Lightweight Channel Estimation Networks for OFDM Systems
Jinbao Li, Qi Peng
Abstract
Channel estimation using neural networks has proven to be a breakthrough technology in the communications field. However, to achieve good performance, the existing studies mostly focus on building a complex neural network with a large number of layers, which consumes a lot of storage space and computing resources. In this letter, a lightweight channel estimation Transformer (LCET) is proposed for an orthogonal frequency division multiplexing (OFDM) system with pilots, based on the application of a convolutional neural network (CNN) and Transformer. The network uses a lightweight feature extraction CNN (LFEC) to extract channel features and then feeds the processed features into a lightweight-adjusted Transformer (LAT) for channel estimation. Specifically, LFEC extracts deep channel features at a low computational cost and retains the high frequency information. The LAT solves the problem of long-range feature information loss with low resource consumption. The experimental results revealed that LCET performed better than the traditional least squares estimation algorithm and state-of-the-art learning network.