Crosstalk and SNR Measurements using a Multi-Antenna RFID Reader with Active Carrier Compensation Robert Langwieser * , Christoph Angerer * , Arpad L. Scholtz, and Markus Rupp Christian Doppler Laboratory for Wireless Technologies for Sustainable Mobility * Institute of Communications and Radio-Frequency Engineering Vienna University of Technology, Austria {rlang, cangerer, ascholtz, mrupp}@nt.tuwien.ac.at Abstract - In this paper we show the dependence of the Signal to Self Interference Ratio (SSIR) and of the Signal to Noise Ratio (SNR) on the transmit power level of the reader and the transpon- der position. We describe the conducted measurements with our multi-antenna Radio Frequency Identification (RFID) research and development environment and compare passive transmitter- receiver isolation due to spatially separated antennas with addi- tional active crosstalk compensation. With the active compensa- tion we have achieved an SSIR improvement of more than 55dB. In scenarios with active carrier compensation the evaluation of the interference to noise ratio at the receivers at different transponder positions shows that the compensated crosstalk is at least com- posed of two different parts. One part is the direct crosstalk from the reader transmitter antenna to the receiver antennas and a sec- ond part of the interference is reflected by the transponder and superposed at the receive antennas. I. I NTRODUCTION In the Ultra High Frequency (UHF) domain communication of pas- sive or semi-active Radio Frequency Identification (RFID) systems is based on the technique of backscatter modulation [1]. This means that a part of a transmitted Continuous Wave (CW) Radio Frequency (RF) signal is reflected by a target to a receiver. In RFID systems trans- mitter and receiver are combined into one device called interrogator or reader and the backscattering device is called transponder or tag. The backscattering requires that the transmitter and the receiver of the reader operate at the same frequency at the same time. Therefore, the transmitter can not be switched off during reception. Furthermore, the transmit signal and the receive signal can not be separated by conven- tional filtering. This leads to a strong self interference at the receiver during the whole communication process. In the case of passive RFID the transponders are additionally powered by the CW signal of the reader. The ratio of the received transponder response to the self interfer- ence, Signal to Self Interference Ratio (SSIR), can be -60 dB [2] and worse depending on the antenna configuration and the distance of the transponder. To reduce the linearity requirements of the receiver and to avoid large DC offsets at the receiver the isolation between trans- mitter and receiver has to be improved either with passive or active techniques [3]- [8]. In this contribution we discuss the system perfor- mance utilizing an experimental multi-antenna reader and increased isolation based on separated transmit and receive antennas and further- more with active carrier compensation based on vector modulators [9]. The approach of two receivers allows to measure at two spatial sep- arated receive positions at the same time. Moreover, multi-antenna scenarios like beam forming or diversity techniques are discussed for RFID [10], [11]. Furthermore we look into the influence of the trans- mit power level onto the SSIR and the Signal to Noise Ratio (SNR). In Section II. we describe our measurement environment and the measurement procedure and in Section III. we present and discuss our results. Finally, in Section IV. we conclude our paper. II. MEASUREMENT ENVIRONMENT We performed the measurements presented in this paper with our multi-antenna RFID research environment [12] [13]. This environ- ment allows for a very flexible reader configuration with up to two transmitters and two receivers. The experimental reader can be di- vided into three building blocks: the digital baseband [14], the RF frontends, and the antennas. A system overview is illustrated in Fig- ure 1. The protocol issues were processed in a DSP of the digital TI DSP (TMS320 C6416) Protocol Stack Xilinx Virtex II FPGA Signal Processing DAC ADC Digital Baseband UHF-TX1 UHF-RX1 UHF-RX2 RF Frontends Antennas DAC ADC BLF SNR, SSIR INR FIGURE 1-MEASUREMENT SETUP OVERVIEW baseband unit and the transmit sequence for powering the transpon- der and to communicate with the transponder is generated by a Field Programable Gate Array (FPGA). Via a Digital-to-Analog Converter (DAC) the analog UHF-transmitter is connected. Finally, the anten- nas are placed in a separate room and are connected with cables to the transmitter and the receivers. Two analog UHF-receivers are con- nected to the two Analog-to-Digital Converters (ADCs) of the digital baseband. The samples are directly transferred and stored to a personal computer (PC). The data processing is performed later on off-line in Matlab. We have measured in the UHF band at a frequency of 866MHz with a commercially available passive transponder for all conducted measurements. In Figure 2 the RF frontend configuration with crosstalk compensation is illustrated. Transmitter and receivers per- form as linear frequency transponders. This transponders are con- nected to the digital baseband at 13.33 MHz and perform the frequency up-conversion to the desired UHF band or the down-conversion re- spectively. At the UHF-transmitter the output power level can be ad- justed continuously over a range of 55 dB. Additionally, two crosstalk or Carrier Compensation Units (CCU1 and CCU2) are part of the fron- tend setup. A fraction of the transmit signal is distributed to the two vector modulators of the CCUs. The outputs of the CCUs are added via the directional couplers into the receivers RX1 and RX2 and used to compensate the crosstalk from antenna TXA1 to the antennas RXA1 and RXA2 at the receivers. The two vector modulators allow to adjust separately the compensation signals for the two receivers. For perfect 66 The Third International EURASIP Workshop on RFID Technology