AI based Pilot Contamination Analysis for 5G MIMO based on Multi Antenna Routing Networks and Multi-User Pilot Scheduling

Abstract

Pilot contamination, a type of inter-cell interference, limits performance of large multi-input multi-output (MIMO) antenna systems.A drawback that results in ineffective bandwidth use is the burden of pilots who must estimate the channel regularly due to the acquisition of channel state data for channel estimation.Thus, there is a trade-off between spectral efficiency (SE)as well as quantity of pilots needed to evaluate channel.This research proposes novel technique in pilot contamination analysis (PCA) for 5G network based on MIMO by multi antenna routing system. The main aim is to detect pilot contamination and enhance spectral efficiency of the network. Here pilot contamination is detected using multi-user pilot scheduling with convolutional adversarial training model. As a result, a security breach occurs when crucial information slips to Eve during downlink transmission.Ability of the legitimate user to maintain secrecy can be greatly enhanced by knowing of an active eavesdropper.We also analyse the likelihood of detection, the likelihood of a false alarm, and the likelihood of a detection error.Simulation results show that suggested strategy to find PCA is effective.the proposed technique attained SINR of 72% , spectral efficiency of 85%,normalized MSE of 73%,PCA detection accuracy of 95%.