Design and Implementation of a Passive UHF RFID-Based
`Real Time Location System
`
`Ting-wen Xiong, Jun-juan Liu, Yu-qing Yang, Xi Tan, Hao Min
`
`State Key Laboratory of ASIC & System, Fudan University
`
`NO.825 Zhangheng Road, Shanghai 201203, China
`
`Abstract—A passive Ultra-High Frequency
`(UHF) Radio
`Frequency Identification (RFID) based Real Time Locating System
`(RTLS) is presented in the paper. For the validation, a reader has
`been developed and complemented with a tag modulator to form a
`complete UHF RFID-based real time locating system focused on
`backscattering communication. PN sequences have been used to
`measure the TOA and estimate the distance between reader and tag.
`A TX-to-RX leakage canceller has been applied in the reader RF
`front-end to increase the isolation and enhance the sensitivity.
`Measurement results indicate that the estimation accuracy in real
`time is about 1.6 meters and the mean estimation error of 1000
`times’ measurement in 1 second is less than 0.5 meter.
`
`
`Index Terms—Radio Frequency Identification (RFID), Real
`Time Locating System
`(RTLS), Time of Arrival
`(TOA),
`backscattering
`
`U
`
`I. INTRODUCTION
`ltra-High Frequency (UHF) Radio Frequency Identification
`(RFID) is promising and increasingly interests many people
`due to the merits of long communication distance, high speed and
`large information capacity [1]. Depending on the existence of a
`built-in power source, tags in RFID system can be classified as
`passive, semi-active, and active. Active tags in RFID system
`contain built-in power sources for their proper operation. Passive
`tags, however, due to lack of built-in power sources, have to
`harvest power from the electromagnetic wave transmitted by the
`reader to supply their internal circuits. The communication
`mechanism of passive UHF RFID is based on backscattering, as
`shown in Fig.1.
`Real Time Locating System (RTLS) is a wireless system with
`the ability to locate the position of an item or a person anywhere
`in a defined space at a point in time that is, or is close to, real time
`[2].Many different RFID-based localization systems have been
`developed and applied to lots of areas in recent years, like
`emergency rescuing, patient tracking, asset or livestock tracking,
`and many other areas [3-5].
`According to the techniques of distance estimation, RTLS can
`be classified as: Time of Arrival (TOA), Angle of Arrival (AOA)
`or Direction of arrival (DOA), Received Signal Strength
`Indicator (RSSI) and so on. RTLS based on TOA technology
`detects transfer time of the signal to estimate the distance
`between the transceiver and the target. Some technologies of
`TOA-based RTLS have been reported. An indoor location
`system using RFID and ultrasonic sensors is realized in [6], but it
`needs additional sensors. Reference [7] introduces an active
`
`Fig. 1 Principle of UHF RFID
`
`RFID location system based on time difference measurement
`using a linear FM chirp tag signal, but a linear FM chirp source
`should be included in the active tags. Multi-frequencies
`continuous-wave (CW) RFID Radar systems have been
`researched and discussed in [8], including a linearized model of
`the tag’s reflection coefficient during backscatter modulation.
`In this paper a TOA-based RTLS using passive UHF RFID is
`proposed. The rest of the paper is organized as follows: Section II
`describes the read architecture. Guidelines for the design of the
`system are put forward in this section. Section III presents
`measurement setups of the system. A backscatter modulator is
`developed in this section to represent a passive tag. Section IV
`discusses the measurement results. And finally, the conclusions
`are drawn in the last section.
`
`II. SYSTEM ARCHITECTURE
`In order to locate a tag in the effective operation area, RFID
`reader need to detect the distance and here Pseudo-random
`Number Ranging (PNR) method is introduced. Block diagram of
`the reader and flow chart of the system are shown in Fig.2 and
`Fig.3, respectively. RFID reader communicates with a tag
`entering the interrogation area and judges whether it need to
`locate or not. If so, a Pseudo-random Number (PN) sequence is
`transmitted by the transmitter (TX block) and at the same time,
`the tag modulates the received signal with a square signal and
`backscatters it to the reader. The received PN sequence is
`demodulated by the reader receiver (RX block) and correlated
`with the transmitted one in Digital Signal Processing (DSP) to
`get the TOA. As a result, the proposed TOA-based real time
`locating system contains a versatile reader and, however, an
`ordinary tag, to locate an object.
`
`978-1-4244-5271-2/10/$26.00 ©2010 IEEE
`
`95
`
`RFC - Exhibit 1014
`
`

`
`
`
`Fig. 2. Block diagram of the reader
`
`Fig. 3. Flow chart of the system
`
`
`
`The system architecture of the reader is illustrated in Fig.4,
`including a direct conversion transmitter and a homodyne
`receiver. Only one local oscillator is used for both transmission
`and reception, which reduces the complexity of the reader and
`avoids the image frequency problems. However, several inherent
`constraints of the architecture should be considered.
` One of the main challenges is the above mentioned TX-to-RX
`leakage. Generally a circulator or a directional coupler is utilized
`to separate the transmission path from reception path. However,
`both of them are difficult to increase the isolation to 25 dB. The
`large leakage will cause desensitization and block the receiver. A
`narrowband TX-to-RX leakage canceller covering UHF RFID
`frequency band is proposed to improve the isolation, as shown in
`Fig.4. The signal transmitted by the reader is routed through a
`directional coupler and breaks into two paths. One path is fed to a
`circulator. The other passes through a phase shifter and an
`attenuator, both adjustable, to achieve the anti-leakage signal.
`The two paths are mixed in a balanced power combiner to cancel
`the leakage. The proposed TX-to-RX leakage canceller is
`supposed to increase the isolation by about 20 dB, which will
`greatly suppress the leakage and improve the receiver’s
`sensitivity [9].
`
`Fig. 4. System architecture of the reader
`
`UHF RFID system also suffers from the detection gaps since
`the phase in the backscattered wave varies along time and
`distance [10]. The leakage, as well as other strong reflections,
`will mask the received signal and destroy the coherent detection.
`Thus an in-phase/quadrature (I/Q) demodulator is introduced to
`avoid detection gaps. When one path is at minimum sensitivity,
`the other is at the maximum and the communication is always
`guaranteed. In addition, a calibration should be carried out to
`cancel the phase shift caused by the target tag to get a real TOA.
`Another constraint of the architecture is DC offset due to
`self-mixing or large reflections received at the antenna. An ac
`coupling stage is added in as the first element in the baseband
`chain to eliminate the offset. It avoids the signal from distortion
`due to saturation in baseband stages. However, it can’t eliminate
`the leaking PN sequence in this system and a variable gain
`amplifier (VGA) with a narrow gain range (0~20dB) is required.
`12 bits Analog-to-Digital Converters (ADCs) with a wide
`dynamic range are applied to avoid the baseband signal from
`being ruined by quantized noise. The anti-aliasing filter and the
`ADCs are designed to work up to 3 MHz baseband bandwidth.
`
`
`III. MEASUREMENT SETUP
`In order to validate the designed reader and the general system,
`a backscatter modulator has been designed and fabricated to
`represent the target tag, as shown in Fig.5. It contains a tag
`antenna, an RLC network, a switch, a frequency divider and a
`crystal oscillator. Although a power source is included in the
`modulator, the communication mechanism is still based on
`backscattering, just like a passive tag. The tag antenna was
`designed using Ansoft HFSS for antenna gain and impedance
`calculation, and then fabricated and measured by an Aglient
`network
`analyzer.
`The measured
`impedances
`are
`39.6+j160.3 : @900MHz, 39.9+j166.8 : @915MHz,
`40.8+j175.2 : @925MHz. The values for the inductance and
`capacitance in Fig.5 are 10 nH and 2.5 pF, respectively. The
`fabricated tag modulator switches the input impedances between
`two states according to the data signal. The switch’s on-off
`frequency varies among 40 kHz, 80 kHz and 160 kHz.
`
`96
`
`

`
`
`
`Fig. 5. Backscatter modulator
`
`Fig. 6. Measurement setup scheme
`
`A testing plan has been established and measurement setups
`have been carried out. The testing process can be divided into the
`following two steps. The first step is to check the function of the
`designed reader. The measurement setup scheme is illustrated in
`Fig.6. In this step, a continuous sine wave is generated and
`transmitted by the reader. Basic performances of a passive UHF
`RFID system are tested and analyzed, including the output power,
`isolation, reader sensitivity and etc.
`The second step is to validate the proposed real time locating
`system. To simplify the communication process, we start from
`the forth step of the chart flow in Fig.3, assuming that the reader
`has interrogated a tag which needs to be located. PN sequences
`are loaded into a Vector Signal Generator (Agilent E4438C) and
`sent to the RFID transmitter through I/Q channels. The received
`signal is down-converted by a I/Q demodulator and stored in
`Digital Storage Oscilloscope (Agilent DSO91204A). The length
`of the PN sequence (Golden code) is 31, while the symbol rate is
`2.48 MHz. However, the sampling frequency is 158.72 MHz,
`with an over-sampling ratio of 32. The demodulated data is then
`processed in Digital Signal Processing to calculate the TOA and
`to estimate the distance.
`
`IV. RESULTS AND DISCUSSION
`A. Basic performances
`
`The fabricated reader and tag are shown in Fig.7 and basic
`performances of passive UHF RFID system are measured and
`provided by Table I. Measurement results without leakage
`canceller are also presented in the table to compare with the final
`values. The TX-to-RX isolation has been increased from 25 dB
`to 42.5 dB within UHF RFID frequency band of 920-925 MHz.
`
`Fig. 7. Fabricated reader and tag
`
`TABLE I
`BASIC PERFORMANCES
`
`Without leakage
`canceller
`
`With leakage
`canceller
`
`18.2%
`78.6 Kbps
`25 dB
`29 dBm
`40.6 dB
`-52.2 dBm
`
`18.2%
`78.6 Kbps
`42.5 dB
`26 dBm
`46.2 dB
`-66.4 dBm
`
`Parameter
`
`AM modulation depth
`Data rate
`Isolation
`Output power
`Channel loss
`Reader sensitivity
`
`
`
`As a result, reader sensitivity has been improved from -52.2 dBm
`to -66.4 dBm. It should be noticed that reader sensitivity is not
`measured in the real communication channel because it will be
`degraded due to the effect of reflections and interferences. Thus,
`it is hard to precisely measure it in the real scenario. Instead, the
`reader and tag are directly connected by a variable attenuator.
`The measured output power of the reader with leakage canceller
`is 26 dBm. Considering a maximum BER of 1.0e-3, the
`maximum channel attenuation is 46.2 dB. Thus, reader
`sensitivity can be calculated by [10]:
`P
`P
`P
`G
`r
`
` (1)
`
`
`2
`sensitivity
`output
`channel
`reader
`where Greader represents the reader antenna gain and r represents
`the reflection coefficient of the target tag. The measurement
`sensitivity is worse than the theoretical value in that the AM
`modulation depth of the tag and the output power are less than
`the anticipated values.
`B. System verification
`
`The final measurements have been carried out in a real
`scenario to validate the proposed real time locating system. The
`environment where the measurements have taken place is a
`typical multi-path environment composed of tables, ceiling, floor
`and walls. Thus, multiple bad effects will take place during the
`measurements, which deteriorate the system performances and
`reduce the overall accuracy. Fortunately main of these effects
`have been considered and modeled in the system.
`
`97
`
`

`
`
`
`Fig. 8. Measurement results
`
`To get a correct measurement, the reader needs to know about
`delays through the antenna, the cables and the amplifiers. Any
`change in cable length is going to cause an error which needs to
`be corrected. Thus a calibration is required to eliminate the
`system error and the above mentioned phase shift caused by the
`tag modulator. The calibration steps are as follows. First, put a
`tag at the calibration distance (5 meters, for example) and
`measure the TOA under this setting. Second, calculate the
`inherent time delay in the system and record it. Practically, the
`calibration settings are saved on the EPROM in the processor and
`they will be valid for all further measurements until there is a
`change in cabling or antennas.
`Fig.8 illustrates received power at tag and estimation distances
`versus real distances. The theoretical values are also given to
`compare with these results. It is clear that the maximum read
`range is 6.3 meters, and that the corresponding power at tag is
`-11.2 dBm, which represents the system sensitivity since the
`reader sensitivity can’t be tested in a real scenario. This result is
`larger than that of the scenario without antenna (-16 dBm) due to
`the mismatch at the tag antenna.
`Fig.8 also indicates that the estimation accuracy in real time is
`about 1.6 meters and the mean estimation error of 1000 times’
`measurement is limited in 0.5 meter. The mean estimation error
`at 5 meters is nearly zero because it is the calibration distance.
`Since the total time of 1000 times’ measurement is less than 1
`second, mean values can be reported every 1 second, and the
`overall accuracy will be kept in 0.5 meter. The standard deviation
`of the estimation error is given by [11]:
`cT
`V'
`
` (2)
`
`1
`mT C N
`˜
`/
`4
`/m s
`where c is the light speed
`, CT is the symbol rate
`u
`8
`3 10
`of PN sequence, T is the average time, m is the over-sampling
`/C N is received Signal-to-Noise ratio.
`rate, and
`
`C
`
`0
`
`0
`
`V. CONCLUSIONS
`A TOA-based real time locating system using passive UHF
`RFID has been presented in the paper. A reader architecture
`based on a direct conversion transceiver and a tag modulator
`based on backscattering have been developed and implemented
`to make up of a complete passive UHF RFID-based real time
`locating system. PN sequences have been used to measure the
`TOA and estimate the distance between reader and tag. A
`TX-to-RX leakage canceller has been applied in the reader RF
`front-end to increase the isolation and enhance the sensitivity.
`Measurement setups have been built in a typical multi-path
`environment. Measurement results indicate that the maximum
`read range of the system is 6.3 meters, and that the estimation
`accuracy in real time is about 1.6 meters. However, the mean
`estimation error of 1000 times’ measurement in 1 second is less
`than 0.5 meter. Thus, the distance of the tag can be derived with
`high accuracy in time that is close to real time.
`As a result, the proposed passive UHF RFID-based RTLS has
`proved to range all the tags in its effective operation area in
`nearly real time, and it will locate and track the tags if Direction
`of Arrival (DOA) should be precisely detected. The latter will be
`future research work.
`
`[4]
`
`REFERENCES
`in
`[1] Finkenzeller, RFID Handbook: Fundamentals and Applications
`Contactless Smart Cards and Identification, 2nd edition. New York: Wiley,
`2003.
`[2] Doo-Jin Park, Young-Bok Choi, Ki-Chan Nam, “RFID-Based RTLS for
`Improvement of Operation System in Container Terminals,” IEEE
`Asia-Pacific Conference on Communications, Aug. 2006, pp. 1-5.
`[3] Woochul Jeon, Melngailis, J., Newcomb R.W., “Disposable CMOS passive
`RFID transponder for patient monitoring," IEEE International Symposium
`on Circuits and Systems, 2006, pp. 5071- 5074.
`Inaba, T., “Value of Sparse RFID Traceability Information in Asset
`Tracking during Migration Period,” IEEE International Conference on
`RFID, Apr. 2008, pp. 183–190.
`[5] Pandey S., Singh D., “Efficient and Cost Effective Tracker in Monitored
`Area Network Using RFID,” IEEE International Conference on Wireless
`and Optical Communications Network, Apr. 2006, pp. 1-5.
`[6] Dai Hepeng, Su Donglin,”Indoor location system using RFID and
`ultrasonic sensors,” IEEE International Symposium on Antennas,
`Propagation and EM Theory, Nov. 2008, pp. 1179 – 1181.
`[7] Yanchuan Huang, Brennan P.V., Seeds A., “Active RFID location system
`based on time-difference measurement using a linear FM chirp tag signal,”
`IEEE 19th International Symposium on Personal, Indoor and Mobile Radio
`Communications, Sep. 2008, pp. 1-5.
`[8] Arnitz D., Witrisal K., Muehlmann U., “Multifrequency Continuous-Wave
`Radar Approach to Ranging in Passive UHF RFID,” IEEE Transactions on
`Microwave Theory and Techniques, vol. 57, no. 2, pp. 1398 – 1405, May
`2009.
`Jin-WOO Jung, Hyoung-Hwan Roh, and Jong-Cheol Kim, “TX Leakage
`cancellation via a Micro Controller and High TX-to-RX Isolations Covering
`an UHF RFID Frequency Band of 908-914 MHz,” IEEE Microwave and
`Wireless Components Letters, vol. 18, no.10, Oct.2008.
`[10] Iker Mayordomo, Roc Berenguer, and Andres Garcia-Alonso, “Design and
`Implementation of a Long-Range RFID Reader for Passive Transponder,”
`IEEE Transactions on Microwave Theory and Techniques, vol. 57, no. 5,
`May 2009.
`[11] Pratap Misra and Per Enge, Global Positioning System, Signal,
`Measurements, and Performance, 2nd edition. Ganga-Jamuna Press, 2008.
`
`[9]
`
`
`
`98