The Study for Doubly Selective Channel Estimation Algorithm Based on Cluster Characteristics

Abstract

According to the wideband channel with the sparse characteristics of time frequency doubly selective fading, and the sparse energy mainly concentrating on a few clusters, this paper presents the compressed sensing algorithm, that is cluster-regularized adaptive matching pursuit (CRAMP) based on cluster characteristics. Through analyzing the time frequency doubly selective channel in orthogonal frequency division multiplexing (OFDM) system, and using the time domain and frequency domain correlation in channel, the algorithm uses the 2D-DFT as the observation matrix in compressed sensing channel. And then utilizing clustering characteristics, it estimates for each cluster rather than estimates for each element in channel. The simulation results show that the CRAMP algorithm in channel reconstruction accuracy and bit error ratio (BER) respects are superior to conventional compressed sensing ones.