Bayesian Channel Estimation in Massive MIMO System

Data trafficking is becoming a challenging task day by day due to an exponential increase in data transfer and handling. Large-scale antenna system also is known as Massive MIMO is deployed. Because of dense network deployment and restricted orthogonal pilot sequences in Massive MIMO systems, multiple neighboring cells may use the same pilot sequence which causes pilot contamination. In order to reduce the impact of pilot contamination an integrated Bayesian estimation technique is proposed, that combine uplink and downlink measurements. In this method, it is concluded that the source of pilot pollution for uplink and downlink is entirely different, hence the reciprocity is not effective. Hence we take an amalgamation of both direction measurement. The channel will remain static through a coherent interval and therefore, the channel vector among the target base station and the target terminal will also remain unaffected. By using this characteristic and Bayesian estimation method, we can obtain the channel vector for both targets and interfere base station. BER and MSE over SNR are calculated, simulation results show that this technique will significantly decrease the impact of pilot pollution and the channel vector we obtain can hold information of interference which will offer help for future's base station assistance.