Compressive Sensing Based Channel Estimation for LTE System in the Presence of Co-channel Interferences

In this paper, we consider the Long Term Evolution (LTE) system in the presence of co-channel interferences (CCIs) and a channel estimation method based on compressive sensing is proposed. Compared to the existing work on the interfered channel estimation, we make use of interferes’ signal structure which is distinguished with the Gaussians noise, a model of multi-cell channel estimation (MCCE) is set up. Promoted by the sparseness of wireless channel in the wideband system like Long Term Evolution, we develop the Joint Modified Subspace Pursuit (J-MSP) algorithm which designs new sensing matrix to overcome column highly coherent problem in the channel estimation formulation and exploits the multi-input multi-output (MIMO) channel sharing common support characteristic based on the existing Subspace Pursuit (SP) within the compressive sensing framework. After the joint estimation of channel and unknown interference symbols, the channel estimation can be obtained by using the multi-cell model. Simulations and analysis show that the proposed method I) outperforms single-cell channel estimation method greatly when there is synchronous interference; II) is comparable to Maximum Likelihood Estimator (MLE) in case of asynchronous interference; III) can also be applied to the noise-only scenario.