Model-Based Compressive Channel Estimation over Rapidly Time-Varying Channels in OFDM Systems

Yi LIU  Wenbo MEI  Huiqian DU  

Publication
IEICE TRANSACTIONS on Communications   Vol.E97-B   No.8   pp.1709-1716
Publication Date: 2014/08/01
Online ISSN: 1745-1345
DOI: 10.1587/transcom.E97.B.1709
Type of Manuscript: PAPER
Category: Wireless Communication Technologies
Keyword: 
model-based compressive sensing,  channel estimation,  OFDM,  block-sparse model,  rapidly time-varying channel,  

Full Text: PDF(1006.6KB)>>
Buy this Article




Summary: 
By exploiting the inherent sparsity of wireless propagation channels, the theory of compressive sensing (CS) provides us with novel technologies to estimate the channel state information (CSI) that require considerably fewer samples than traditional pilot-aided estimation methods. In this paper, we describe the block-sparse structure of the fast time-varying channel and apply the model-based CS (MCS) for channel estimation in orthogonal frequency division multiplexing (OFDM) systems. By exploiting the structured sparsity, the proposed MCS-based method can further compress the channel information, thereby allowing a more efficient and precise estimation of the CSI compared with conventional CS-based approaches. Furthermore, a specific pilot arrangement is tailored for the proposed estimation scheme. This so-called random grouped pilot pattern can not only effectively protect the measurements from the inter-carrier interference (ICI) caused by Doppler spreading but can also enable the measurement matrix to meet the conditions required for MCS with relatively high probability. Simulation results demonstrate that our method has good performance at high Doppler frequencies.