Frequency domain sensing system using random modulation pre-integrator

A major challenge in wideband analog signal processing is the requirement for very high sampling rate. The emerging compressed sensing (CS) theory makes processing wideband signal at its information rate possible if the signal has a sparse representation in a certain space. This study introduces a frequency domain sensing system based on CS. First, the proposed system employs a random demodulator to sensing the signal in frequency domain and an integrator to compress the signal. Second, the incoherence between the proposed measurement matrix and the sparse representation basis is proved mathematically. Different from the conventional analog-to-information conversion, the integrator of the system does not need to be reset after each acquisition. Experimental results indicate that, for spectrally sparse signal, the system is able to recover analog waveform at a high equivalent sampling rate from a small number of low-speed samples.