Research On Digital Signal Processing Of Passive Wireless SAW Sensors

Wireless Passive SAW sensors can not only measure various physical and chemical quantities, but also hold such excellence as resisting interference, durability, bearing hostile environment, minisize, passive wireless measurement etc. So people are interested in it. Signal processing is one of the kernels of this kind of sensors and the main aspect of effectiveness. Digital signal processing is studied and used to extract the characteristic frequency of the measurand from the response of wireless passive sensors using SAW resonators, that is the echo to RF interrogation signal. At first, the response of the sensors is analyzed, and its characteristic and primary components are discussed. It illuminates two methods of direct and indirect measurement for resonant frequency and the tasks of signal processing corresponding to them, for measuring the characteristic quantity varying with the measurand varied. An algorithm of signal detection combinedly based on time domain and frequency domain is designed. While the signal is detected, the algorithm of signal processing varying with the characteristic quantity to be extracted is different. Algorithms such as notch filtering, digital phase detection, denoising with hard threshold based on multiresolution analysis, FFT, and rectifying frequency etc., are used to evaluate the primary frequency of modulation signal, which shows the difference between the carrier frequency of exciting signal and resonant frequency, realizing the indirect measurement for resonant frequency. While preprocessing, notch filter is used to eliminate the exciting frequency interference. Digital phase detection (DPSD) is used to extract the echo envelope. The relation between the primary frequency of the output by digital phase detection and that of modulation signal from the echo is derived. The components of 'noises' in the decimation output are analyzed. Denoising with hard threshold based on multiresolution analysis is used to decrease or eliminate these 'noises'. Finally, the primary frequency of the modulation signal is evaluated by using FFT and the algorithm of rectifying frequency from the denoised output. These algorithms being implemented, the requirements and effects for using them are analyzed. And in order to evaluate the primary frequency of the modulation signal, some arguments are selected and the algorithms are designed corresponding to the sensingsignal features. Finally, the effects on the primary frequencies of the modulation signal by the above algorithms are analyzed. The processed results show that the algorithms can evaluate the primary frequency of modulation signal from the echo with higher signal-noise ratio. According to the principle of frequency measurement in an electronic counter, the frequency measurement by a software method is used to measure the carrier frequency of the echo, realizing direct measurement of the resonant frequency. The frequency resolution for this method is analyzed, and the interpolation by software method is used to improve the frequency resolution and frequency precision. A virtual instrument implementing this measurement is developed in LabVIEW6i. Comparing the algorithm to evaluate the primary frequency of the modulation signal with the algorithm to measure the carrier frequency of the echo, the direct measurement is simpler and more practical, while the indirect measurement is complex, and computed with a great deal of time. Consequently the indirect measurement is difficult to be implemented in realtime.