A Low Complexity Near-Optimal Detector Based on Teaching-Learning Algorithm for Massive MIMO

Amiriara, Hamid; Zahabi, Mohammadreza

doi:10.22091/jemsc.2024.8730.1167

A Low Complexity Near-Optimal Detector Based on Teaching-Learning Algorithm for Massive MIMO

Document Type : Original Article

Authors

¹ Electrical Engineering, Sharif University of Technology, Tehran, Iran.

² Faculty of Electrical and Computer Engineering, Babol Noshirvani University of Technology, Babol, Iran

10.22091/jemsc.2024.8730.1167

Abstract

Despite the advantages of massive multi-input multi-output (MIMO) technology, traditional multi-antenna detection algorithms are not suitable for systems with large-scale antennas, and the use of this technology requires a significant increase in computational costs. In this paper, a low-complexity receiver is proposed using a Teaching-Learning based optimization (TLBO) heuristic algorithm for a large-scale system. The TLBO algorithm, as one of the advanced methods of intelligence, is very useful for large-scale problems. Therefore, we use this method to search for the optimal solution vector in the modulation alphabet. In order to prove the accuracy and efficiency of the detector, it was suggested to simulate the system with different dimensions from 64×64 to 1024×1024. The proposed TLBO detector, in a limited time, achieves a bit error rate (BER) 10^(-5) in the average signal-to-noise ratio of 12 dB, which is approximately equal to the optimal detector performance, and maximum likelihood. As a result, the proposed detector is very efficient for use in massive MIMO systems.

Keywords

Main Subjects