Comparison of POA and TOA Based Ranging Behavior for RFID Application

Transcription

1 Comparison of POA and TOA Based Ranging Behavior for RFID Application Yongtao Ma, Member, IEEE, Kaveh Pahlavan, Fellow, IEEE, and Yishuang Geng, Student Member, IEEE School of Electronic Information Engineering, Tianjin University, Tianjin, 372 China Center for Wireless Information Network Studies, Worcester Polytechnic Institute, Worcester, MA 69 USA Abstract RFID can play great roles in future context-aware applications such as ambient intelligence and internet of things. Ranging technology will be one of interesting topics for RFID application. In the paper we compare the phase of arrival (POA) and time of arrival (TOA) based ranging behavior for RFID application. We characterize the ranging behaviors from the view of Cramer-Rao lower bounds (CRLB), ray tracing and empirical measurements. First, we analyze and compare the CRLB for POA and TOA based ranging in AWGN environment theoretically. Second, we use ray tracing method to model the distance and bandwidth influence on POA and TOA based RFID ranging in multipath environment. Third, we perform measurement to validate the noise and multipath influence on POA and TOA based RFID ranging. The parameters in ray tracing and measurements are also provided to further support the behavior comparison between TOA and POA ranging. The results show that in short range application such as RFID, POA based ranging has a comparatively better performance than TOA. It will benefit the future selection of POA or TOA technique, narrowband or wideband ranging method regarding specific scenarios for RFID application. Keywords Ranging, Phase of Arrival, Time of Arrival, RFID I. INTRODUCTION RFID can play an important role in context-aware applications. These contexts are based on the ubiquitous presence of objects carrying some intelligence that can communicate with the surrounding objects and environments in real time. RFID-based ranging will be a promising technique and become increasingly popular []. Compared with those short distance wireless positioning technology such as mobile cell, Wi-Fi, UWB etc., RFID ranging has significant advantages, because in short distance, RFID tag has the unparalleled advantages of very low price, easy deployment, energy efficiency, and etc. There are already some papers discussing the RFID position subject. These existing works offered overview of the most recent approaches for ranging and locating RFID tagged objects [2] [3] [4]. Locating and discovering objects is a challenging technique for RFID-based localization. Some articles present RFID-based localization and tracking technologies, including tag-based (e.g., LANDMARC), reader-based (e.g., reverse RFID), transceiver-free, and hybrid approaches. These technologies mainly use the available resource of RSS or TOA information to localize the target objects. A number of well-known approaches and their limitations are introduced [5] [6] [7]. Existing literature also indicates the challenges including multipath propagation, interference, and localizing multiple objects, etc [8]. Most of these challenges appeared not only for RFID-based localization, but also for other RFbased localization technologies. Phase of arrival is a feasible metric for ranging with RFID [9]. In indoor geo-location systems, scholars have already studied the receivers phase error for navigation performance analysis []. Some papers present continuous wave (CW) RF ranging technique for RFID. The proposed system concept relies on exact phase information. A linearized model of the tag s reflection coefficient has been proposed to bridge the gap between the nonlinear reality and the linear CW radar theory []. Estimation errors are derived and effects caused by noise and multipath propagation are analyzed in detail [2] [3] [4]. Certain scenarios such as robots application successfully achieved phase of arrival for ranging and tracking [5] [6]. It is also verified that the system performs well under lineof-sight conditions, and also can be possible in multipath or NLOS propagation environments [7]. Although there are many papers describing POA based ranging and positioning, one subject is not discussed until now. That is the comparison of POA and TOA based ranging behaviors for RFID in a certain bandwidth and distance. Below is the main work of this paper. First, we analyze phase of arrival based ranging principle. Second, we analyze the CRLB for two tones phase based ranging method, and compare the bounds of TOA based ranging. Third, we use ray tracing and measurement method to model the distance, bandwidth and multipath influence on POA and TOA based RFID ranging. We study the noise and multipath influence on RFID ranging in narrowband and wideband. Also our measurements results can be used to calibrate RT to get a deliberated channel model for narrowband and wideband RFID ranging. II. POA RANGING PRINCIPLE FOR RFID Fig. shows the single tone distance measurement. The phase difference between a transmitted and received tone can be used to estimate distance. A RFID reader transmits an uninterrupted tone. The tag locks a local oscillator to the incoming tone and retransmits it. The reader measures the phase delay of the received tone compared with the transmitted one and calculates the one way distance as: d = λ 2 ( θ + n) () 2π where λ is the tone wavelength, θ is the phase delay, and n is an integer. Since the range of the phase measurement is

2 Fig.. Single tone distance measurement TABLE I. PARAMETERS OF ISO/IEC 8-6C UHF RFID PROTOCOLS Frequency 86-96MHz, USA92-928MHz Channels Frequency-hopping [FHSS] Rates Kbps Reader Modulation DSB-ASK, SSB-ASK or PR-ASK to tag Tari Reference time interval: 6.25μs to25μs RF Tari pulsewidth Channels Frequency-hopping [FHSS] Tag to FM : 4-64 Kbps reader Rates Subcarrier Miller modulation: 5-32 Kbps Modulation DSB-ASK, SSB-ASK or PR-ASK between and 2π, it would be necessary to recognize the passed number of whole cycles n in order to determine the one-way distance greater than λ/2. The ambiguity can be eliminated by sending two tones and measuring the difference between their received delay phases. Substituting frequency for wavelength equation () we get: θ =2π( 2df c n) (2) θ 2 =2π( 2df 2 n) (3) c where c is the speed of light. Subtracting (2) from (3), we get d: d = c θ 2 θ = c 2π(f 2 f )τ = cτ (4) 4π f 2 f 4π f 2 f 2 The ambiguity in range is eliminated. The span of the measurement of θ is 2π, and the maximum value of d that can be measured using phase difference of two tones is conditioned on a maximum two tones frequency bandwidth. Table I shows the parameters of ISO/IEC 8-6C UHF RFID protocols. Fig.2 shows the distance ambiguity related with bandwidth using POA based ranging. For example, if interval of f 2 between f is 2MHz, the maximum measurable one-way distance is 7.5m. Let δθ equals to a given phase difference measurement error. The distance measurement error δd is: δd = c δθ (5) 4π Δf Fig. 2. Distance ambiguity related with bandwidth where Δf = f 2 f. It is clear from (5) that the error increases in inverse proportion to the frequency difference. As f 2 f is made smaller to accommodate longer range, the resolution or range error increases. III. CRLB COMPARISON OF POA AND TOA BASED RANGING IN AWGN CHANNEL According to equation (4), phase and time delay are equivalent if two tones values are defined. We may measure the time delay for POA-based ranging. Two tones spectrum for calculation of CRLB for POA-based ranging shows in Fig.3. The transmitted pulse is s(t), and the observed signal at the receiver is given by: r(t) =s(t τ)+η(t) (6) where τ is the time of flight of the signal between the transmitter and the receiver, and η(t) is the additive white Gaussian noise, with a spectral height of N /2 observed at the receiver. To form the probability density function of the observation given the value of the parameter τ, we should note that we are observing the entire pulse in a Gaussian noise with variance of σ 2, as if we were observing K points on the signal when K grows to infinity. The probability density function of Fig. 3. Two tones spectrum for calculation of CRLB for POA-based ranging

3 the observation would be: K f(r K /τ) = N π exp{ [r k s k (τ)] 2 } (7) 2 N /2 k= If we let K, f(r/τ) is not well defined. We can divide a likelihood function by anything that does not depend on and still have a likelihood function. We can divide by f(r K /H )= K N π exp{ [r k ] 2 2 N /2 } (8) k= before letting K, because this function does not depend on τ, it is legitimate to divide by it from here. f(r K τ) = f(r K τ) (9) f(r K H ) Substituting into this expression, cancelling common terms, letting K, and taking the logarithm we obtain ln f (r K τ) = 2 T r(t)s(t τ)dt T s 2 (t τ)dt N N () E[ 2 ln f (r K τ) τ 2 ]= 2 N {E T [r(t) s(t τ)] 2 s(t τ) τ 2 dt E T 2 s(t τ) τ 2 } () where we assume the derivatives exist. In the first term we observe that E[r(t) s(t τ)] = E[η(t)] = (2) The Fisher matrix for the time delay estimation is given by F τ = E[ 2 ln f (r K τ) τ 2 ]= 2 N T [ s(t τ) ] 2 dt (3) τ According to Parsevals theorem and Fourier transform characteristic, we can get F τ = ω 2 S(ω) 2 dω (4) πn Therefore the CRLB representing the variance of the estimate is given by CRLB = Fτ πn = ω2 S(ω) 2 dω σ d c 2 8π 2 (f 2 + W 2 4 )SNRWT (6) Compared with the CRLB of TOA based ranging [8]: σ d c 2 8π 2 (f 2 + W 2 2 )SNRWT (7) Simulation results from the two tones POA and TOA ranging CRLB equations is provided in Fig. 4 and Fig. 5. Supposed WT =, W is bandwidth of the transmitted signal, T is the duration of time for observing the signal. We can get the simulation results in Fig. (4). Fig.4 demonstrates the CRLBs of POA with two tones and TOA based ranging. It demonstrates the difference of CRLBs of POA with two tones and TOA based ranging. As the bandwidth increases, the POA with two tones is more accurate than TOA based ranging. Fig. 4. CRLBs of POA with two tones and TOA based ranging = f+ W 2 f W 2 πn (2π) 2 S 4 W [δ(f f + W 2 )+δ(f f W 2 )]2πdf = 8π 2 (f 2 + W 2 4 )SNRWT So, CRLB of two tones POA based ranging is (5) Fig. 5. CRLBs of POA with two tones and TOA based ranging Since T is the waveform interval which represents the signal waveform duration cycle, supposed T =/k, we can get the simulation results in Fig. (5). Fig.5 demonstrates the CRLBs of POA with two tones and TOA based ranging. It

4 demonstrates the difference of CRLBs of POA with two tones and TOA based ranging. As the bandwidth increases, the POA with two tones is more accurate than TOA based ranging. We can make a conclusion that in free space propagation, two tones POA based ranging is more accurate than TOA based ranging. For RFID ranging application, the transmitted signal frequency intervals are not so greater than UWB frequency bandwidth. In accordance with RFID application restrictions, POA based ranging is a better choice than TOA based ranging. IV. COMPARISON OF POA AND TOA BASED RANGING WITH RAY TRACING IN MULTIPATH ENVIRONMENTS (a) Fig. 6. One sketch scenario for ranging a tag with Ray Tracing Ray tracing is a good way to evaluate the performance of multipath impact on indoor localization. In order to compare POA and TOA based ranging, we use ray tracing method to study the ranging performance in one scenario in Fig.6. We suppose that there are four paths in the time channel profile, which are marked in Fig.6. The results of ray tracing simulation are shown in Fig.7. In order to explain the ranging behavior of TOA and POA methods more clearly, we draw the table II. Table II shows parameters in comparison of POA and TOA based ranging with Ray Tracing in multipath environment. (b) Fig. 7. (a)time channel profile with 4 paths for 2MHz bandwidth; (b)distance measurement using TOA and POA TABLE II. PARAMETERS IN COMPARISON OF POA AND TOA BASED RANGING WITH RAY TRACING IN MULTIPATH ENVIRONMENT Ranging Bandwidth No. of Meas. Meas. method (MHz) Freq. points Mean(m) Variance(m) TOA POA Obviously, for a comparatively low bandwidth, POA based ranging is more accurate than TOA. POA based ranging method is suitable for short distance position such as RFID. V. COMPARISON OF POA AND TOA BASED RANGING WITH MEASUREMENT IN MULTIPATH ENVIRONMENTS To further compare the performance of POA and TOA based ranging method in the appearance of multipath, we also performed the measurement in a typical indoor environment. We use Agilent network analyzer E8363B and a pair of UHF Fig. 8. One real measurement experiment scenario for ranging antennas to simulate the RFID ranging scenario. Fig.8 is the photograph of experiment of ranging a tag on a book on the shelf. Fig.9 is the comparison results of distance measurement. In Table III, we show the parameter details for comparison of POA and TOA based ranging with measurement in the real environment. Similar as Table II, we also draw the table III. Parameters are showed in table III for comparison of

5 Fig. 9. TABLE III. Distance measurement using POA and TOA COMPARISON OF POA AND TOA BASED RANGING WITH MEASUREMENT IN A REAL ENVIRONMENT Ranging Bandwidth No. of Meas. Meas. method (MHz) Freq. points Mean(m) Variance(m) TOA POA POA and TOA based ranging with measurement in multipath environment. VI. CONCLUSION In this paper we address the comparison of POA and TOA based ranging method for RFID application and characterize the ranging behavior from the view of CRLB, ray tracing and experimental results. We first analyze phase of arrival based ranging principle. And then we analyze the CRLB for two tones phase based ranging method, and compare it with the bounds of TOA based ranging. At last, we respectively implement ray tracing and measurement method to model the distance, bandwidth and multipath influence on POA and TOA based ranging. The result will benefits the future reference use of choosing ranging method for RIFD application. [5] N. A. Alsindi, B. Alavi, and K. Pahlavan, Measurement and modeling of ultrawideband toa-based ranging in indoor multipath environments, Wireless Communications, IEEE, vol. 8, no. 2, pp , 29. [6] J. D. Griffin and G. D. Durgin, Complete link budgets for backscatterradio and rfid systems, Antennas and Propagation Magazine, IEEE, vol. 5, no. 2, pp. 25, 29. [7] A. Lázaro, D. Girbau, and D. Salinas, Radio link budgets for uhf rfid on multipath environments, Antennas and Propagation, IEEE Transactions on, vol. 57, no. 4, pp , 29. [8] Y. Geng, J. He, H. Deng, and K. Pahlavan, Modeling the effect of human body on toa ranging for indoor human tracking with wrist mounted sensor, IEEE 6th Wireless Personal Multimedia Communications Symposium(WPMC), 23. [9] A. Bensky, Wireless positioning technologies and applications. Artech House, 27. [] I. F. Progri, An assessment of indoor geolocation systems, 23. [] D. Arnitz, K. Witrisal, and U. Muehlmann, Multifrequency continuouswave radar approach to ranging in passive uhf rfid, Microwave Theory and Techniques, IEEE Transactions on, vol. 57, no. 5, pp , 29. [2] E. DiGiampaolo and F. Martinelli, A passive uhf-rfid system for the localization of an indoor autonomous vehicle, Industrial Electronics, IEEE Transactions on, vol. 59, no., pp , 22. [3] P. V. Nikitin, R. Martinez, S. Ramamurthy, H. Leland, G. Spiess, and K. Rao, Phase based spatial identification of uhf rfidtags, inrfid, 2 IEEE International Conference on. IEEE, 2, pp [4] M. Scherhaufl, M. Pichler, E. Schimback, D. Muller, A. Ziroff, and A. Stelzer, Indoor localization of passive uhf rfid tags based on phase-of-arrival evaluation, Microwave Theory and Techniques, IEEE Transactions on, vol. 6, no. 2, pp , 23. [5] E. Di Giampaolo and F. Martinelli, Mobile robot localization using the phase of passive uhf-rfid signals, Industrial Electronics, IEEE Transactions on, vol. 6, no., pp , 24. [6] S. Sarkka, V. V. Viikari, M. Huusko, and K. Jaakkola, Phase-based uhf rfid tracking with nonlinear kalman filtering and smoothing, Sensors Journal, IEEE, vol. 2, no. 5, pp. 94 9, 22. [7] Y. Ma, L. Zhou, K. Liu, and J. Wang, Iterative phase reconstruction and weighted localization algorithm for indoor rfid-based localization in nlos environment, Sensors Journal, IEEE, vol. 4, no. 2, pp , 24. [8] K. Pahlavan and P. Krishnamurthy, Principles of Wireless Access and Localization. John Wiley & Sons, 23. ACKNOWLEDGMENT The authors wish to thank Mr. Yang Zheng, master student of CWINS in WPI, for his kind help in channel measurements and useful discussions. REFERENCES [] A. Costanzo, D. Masotti, T. Ussmueller, and R. Weigel, Tag, you re it: Ranging and finding via rfid technology, Microwave Magazine, IEEE, vol. 4, no. 5, pp , 23. [2] J. Heidrich, D. Brenk, J. Essel, S. Schwarzer, K. Seemann, G. Fischer, and R. Weigel, The roots, rules, and rise of rfid, Microwave Magazine, IEEE, vol., no. 3, pp , 2. [3] R. Miesen, R. Ebelt, F. Kirsch, and T. Schafer, Where is the tag? Microwave Magazine, IEEE, vol. 2, no. 7, pp , 2. [4] L. Ni, D. Zhang, and M. Souryal, Rfid-based localization and tracking technologies, Wireless Communications, IEEE, vol. 8, no. 2, pp. 45 5, 2.