Research On Time-Domain Sparse Channel Estimation Methods Of OFDM System Based On Compressed Sensing

Compressed Sensing appeared in recent years is different from Nyquist’s samplingtheory, and it can reconstruct sparse signal from very few samples with high probability.It has been introduced in many fields gradually. In this paper, sparse channel estimationmethod of pilot-aided OFDM system is researched based on Compressed Sensing.The paper introduces the basic principles of OFDM system and traditional channelestimation theory. Then it studies the three important aspects of CS: the indication ofsparse signal, the design of observation matrix and the algorithm about signalreconstruction. The focuses are MP, OMP, StOMP, BP algorithms and reconstruction ofsparse signal based on DFT. And a method of time-domain sparse channel estimation ofOFDM system on the basis of four algorithms mentioned is realized. Finally, it conductsfour groups of simulation experiment, respectively for performance of channelestimation, the least numbers of pilot frequency, the inserted modes of active pilot andthe influence on performance by the priori sparsity of channel, when OMP, StOMP, BPalgorithms are applied to the OFDM system. Results show that the Compressed Sensingapplied to the OFDM systems time-domain sparse channel estimation can greatly savethe frequency, and improve spectrum utilization.