Abstract: Signal sampling is a vital component in modern information technology. As the signal bandwidth becomes wider, the sampling rate of analog-to-digital conversion (ADC) based on Shannon-Nyquist theorem is more and more high and may be beyond its capacity. However the analog to information converter (AIC) based on compressed sensing (CS) is designed to sample the analog signals at a sub-Nyquist sampling rate. A new multi-rate sub-Nyquist sampling (MSS) system was proposed in this article, it has one mixer, one integrator and several parallel ADCs with different sampling rates. Simulation shows the signals can be reconstructed in high probability even though the sampling rate is much lower than the Nyquist sampling rate.

Key words: CS , sub-Nyquist rate, AIC, MSS