Compressed Sensing Based Recursive Estimation of Doubly-Selective Channels for High-Mobility OFDM Systems

Abstract

For an orthogonal frequency-division multiplexing (OFDM) system in high-mobility applications, channel suffers from both frequency-selective and time-selective fading introduced by Doppler shift. Large pilot overhead is needed to estimate the numerous parameters of doubly-selective channel, resulting in low-spectral efficiency. In this paper, considering the correlation of practical wireless channels in high-dimensional signal spaces, a recursive channel estimation scheme based on compressed sensing (CS) is proposed for high-mobility OFDM systems to reduce the pilot overhead. Specifically, by exploiting the sparsity of OFDM channel in basis expansion model (BEM), the sparse BEM coefficients is estimated instead of numerous channel taps. Then we theoretically verify that the BEM coefficients corresponding successive OFDM symbols also share a common support. To utilize this temporal correlation of BEM coefficients, a recursive channel estimation algorithm derived from the classical modified CS algorithm is proposed to improve the present estimation by prior channel information from previous estimation. Theoretical and simulation results demonstrate that the proposed recursive channel estimation scheme preforms better than conventional schemes in various scenarios, even with less pilot overhead.