Sub-Nyquist sampling and parameters measurement based on dual-channel random demodulation for FRI signals

Recent developed finite rate of innovation (FRI) technology provides an efficient sub-Nyquist sampling and parameter measurements method for signals. However, due to the unique distinction of frequency spectrum, the existing FRI systems need to be designed according to the spectrum characteristics, which have poor universality. In this paper, we propose a sub-Nyquist sampling and parameter measurements system for FRI signals with non-ideal low-pass filter (LPF). Firstly in the main channel, a spread spectrum technology of random demodulation is used to distribute the frequency domain information over the entire spectrum. Then the spectrum information is obtained by filtering with non-ideal LPF and sampling with low rate analog-to-digital converter. To solve the problem of low reconstruction accuracy caused by the non-ideal effects of LPF, a dual-channel parallel measurement structure is proposed to obtain partial spectrum information of the basic signal. Finally, the random demodulation and filtering process are converted to a minimum L0 norm optimization problem, as well as a parameter estimation algorithm based on sparsity is proposed. We further design the hardware platform of the proposed system and confirm the validity through simulations and hardware experiments. The results demonstrate that the sampling method not only enhances the accuracy of parameters measurement, but also improves the flexibility of the sampling system.