A new threshold determination algorithm for SAW resonant sensors

摘要

Wireless SAW resonant sensors have been widely used in many applications, especially in harsh environment. The sensor is detected when the echo signal spectrum processed by Fourier transform exceeds a threshold. Because of complicated noise and electromagnetic interference in wireless channel, the threshold should be adaptive to interference in order to detect the sensor accurately. However, few researches have been reported on the threshold determination of SAW resonant sensors. The only reported minimum identification error probability method for SAW RFID-tags needs prior probability density functions of both the background noise and the signal. In addition, since the resonance frequency of sensor and the reading distance are unknown when the reading system interrogates a SAW resonant sensor, it is difficult to estimate the signal-to-noise ratio (SNR) in real-time before a threshold is determined. A false alarm means the detection of a sensor when it actually does not exist, and it will cause wrong frequency estimation in the reading system; so the false alarm is more unacceptable than the miss-detection. According to the constant false alarm rate (CFAR) method in radar signal processing, a threshold determination algorithm for SAW resonant sensors is proposed, which pursues the detection probability to reach a maximum in complicated noise, and only the probability density distribution of noise is needed to estimate in the algorithm.