Sparse Channel Estimation in TDS-OFDM Systems

This chapter introduces a low-complexity sparse channel estimation (CE) scheme for time-domain synchronous (TDS)-OFDM systems to address the performance challenge in the presence of doubly selective fading channels. To tackle the issue, an overlap-add technique in the time-domain transmission symbol (TS) processing is adopted to coarsely estimate the channel length, path delays, and path gains. Even though the severe fading channel has a long delay spread, the robustness of the coarse CE can be improved by leveraging the temporal correlation of channels. Based on the coarse results, a priori-information aided iterative hard threshold (PA-IHT) algorithm is introduced to facilitate accurate channel estimation of doubly selective fading channels. Note that the considered method eliminates the convergence constraint and reduces the iteration numbers compared to the classical IHT algorithm, and has lower computational complexity and higher accuracy than the existing compressive sensing-based CE solutions. Simulation results verify the superiority of the considered methods for CE of TDS-OFDM systems over the existing schemes, particularly when dealing with severely doubly selective fading channels.