Research on the Application of Multilayer Perceptron in Improving Massive MIMO Channel Estimation

Hoai Trung Tran

Massive MIMO applications hold significant potential in 6th generation mobile communication systems due to their ability to optimize transmission space utilization. The conversion of real-world physical channel matrices into spatially discrete channel models has been studied. For this channel model, the concept of a virtual channel matrix emerged. Algorithms OMP (Orthogonal Matching Pursue) to find channel matrices exist. However, the accuracy of evaluating the coefficients of these channel matrices is not high depending on matrix computation. Furthermore, the OMP applies to each virtual path, so the computational complexity is high. This paper uses the Multilayer Perceptron (MLP) estimation algorithm to determine a virtual channel matrix with higher accuracy using received signals as input. Furthermore, the algorithm's complexity is significantly reduced compared to the OMP algorithm.

Select Volume / Issues:

• Volume 17,Issue 1, Jan. 2026

Year:

2026

Type of Publication:

Article

Keywords:

Beamforming; OMP; MLP; Virtual Channel Matrix

Journal:

IJECCE

Volume:

17

Number:

1

Pages:

1-6

Month:

January

ISSN:

2249-071X