Performance analysis of large scale MIMO system using noise and relevancy aware low complexity detection algorithm

Abstract

Complexity reduction for Large Scale Multiple Input and Multiple Output (LM-MIMO) system is a most concentrated research issue to ensure the optimal and reliable signal processing. In the existing work, optimal low complexity detection procedure has been carried out by using Pruning based Maximum Likelihood Detection using Low Complexity detection Algorithm (PRUN-MLD-LCDA). The existing work might lacks in its performance with the reception signals with increased noise level and more missing information. This problem can be overcome in the proposed method namely Noise and Relevancy aware Low Complexity Detection (NRLCD). The objective of this research method is to ensure the accuracy recover of signals with reduced complexity. In this work, initially signal pruning is performed to filter out the alike signals. This is done by Normalized Cross Correlation (NCC) based pruning which is used to avoid the irrelevant signals. And then Adaptive filter is used to remove the noises present in the signals before analyzing the complexity. Then, Hybrid Genetic-Branch-and-Bound (Hybrid Genetic-BB) detection algorithm is utilized to detect the received signals with low complexity. The overall simulation of the research methods are performed using Matlab from which it is inferred that NRLCD research method significantly achieves better Bit Error Rate (BER) performance compared to the existing methods.