Channel Estimation Based on Compressed Sensing for Massive MIMO Systems With Lens Antenna Array
Elham Sharifi,Mahmood Mohassel Feghhi,Ghanbar Azarnia,Sajjad Nouri,Duehee Lee,Jalil Piran,Md. Jalil Piran
DOI: https://doi.org/10.1109/access.2023.3297884
IF: 3.9
2023-08-05
IEEE Access
Abstract:With the emergence of fifth-generation cellular networks (5G), there has been significant interest in multi-input-multi-output (MIMO) systems. MIMO systems aim to achieve several key objectives, including increasing capacity, mitigating the negative effects of multi-path propagation, minimizing interference, and achieving higher data rates. Furthermore, the utilization of millimeter-wave (mmWave) technology and high bandwidth can address traffic congestion and interference challenges, leading to substantial improvements in data rates, spectral efficiency, and overall bandwidth. In the realm of mmWave communications, massive MIMO systems incorporating lens antenna arrays have proven effective in reducing the number of required radio-frequency chains. This paper presents two distinct approaches to address the challenge of massive MIMO channel estimation in mmWave communications. The first approach proposes a compressed sensing (CS) scheme based on convex optimization, which offers accurate and low-complexity channel estimation. The second approach introduces an estimation algorithm based on the greedy method, which provides fast reconstruction, a straightforward geometric interpretation, and a mathematically efficient framework. Extensive simulations demonstrate the superior performance of the proposed algorithms compared to similar methods such as support detection (SD), orthogonal matching pursuit (OMP), and sparsity mask detection (SMD). The proposed algorithms exhibit higher channel estimation accuracy, better recovery quality, and fast convergence rates.
computer science, information systems,telecommunications,engineering, electrical & electronic