The phase noise of the local oscillator used in a passive UHF RFID system can prove to be a key limiting factor on the ultimate interrogation range of the system.

Radio-frequency-identification (RFID) systems have become widespread as reliable means of storing and remotely retrieving data through the use of compact RFID tags. In particular, the use of ultrahigh- frequency (UHF) passive RFID is appealing for many applications since it enables recognition from a reasonable distance. The technology is ideal for supply-chain management and several major firms, such as Wal-Mart and Tesco, are planning to mandate the use of UHF RFID in their supply chains.^{1,2}

The performance of a UHF RFID system is usually characterized by its interrogation range, which is defined as the maximum distance at which an RFID reader can recognize a tag. This can be divided into two categories: the forward-link interrogation range (FIR) and the reverse-link interrogation range (RIR). In UHF RFID systems, the forward-link refers to the communication link from a reader to a tag, whereas the reverse-link is that from the tag to the reader. FIR is defined as the maximum distance at which the tag receives such power as to turn on and back-scatter, and RIR is the maximum distance at which the reader can decode the data of the tag satisfying a SNR requirement. Since the actual interrogation range is determined by the smaller value of FIR and RIR, both values should be considered simultaneously when deploying UHF RFID systems.^{3-6}

*Figure 1* shows UHF RFID link concepts compared to a typical wireless communication system such as code-division- multiple-access (CDMA) or Global System for Mobile Communications (GSM) cellular systems. In a typical wireless communication system, the forward link refers to the communications link from a base station (BS) to a mobile station (MS), whereas the reverse link is that from the MS to the BS. The noise levels at both the links are determined by thermal noise power, which is

*P _{N, thermal} = 4kTB (1)*

where

k = Boltzman's constant,

T = the absolute temperature (in K), and

B = the bandwidth.

Generally, wireless communication systems exhibit link balance between the forward and reverse links, where the dynamic range between two links is almost the same. Therefore, the forward-link coverage corresponds closely with the reverse-link coverage, although the transmit powers for the forward and reverse links may differ.

In contrast, the communications link of a passive UHF RFID system has an imbalance between the forward and reverse links (*Fig. 1*). This is because the RFID tag has no internal power supply and must harvest energy from a continuous-wave (CW) signal transmitted by the RFID reader. Consequently, the FIR is mainly dependent on the threshold power necessary to power up the tag. Another major difference is that the phase noise of the transmitter (Tx) leakage at the reader's circulator has greater influence on system noise than the thermal noise at the reader's receiver (Rx). Therefore, there is the possibility that the RIR has a smaller value than FIR especially for poorly designed stationary readers or handheld readers.

In this examination of UHF RFID systems, it will be assumed that the RFID reader's antenna employs polarization that matches that of the tag's antenna. If r is used to denote the operational distance between an RFID tag and the reader operating in free space, then the power received by the RFID tag, P_{Rx}, can be found by applying the Friis electromagnetic (EM) wave propagation equation:

where

λ = the wavelength in free space,

P_{Tx} = the signal power feeding into the reader antenna by the transmitter,

G_{R} = the gain of the reader antenna,

G_{T} = the gain of the tag antenna,

One portion of the power P_{Rx}is absorbed by the tag for DC power generation and the other portion of P_{Rx}is backscattered for the reverse link. In order to ensure correct operation of the tag, the absorption power must be larger than the minimum operating power required for tag operation, P_{TH}} In the case of a tag with amplitude-shift-keying (ASK) modulation, the time-averaged absorption power of the tag is given by7:

where

m = the modulation depth.

Generally, P_{TH} is determined according to the tag chip design and antenna matching conditions. The FIR, T_{forward}, can then be derived using Eq. 4:

The FIR calculated by Eq. 4 has a value of 8 m as shown in *Fig 2*. In Eq. 4, the FIR is proportional to the square root of the transmitted effective isotropic radiated power (EIRP) P_{Tx}G{T, and the tag antenna's gain, GR, and is inversely proportional to the square root of the tag's power threshold level, P_{TH}. From experience, it is known that the RF threshold power level required to turn on a tag ranges from 10 } W (-20 dBm) to 50 W (-13 dBm).^{8} The modulation depth, m, is chosen to be an average value between 0.1 and 0.9.

In the reverse link, the backscattered signal from a tag should be strong enough that the reader's demodulation output signal will meet the system's signal-to-noise-ratio (SNR) requirement. To calculate the SNR of the demodulation output signal of the reader, consider the conventional reader architecture of *Fig. 3*. The RFID reader is composed of an LO, transmitter, receiver, and antenna with a circulator. The circulator is a nonreciprocal three-port device, where the signals travel from the transmitter port to the antenna port or from the antenna port to the receive port. In practice, the circulator cannot totally isolate the transmitter from the receiver due to the inherent leakage between its ports. Generally, the Tx/Rx isolation ranges from 20 to 50 dB.10 Therefore, the phase noise of the Tx leakage power is much stronger than the thermal noise, to a degree that the RIR mainly depends on the Tx/Rx isolation level. On the other hand, in a typical wireless communication system, the Tx leakage is normally not a major problem because duplexing techniques such as frequency division duplexing (FDD) and time division duplexing (TDD) are applied.

As shown in *Fig 3*, the LO provides two identical frequency signals: one for the transmitter and the other for the receiver. Neglecting the amplitude noise, the LO signal can be expressed as

where

A_{LO} = the amplitude of the LO signal,

ω = the angular frequency, and

θ _{LO}(t) = the phase noise of the LO signal.

The RFID system's power amplifier (PA) boosts the level of the LO signal. This amplified signal feeds the reader antenna via the circulator and then is radiated into free space. Simultaneously, the reader antenna receives backscattered signals from the tag.

Continue to page 2

### Page Title

As can be seen in *Fig 3*, the circulator cannot totally isolate the transmitter from the receiver due to the inherent leakage between its ports. The TX leakage signal is delayed by and given as follows:

where

A_{U} = (2ηR_{0}P_{Tx})^{0.5} = the signal amplitude,

η = the Tx/Rx isolation level,

R_{0} = the input resistance of the RFID reader's receiver, and

Δt = the round-trip delay between the Tx leakage and the LO signal.

It is important to recognize that θ_{LO}(t) in Eqs. 5 and 6 is related with the phase noise from the same LO source, except for the time delay. It is also assumed that the baseband bandpass filter (BPF) used in the RFID reader's receiver has sharp frequency selectivity. To simplify the forthcoming analysis, the impulse response of the BPF is characterized approximately with an ideal rectangular transfer function and its low-end frequency and high-end cut-off frequency are denoted by f_{L} and f_{H}, respectively. With a quadrature receiver, the Tx leakage signal and LO signal are mixed and the output is low-pass filtered. The resulting phase noise of the Tx leakage signal can then be given by

where

G_{YX} = the transfer coefficient of the receiver that takes into account the total gain of the RF circuitry and φ{} = the operator for calculating the power spectral density (PSD) of a random process, such as the Fourier transform of the auto-correlation function of a random process.

If the RFID system simply transmits a UHF CW signal, then level of P_{N}, pn varies dramatically with time delay.^{11} In the worst case, where ωΔt takes on a value of

If the same LO is used for transmission and reception, the phase noise of the received signal is correlated with the LO, where the level of correlation depends on the time difference between the two signals. If the time difference is short, the corresponding effect greatly abbreviates the phase noise spectrum at baseband. In radar applications, such as RFID, this phase-noise-reducing effect is called range correlation12 According to ref. 12, the baseband PSD at the offset frequency Δf_{c} with a round-trip delay of Δt can be found (from ref. 12) by referring to Eq. 10 on p. 88.

*Figure 4* shows an example of a typical PSD for the LO itself and the phase noise reduction effects due to the range correlation with a round-trip delay of 1 m. The typical PSD of the LO is selected by considering state-of-the-art UHF RFID LO performance. The effect of the range correlation on the phase noise for different offset frequencies was estimated using Eq. 10. For example, with an offset frequency of 10 Hz, the phase noise is reduced by 130 dB. The phase noise reduction is proportional to the square of the round-trip delay, r, and the square of the offset frequency, Δf_{c}. Due to the short round-trip delay (less than 1 m) between the Tx leakage signal and the LO signal, the phase noise effects are dramatically reduced.

Reduction of phase noise by range correlation also depends on the filter bandwidth. In the case of a 160-kb/s data rate, the measured phase-noise reduction values are shown in ref. 6. The measured value is 41 dB and it is almost same with the result reported here. As shown in *Fig. 5*, the phase noise of the Tx leakage signal is at a higher level than the thermal noise, to a degree that the RIR mainly depends on the Tx/Rx isolation level and it is verified that the authors' assumption that the phase noise of the Tx leakage is the dominant factor in determining the RIR. For a closed-loop phase locked loop (PLL), the PSD of the phase noise is filtered by the transfer function of the PLL, and the phase-noise effects become much smaller. Thus, the result for the open-loop voltage-controlled oscillator (VCO) represents the worst case.

The RIR equation can be derived using Eq. 9. The backscattered signal, X_{M}(t) in the ASK case can be expressed as

where

A_{M} = the backscattered signal's amplitude,

s(t) = the RFID tag's binary data sequence of 0's and 1's, and

2r/c = the roundtrip delay between the RFID reader and the RFID tag. Again using the Friis EM wave propagation equation, A_{M} can be found by

With a quadrature receiver, the received signals (Eq. 6) and (Eq. 11) and LO signals (Eq. 5) are mixed and the output is low-pass filtered. The resulting baseband signals will be

respectively. Generally, the Tx leakage magnitude A_{U} is larger than the level of the backscattered signal, A_{M}, and it is possible to neglect the phase-noise contribution of the received signal. As a result, the minimum achievable SNR can be approximately expressed by Eq. 15 (above), which gives the lower boundary of the SNR value of the RFID reader. Substituting Eq. 12 into Eq. 15, in the case of ASK, yields the RIR as

Continue to page 3

### Page Title

Unlike the FIR, the RIR depends on various parameters such as reader antenna gain, circulator performance, and LO phase noise. Figure 6 shows RIR as a function of Tx/ Rx isolation level. The poor isolation of the circulator degrades not only the RFID reader's sensitivity but also saturates the receiver front-end circuitry. Thus, poor isolation can be considered the main cause of RIR reduction. In other words, a large isolation coefficient is recommended for enhancing RIR. An easy way to enhance isolation is to employ two separated antennas for Tx and Rx. However, the RFID reader size and cost will be increased as a result. A circulator of ferrite material or an active CMOS circulator may also mitigate this problem, but the increase in reader cost and poor isolation of the circulators are major obstacles.

*Figure 6* also shows the dependence of the reader antenna gain on the interrogation range. Under the same EIRP, different antenna gains of 13 and 4 dBi are used in the RIR calculation. The transmit power of the RFID reader, P_{Tx}, in the case of an antenna with 13-dBi gain, is lower than that for an antenna with 4-dBi gain. Because increasing the transmitter power also induces a Tx leakage power increment, the case of 13-dBi antenna gain is more effective than the case of 4-dBi antenna gain with respect to increasing RIR. The authors think that these results are useful in determining UHF RFID reader specifications such as Tx/Rx isolation and the phase noise of VCO.

When deploying a UHF RFID system, the interrogation range of the RFID reader is a key design parameter. In order to maintain good performance and stable operation of the RFID system, the FIR and RIR should first be well understood, and the interrogation range equations derived here can help. in this article. Their effectiveness was shown via numerical results of FIR and RIR. In particular, the authors analyzed the relationship between the Tx leakage power and the phase noise of the LO in determining the RIR performance. From the results shown here, it can be concluded that the transmit power of a RFID reader is the dominant factor in determining FIR and the reader antenna gain, the LO phase noise, and the Tx/Rx isolation are the dominant factors in determining RIR. These results should serve as a useful reference in designing and deploying RFID systems.

REFERENCES

1. Klaus Finkenzeller, RFID Handbook, 2nd ed., Wiley, New York, 2003.

2. Daniel M. Dobkin, The RF in RFID-Passive UHF RFID in Practice, Elesvier, New York, 2007.

3. Udo Karthaus and Martin Fischer, "Fully Integrated Passive UHF RFID Transponder IC with 16.7-uW Minimum RF Input Power," IEEE Journal of Solid-State Circuits, vol. 38, No. 10, 2003, pp. 1602-1608.

4. Curty J.-P. et al., "Remotely powered addressable UHF RFID integrated system," IEEE Journal of Solid-State Circuits, vol. 40, No. 11, 2005, pp. 2193-2202.

5. D. Y. Kim, B. J. Jang, H. G. Yoon, J. S. Park, and J. G. Yook, "Effects of reader interference on the RFID interrogation range," in the 2007 European Microwave Conference (EuMC'07), October 2007, pp. 728-731.

6. Z. G. Fan, S. Qiao, J. T. Huangfu, and L. X. Ran, d"Signal descriptions and formulations for long-range UHF RFID readers," Progress in Electromagnetics Research, PIER 71, 2007, pp. 109-127.

7. G. De Vita and G. Iannaccone, "Design criteria for the RF section of UHF and microwave passive RFID transponders," Transactions on Microwave Theory & Techniques, vol. 53, No. 9, 2005, pp. 2978-2990.

8. Udo Karthaus, "Fully Integrated Passive UHF RFID Transponder IC with 16.7-uW Minimum RF Input Power," IEEE Journal of Solid-State Circuits, vol. 38, No. 10, 2003, pp. 1602-1608.

9. EPCglobal, Inc., "EPCTM radio-frequency identity protocols Class-1 Generation-2 UHF RFID protocol for communications at 860-960MHz," ver. 1.0.9, Jan. 2005.

10. Wan-Kyu Kim, Moon-Que Lee, Jin-Hyun Kim, Hyung-sun Lim, Jong-Won Yu, Byung- Jun Jang, Jun-seok Park, "A Passive Circulator for RFID application with High Isolation using a Directional Coupler," 2006 European Microwave Conference, September 2006, pp. 196-199.

11. J. H. Bae, J. C. Kim, B. W. Jeon, J. W. Jung, J. S. Park, B. J. Jang, H. R. Oh, Y. J. Moon, and Y. R. Seong, "Analysis of Phase Noise Requirements on Local Oscillator for RFID System Considering Range Correlation," in the 2007 European Conference on Wireless Technology(EcWT'07), October 2007, pp. 385-388.

12. A. D. Droitcour, O. Boric-Lubecke, V. M. Lubecke, J. Lin, and G. T. A Kovac, "Range Correlation and I/Q Performance Benefits in Single-Chip Silicon Doppler Radars for Noncontact Cardiopulmonary Monitoring," IEEE Transactions on Microwave Theory & Techniques, vol. 52, 2004, pp. 838-848.