pav = Re(E × H∗ ) · ds S

where S is a spherical surface surrounding by the antenna [12 pp. 38-39]. Equations (1) and (2) are valid for far-field. Farfield is located at grates radial distances r from the antenna. For simple wire antennas, the radial distances r for far-field is at least a half wavelength from the antenna’s radiating center. All our measurements are executed in ‘far-field’. The quarter-wavelength monopole, as is used in our experiment, can be represented as a half-wavelength dipole, explained in [11 pp. 2-17]. For half-wavelength dipole, E- and H-plane can be written as [12 pp. 182]: I0 e− j kr cos π2 cos θ (3) Eθ j η 2πr sin θ I0 e− j kr cos π2 cos θ Eθ Hϕ j (4) η 2πr sin θ

1952


y Radiation patterns E-planes

y'

y Radiation patterns E-planes

x'

Rx3

Rx3

Rx4 x Rx2

x Rx2

Rx4

Rx1 Rx1 φ φˆ = 0

φ = φˆ Tx

Tx ω

ω RSS1-Rx1 RSS1-Rx2 RSS1-Rx3 RSS1-Rx4

RSS1-Rx1 RSS1-Rx2 RSS1-Rx3 RSS1-Rx4 (a)

(b)

Fig. 4. Top view of measuring device, and the coordinate system for measuring AoA. (a) Initial set-up with unknown position of transmitter (Tx), bar graphs show the measured RSSI. (b) Scenario where the true AoA is obtained. The minimizations of the measured RSSI on Rx1 and Rx3 can be observed on the bar graphs.

where η is the intrinsic impedance of the medium, r is the radius distance, θ is the azimuth angle and is the latitude angle. I0 is a traveling wave current. The 3-D omnidirectional radiation pattern in Fig. 1 has an infinite number of principal E-planes (elevation planes), and one principal H-plane (azimuthal plane), as shown in Fig. 2 [13]. The microstrip feeding quarter-wavelength monopole antenna [14] was used in our experiment. The microstrip antennas are low-profile, low weight, easy to produce, low cost, and mechanically robust [14]. Cooper printed in the FR4 laminate presents the microstrip feeding monopole antenna, as shown in Fig. 3 [15]. The antenna used in our experiments is a part of the MRF24J40MA ZigBee module [13]. Our antenna can be considered as an omnidirectional monopole antenna because of named characteristics. The specific characteristics of the used MRF24J40MA’s antenna’s radiation pattern are shown in Fig. 2. The radiation patterns in Fig. 1 and Fig. 2 are taken from the manufacture’s datasheet [13]. It can be observed that the pattern have all the characteristics of an omnidirectional monopole or dipole, respectively. The radiation pattern consists of two zones of decreased intensity at 90 and −90 degrees, as observed along the x axis on Fig. 1 and in Fig. 2. The strength of the received signal depends on the radiation pattern of an antenna. RF signal, which arrives in the direction of a radiation pattern with less sensibility, has lower RSSI than a signal that arrives in the direction of a larger antenna sensibility. In order to find the minimal measured value of RSSI the omnidirectional antenna is physically rotated within an E-plane. The azimuth angle of rotated antenna where the RSSI is minimal, is assume to represents estimate AoA. The radiation pattern is symmetrically regarding as an antenna axis therefore the AoA estimation can be mismatch by 180°.

In order to solve this problem, two antennas are needed, however the proposed novel method uses four antennas placed in four directions, as shown in Fig. 4. Four receiving antennas Rx1-Rx4 are symmetrically mounted on a circular plate with diameter d = 12 cm, as shown in Fig. 4. The plate can be rotated counter-clockwise by ϕˆ over a 3.6 degree step, using custom-made hardware, as described in Section III. For the explanation of this method, it is first assumed that the receivers and transmitter lie within the same plane. This situation is common in many applications. Fig. 4a shows the initial position of the receiver with four antennas and a transmitter. The AoA is measured with reference to a local coordinate system which is defined by its initial position. Fig. 4b outlines the proposed approach. It shows that minimum RSSI is obtained when the x axis of the receiving antenna lies on the same line as the transmitting antenna. The bar-graphs of the measured RSSI for the scenarios, are shown in in Fig. 4a and Fig. 4b. By rotating the antennas, the true . azimuth angle ϕ is estimated: ϕ = ϕ. ˆ It can be seen that the minimal RSSI values in Fig. 4b correspond to the antenna Rx1 and Rx3, which are aligned with the Tx-antenna, however Rx3 received a smaller value due to its greater distance from Tx, and due to differences in the receiving sensibility in x axis, see Fig. 2. This is the main reason for using two opposite antennas for the estimation of true AoA. The positions of Rx1 and Rx3 therefore determine the estimated AoA ϕ. ˆ The true angle was, in our design case, known in advance. III. H ARDWARE The proposed device for AoA measurement consists of a four transceivers placed on circle plate (Fig. 5a and Fig. 6a), a stepper-motor (Fig. 6c), a driver for the stepper-motor


1953

Start of algorithm

Transmitter [Fig. 6(f)] sends data packet

(a) (b) Fig. 5. (a) AoA measuring device and (b) SPaRCMosquito module as transmitter.

Receiving AoA device [Fig. 6(a) and (b)] receives packet in all antennas

Board in [Fig. 6(b)] sends data packet with RSSI to board in [Fig 6(e)]

End of procedure on measurement device

Searching for minimum RSSI on PC collected RSSI values (curves) for each antenna

RSSI curves of each antenna are shifted and averaged

40 received packet?

Fig. 6. Hardware set-up (a) plate with four MRF24J40MAs, (b) processor board 1 which controls four transceivers, (c) stepper-motor, (d) driver for stepper-motor, (e) processor board-2 which controls the stepper-motor and serves as a wireless interface between the laptop and processor board-1, and (f) transmitter node with MRF24J40MA.

(Fig. 6d), a processor board-1 (Fig. 6b), and board-2 (Fig. 6e). Both boards are based on our SPaRCMosquito WSN module with a Cortex M3 NXP microcontroller, DC/DC converter, integrated MRF24J40MA transceiver and battery-powered source, as shown in Fig. 5b [16]. The transceivers are the Microchip MRF24J40MA [13] with microstrip monopole antenna operating in the ISM 2.4 GHz band. A 200 stepsper-revolution motor is used to rotate the plate with transceivers. The stepper-driver is the well-known L298. The plate with four transceivers is attached on the board-1. The entire device for measuring AoA, together with the power-source, is mounted on the axis of the stepper motor. The slipping rings for communications and power supply are eliminated because of this. Board-2 is also provided for controlling the stepper-motor. During the experiments, the stepper motor turns the plate using receivers with resolutions of 3.6°. The transmitter continually transmits on RF signal. For each position of the plate, the receivers measure RSSI and then the data is sent to board-1, which wireless communicates with board-2. Board-2 is connected via USB to laptop where the data is collected and later processed to estimate AoA. IV. E XPERIMENTAL V ERIFICATION OF THE P ROPOSED M ETHOD The presence of a moving object or humans in the vicinities of the transceivers causes large fluctuations in the RSSI

Board in [Fig. 6(e)] computes averaged RSSI for each antenna and send them to PC together with corresponding angle of stepper motor

Board in [Fig. 6(e)] turn stepper for 3.6°

Searching for minimum on averaged RSSI cuurves to determine AoA

Comparing of RSSI of minimums of two opposite antennas to determine the direction of arrival (from one or opposite sides)

Stepper motor angle is 356.4°?

Fig. 7.

End

Algorithm for AoA estimating.

measurements, therefore all the experiments were carried-out within environments without obstacles and moving objects, and in the line of sight (LOS). The experiments were conducted within both indoor and outdoor environments for different distances between transceivers. The completed experiments were repeated three times at the same configurations of the transceivers in both environments. The whole algorithm for estimating AoA with data acquisition on measuring device side is presented in flow chart in Fig. 7. Equations (3) and (4) are not directly linked to algorithm. Equations describe the general radiation pattern of dipole antenna with two decreased sensibility zones. The obtained via RSSI measured radiation patterns of

1954


−35

−45 −50

−40

−55 RSSI (dBm)

RSSI (dBm)

−45 −50 −55

−65 −70

Rx1 Rx2 Rx3 Rx4 true angle

−60 −65 0

−60

50

100

150

200 250 300 AoA (degrees) (a)


−75 −80

350

−85 −300 −200 −100

400

Fig. 10.

−35

RSSI (dBm)

RSSI (dBm)

−40

−50 0

Rx1 Rx2 Rx3 Rx4 true angle 50

100

150

200 250 300 AoA (degrees) (b)

350

−80 6 4 2 Range (m) 0

Angle error (degrees)

−50 −55 −60 Rx1 Rx2 Rx3 Rx4 true angle 0 100 200 AoA (degrees)

300

100

0

200

400

300

Angle (degrees)

Fig. 11. 3-D view of outdoor measured RSSI versus AoA versus distance between measuring device and node. The blue plane represents the estimated AoA, which is almost nondependent on distance.

2

−45 RSSI (dBm)

−70

4

Fig. 9.

400

−60

−40

−75 −300 −200 −100

300

−50

−35

−70

200

RSSI versus Angle versus Range Estimated angle

−40

400

Fig. 8. Arrangement of local minimas. (a) Measured RSSI versus angle of arrival. (b) Measured RSSI with a 90° rotated transmitter.

−65

100

Outdoor measurements of RSSI versus AoA at a distance of 2 m.

−30

−45

0

AoA (degrees)

0 −2 −4 −6 Estimated AoA Mean

−8 −10 400

Arrangement of local minimas.

rotating antennas is shown in Fig. 8a where minimums of strength of received signal are clearly visible. The goal of algorithm is to find the AoA where minimums of RSSI appear. A. Preliminary Measurements The first preliminary measurements were taken within an indoor environment with no obstacles between the

−12

0

1

2

3 4 Distance (m)

5

6

Fig. 12. Error of the estimated AoA from all outdoor tests as a function of the distance between transmitter and receiver. Small circles are the mean of three AoA calculations at each location.

transceivers (LOS) over short distance between transceivers of approximately 15 cm, by putting both the receiver plate and transmitter on a desk. The results of the measurements are shown in Fig. 8a, and represent the values of RSSI for each antenna at certain angles, obtained by rotating the plate. In initial set-up the azimuth angle between the transmitter and receiver was 54°.


1955

Tx Tx

−30

90°

RSSI versus Azimuth angle versus latitude angle

60° −40

30°

Latitude angles

Tx

Receivers 0°

θ Tx −30° Tx −60°

ω

−90° Tx

Azimuth angles φ

RSSI (dBm)

Tx

−50 −60 −70

0 −30 −60 Latitude angle (degrees) −90

(a) Fig. 13.

90 60 30

100

0

(b)

200

300

400

Azimuth angle (degrees)

(a) Different height positions between the receivers and the transmitter. (b) Dependence of RSSI on latitude angle and azimuth angle.

Fig. 8a shows the graph RSSI vs. rotating angle ϕˆ from 0 to 360°. Both the transmitter and receivers are configured for an RF power-level of 0 dBm. The true-angle between the initial position of receiver Rx1 and the transmitter was set at 54°, as shown in Fig. 4a. The RSSI of the corresponding receivingantenna achieved a minimum RSSI when its axis was aligned with the transmitter’s antenna. Experiments were also conducted with a 90° rotated transmitter within the azimuth plane, and obtained practically the same results see Fig. 8b. The RSSI curves mainly depended on the radiation pattern of the receiver and less on the distance to the transmitter. A Search was made for the appearance of the first minimum during AoA estimation. In this case the antennas Rx1 and Rx3 first achieve minimums RSSI approximately at angle of 54°. It can be concluded from the RSSI curves, that antenna Rx1 was closer to the transmitter, because its minimum value was higher than the RSSI of antenna Rx3; thus indicating that the AoA was 54° and not 54+184°. The same conclusion can be made for an angle of around 140° because receiving antennas are physically shifted for angle 90°, where the minimum RSSI values were achieved by antennas Rx2 and Rx4, but the RSSI of Rx2 was higher than the RSSI of antenna Rx4. Imprecise positional initialization involved on offset during AoA estimation. In order to minimize the effect of measurement uncertainty, such as random fluctuation of RSSI and plate-vibration, multiple measurements (40) were taken for each position. For additionally improving accuracy when estimating AoA from the RSSI measurements presented in Fig. 8a, RSSI curves of the antennas vs. the angles of Rx2, Rx3, Rx4 for 90°, 180° and 270° were shifted, respectively. The new graph is shown in Fig. 9. The RSSI curves in Fig. 9 were then averaged and the minimum of averaged curve determined for estimating the AoA. B. Outdoor Experiments The outdoor experiments were carried-out on an asphalt floor in a line-of-sight between transceivers. The transceivers

were placed at a 1 m height from ground, on a rack. The true AoA was set at 54°. Fig. 10 shows the measured RSSI versus rotating-angle at a distance between the transmitter and the receiver of 2 m. The RSSI values of each antenna were processed using the procedure describe in Section II. Then the transmitter was placed at distances from 1 m to 6 m, with steps of 1 m. The dependences between the measured RSSI values, the distance between the transmitter and antenna Rx1, and the rotational angle are depicted in the 3-D graph of Fig. 11. It can be concluded from Fig. 11, that the same algorithm for AoA estimation can be used at each distance, because the distance has little influence on the shapes of the RSSI curves. Fig. 12 shows the average of errors between the estimated and true AoA for three complete estimation procedures at different distances. The maximum absolute mean error of the estimated AoA was about 4°. The next experiment described the scenario when the transmitter and receiver were placed on a parallel plane at different heights. The radiation pattern, as shown in Fig. 1, indicates that the estimation of an azimuth angle is possible with the proposed method at small latitude angles. This indication was confirmed by measurements presented in Fig. 13, where the RSSI curves were measured at different latitude angles between the transmitter and receiver. Measurements were taken at 6 different latitude angles, from −90° to 90° over 30° steps (Fig. 13a). An azimuth angle was set up at 54° for all measurements. At greater altitude angles, the RSSI minimums are not sharp enough for estimating AoA. Fig. 14 shows the errors for the estimated azimuth angle versus latitude angles, and confirms that the error was small for those latitude angles nearer zeros. C. Indoor Experiments The indoor experiments were carried-out in LOS in a 6 mby-8 m room in LOS between transceivers; the true azimuth AoA was set at 54°. The transceivers were placed at 1 m in height. The experiments were carried-out in the same way as

1956


50

−40

Estimated angle

RSSI (dBm)

Azimuth angle error (degrees)

40 30 20 10

−50 −60 −70 −80 6

0

RSSI versus Angle versus Range Estimated angle

−10

4

−20

2 0

−40 −100 −80 −60 −40 −20 0 20 40 Latitude angle (degrees)

60

Fig. 16. 3-D view of indoor measured RSSI versus AoA and distance between measuring device and transmitter.

10

Angle error (degrees)

15

−50 −55

RSSI (dBm)

20

−45

−60 −65 −70 −75 −80

0 100 200 AoA (degrees)

Estimated AoA Mean

5 0 −5 −10


−85 −300 −200 −100

0

400

80 100

Fig. 14. Estimated azimuth angle error versus different latitude angle. When the latitude angle is not close to zero it is hard to accurately estimate the azimuth angle.

Fig. 15.

300 200 100 Angle (degrees)

Range (m)

−30

−15 300

400

Indoor measurements of RSSI as a function of azimuth angle.

the outdoor experiments, as described in Section IVB. Fig. 15 shows the measured RSSI versus azimuth angle. The indoor measurements were influenced by reflected signals from the walls, floor, ceiling, furniture, and other objects. Therefore the RSSI curves were not as smooth, having more local minimums and maximums, and it was more difficult to accurately estimate the AoA than during outdoor measurements. Fig. 16 shows the dependence of antenna Rx1 RSSI values on the rotating azimuth angle and distance. The distances between the receivers and the transmitter were set at 1 m to 5 m, by steps of 1 m. The blue shaded plane presents the true-angle. Distance also had little affect on the accuracy of the AoA estimation, in this case (see Fig. 17). Fig. 17 shows the errors between the estimated and true AoA over different distances. The angle was estimated three times using the complete proposed algorithm. All three estimates were averaged and the errors plotted, as in Fig. 17. The maximum absolute mean error was about 6°. V. C ONCLUSION These results show that it is possible to quite accurately obtain AoA from RSSI measurements by using four antennas

0

1

2

3 4 Distance (m)

5

6

Fig. 17. Errors of estimated AoA from all indoor tests as a function of the distance between transmitter and receiver.

placed circularly on a plate. It is assumed that there are no moving obstacles in the line-of-sight between the transceivers, and that the transmitter and receiver lie within the same azimuth plane. Moving obstacles causes additional time dependent reflections of signals and may change total strength of signal in receivers during measuring cycle. The indoor absolute mean error was smaller than 6° and outdoor smaller than 4°. The error drastically increased with any increasing height between the transmitter and receiver device. This concludes that the indoor measurements are less accurate, which is a consequence of interference with many reflected signals. Distance versus error shows that the increased distances between transceivers had little influence on error.

R EFERENCES [1] I. F. Akyildiz, W. Su, Y. Sankarasubramaniam, and E. Cayirci, “Wireless sensor networks: A survey,” Comput. Netw., vol. 38, no. 4, pp. 393–422, 2002. [2] N. Patwari, A. O. Hero, M. Perkins, N. S. Correal, and R. J. O’Dea, “Relative location estimation in wireless sensor networks,” IEEE Trans. Signal Process., vol. 51, no. 8, pp. 2137–2148, Aug. 2003. [3] S. M. Lanzisera, “RF ranging for location awareness,” Ph.D. dissertation, Dept. Electr. Electron. Eng., Univ. California, Berkeley, 2009.


[4] M. Malajner, K. Benkic, P. Planinsic, and Z. Cucej, “The accuracy of propagation models for distance measurement between WSN nodes,” in Proc. 16th Int. Conf. Syst. Signals Image Process., 2009, pp. 1–4. [5] N. Patwari, “Locating the nodes: Cooperative localization in wireless sensor networks,” Signal Process. Mag., vol. 22, no. 4, pp. 54–69, 2005. [6] S. Maddio, A. Cidronali, and G. Manes, “An azimuth of arrival detector based on a compact complementary antenna system,” in Proc. Wireless Technol. Conf. Eur., 2010, pp. 1726–1729. [7] B. N. Bryan and P. Barooah, “Estimating DoA from radio-frequency RSSI measurements using an actuated reflector,” IEEE Sensors J., vol. 11, no. 2, pp. 413–417, Feb. 2011. [8] M. Abusultan, S. Harkness, B. J. LaMeres, and Y. Huang, “FPGA implementation of a Bartlett direction of arrival algorithm for a 5.8 Ghz circular antenna array,” in Proc. Aerosp. Conf., Big Sky, MT, 2010, pp. 2–10. [9] J. Graefenstein, A. Albert, P. Biber, and A. Schilling, “Wireless node localization based on RSSI using a rotating antenna on a mobile robot,” in Proc. Workshop Position. Navigat. Commun., 2009, pp. 253–259. [10] J. A. Jiang, C. L. Chuang, T. S. Lin, C. P. Chen, C. H. Hing, J. Y. Wang, C. W. Liu, and T. Y. Lai, “Collaborative localization in wireless sensor networks via pattern recognition in radio irregularity using omnidirectional antennas,” Sensors, vol. 10, no. 1, pp. 400–427, 2010. [11] R. D. Straw, The ARRL Antenna Book, 21st ed. Newington, CT: ARRL, 2007. [12] C. A. Balanis, Antenna Theory Analysis and Design, 3rd ed. New York: Wiley, 2005. [13] Microchip, MRF24J40MA Datasheet. (2010) [Online]. Available: http://ww1.microchip.com/downloads/en/DeviceDoc/70329b.pdf [14] N. Nasimuddin, Microstrip Antennas. Rijeka, Croatia: Intech Publishers, 2011. [15] S. N. Makarov, Antenna and EM Modeling with MATLAB. New York: Wiley, 2002. ˘ cej, “Wireless [16] M. Malajner, K. Benki˘c, P. Planin˘si˘c, D. Gleich, and Z. Cu˘ sensor networks module SPaRCMosquito v.2,” in Proc. 7th AIG, 2011, pp. 1–5.

1957

Marko Malajner received the B.Sc. and M.Sc. degrees in electrical engineering from the University of Maribor, Maribor, Slovenia, in 2006 and 2009, respectively. His current research interests include remote controls and wireless sensor network localization.

Peter Planinšiˇc received the B.Sc., M.Sc., and Ph.D. degrees in electrical engineering from the University of Maribor, Maribor, Slovenia, in 1979, 1991, and 2000, respectively. He has been with the University of Maribor since 1986, where he is an Associate Professor of electrical and computer engineering. His current research interests include digital signals and image processing and data compression.

Dušan Gleich received the B.Sc., M.Sc., and Ph.D. degrees in electrical engineering from the University of Maribor, Maribor, Slovenia, in 1997, 2000, and 2002, respectively. He is a Research Scientist with the Laboratory for Signal Processing and Remote Control, Faculty of Electrical Engineering and Computer Science, University of Maribor. He was a Visiting Scientist with German Aerospace Center, Cologne, Germany, from 2004 to 2005 and 2009. He has been an Associate Professor with the University of Maribor since 2010. His current research interests include image processing and data compression and extraction information form synthetic aperture radar images.