Estimation of Sparse Channel Using Bayesian Gaussian Mixture and CS-Aided Techniques for Pilot Contaminated Massive MIMO System

Abstract

Massive MIMO is established by future wireless communication systems to facilitate real-time user applications with having a higher of data transfer rates and spectral efficacy through perfect channel knowledge. Many pilot reuse schemes are working to improve channel performance regarding pilot overhead and channel knowledge. However, the pilot contamination becomes a major challenging issue in massive MIMO applications due to erroneous channel state information between neighboring cells. To tackle this issue, analyzing to the estimation of channel parameters in the preferred paths and interference from neighboring cells in the underdetermined system is essential. For estimating the channel behavior over Sparse Bayesian and Compressed sensing Aided methods are proposed by surviving the Pilot contamination effects by changing the angular to beam domain for sparse channel approximation. Simulation results have demonstrated the effectiveness of Bayesian and compressed sensing by compared with the standard estimators in connection with estimation accuracy, especially in pilot contamination.