Compressive Channel Estimation Using Distribution Agnostic Bayesian Method

Yi LIU  Wenbo MEI  Huiqian DU  

Publication
IEICE TRANSACTIONS on Communications   Vol.E98-B   No.8   pp.1672-1679
Publication Date: 2015/08/01
Online ISSN: 1745-1345
DOI: 10.1587/transcom.E98.B.1672
Type of Manuscript: PAPER
Category: Wireless Communication Technologies
Keyword: 
compressive channel estimation,  OFDM,  compressive sensing,  block-sparsity,  Bayesian method,  

Full Text: PDF(1.3MB)
>>Buy this Article


Summary: 
Compressive sensing (CS)-based channel estimation considerably reduces pilot symbols usage by exploiting the sparsity of the propagation channel in the delay-Doppler domain. In this paper, we consider the application of Bayesian approaches to the sparse channel estimation in orthogonal frequency division multiplexing (OFDM) systems. Taking advantage of the block-sparse structure and statistical properties of time-frequency selective channels, the proposed Bayesian method provides a more efficient and accurate estimation of the channel status information (CSI) than do conventional CS-based methods. Moreover, our estimation scheme is not limited to the Gaussian scenario but is also available for channels that have non-Gaussian priors or unknown probability density functions. This characteristic is notably useful when the prior statistics of channel coefficients cannot be precisely estimated. We also design a combo pilot pattern to improve the performance of the proposed estimation scheme. Simulation results demonstrate that our method performs well at high Doppler frequencies.