Sparse Channel Estimation Algorithm Based On Compressive Sensing In OFDM System

As a novel sampling theory, Compressive Sensing is a major change in information acquisitionand can solve many technical problems encountered in the traditional signal processing system.With inherent sparsity characteristics of wireless multipath channel, channel estimation based onCompressive Sensing theory can greatly reduce the length of the required number of pilot ortraining sequence thus not only improving spectral efficiency, but also enhancing the estimationperformance. Sparse channel estimation algorithm based on Compressive Sensing in OFDM systemis analyzed in this paper. The main work for this paper is as follows:First, the classical reconstruction algorithm-Orthogonal Matching Pursuit is used infrequency-selective channel estimation. Do analysis about the pilot pattern’s influence oncompressed channel estimation, comparing it with the traditional LS channel estimation method.Modeling doubly selected sparse multipath channel system in delay-Doppler domain. In order Toovercome the instability caused by the doubly selective fading characteristics, RegularizedOrthogonal Matching Pursuit algorithm is adopted to estimate the doubly selective channelconsidering the instability of OMP algorithm. Theoretical analysis and simulation show that theROMP channel estimation has obvious improvement and upgrading than OMP.Using the correlation between time domain and frequency domain, an improvedtwo-dimensional compressed channel estimation in the time-frequency group sparse channel isdesigned. The new algorithm searchs by row and column at the same time so as to extract energyconcentration sampling point and Doppler frequency shift, and then restruct all the channelparameters. This method has a higher efficiency, better performance and lower complexity thanconventional compressive sensing algorithm.A new channel tracking method based on the difference transformation Compressive Sensing inslow fading channel is also proposed. By means of differential treatment with adjacent symbolsubcarriers channel pulse response, getting the change of intra channel. Differential treatment canreduce signal sparsity and thus achieve the purpose of reducing the complexity of the reconstructionalgorithm, significantly reduce the number of iterations, improve the channel estimationperformance.