Title: Training-based channel estimation for massive MIMO systems using LAS algorithm

Authors: Mitesh Solanki; Shilpi Gupta

Addresses: Department of Electronics Engineering, Sardar Vallabhbhai National Institute of Technology, Surat, Gujarat, India ' Department of Electronics Engineering, Sardar Vallabhbhai National Institute of Technology, Surat, Gujarat, India

Abstract: Likelihood ascent search (LAS) recently emerged as a promising low-complexity near-optimal detection algorithm for massive multiple-input multiple-output (MIMO) systems. LAS detection algorithm design with perfect channel state information (CSI) consideration is an area of research attention. In the real-time scenario, channel estimation has been challenging in communication systems. This article proposes a lattice reduction-based LAS (LR-LAS) detection for massive MIMO systems under imperfect CSI. Our proposed LR-LAS take the medium-to-strong channel estimation error of antennas into account when computing the maximum likelihood (ML) cost function. Our approach presents a novel iterative detection and training-based channel estimation framework. Simulation results show that the proposed LR-LAS detector significantly improves bit error rate (BER) performance compared to their conventional LAS counterpart. Hence, our proposed detector can be used as an efficient candidate under imperfect CSI in the system.

Keywords: likelihood ascent search; LAS; channel state information; CSI; lattice reduction; massive MIMO; maximum likelihood; training-based channel estimation.

DOI: 10.1504/IJUWBCS.2022.123423

International Journal of Ultra Wideband Communications and Systems, 2022 Vol.5 No.2, pp.77 - 83

Received: 03 Jan 2021
Accepted: 09 Mar 2021
Published online: 20 Jun 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article