Design and research of high-speed channel estimation algorithm in OFDM system based on GCE-RNN algorithm
نویسندگان
1 Anhui Xinhua University, Heifei, 230088, China.
2 Management and Science University, Shah Alam 40100, Selangor, Malaysia.
3 Management and Science University, Shah Alam 40100, Selangor, Malaysia.
doi
10.22105/jarie.2025.526233.1806چکیده
Due to the presence of Inter-Carrier Interference (ICI) and Inter-Symbol Interference (ISI) in Orthogonal Frequency Division Multiplexing (OFDM) systems under high-velocity mobility conditions, channel estimation algorithms are challenged. This article proposes a channel estimation method using a Generalized Complex Exponential-Recurrent Neural Network (GCE-RNN). This method uses the Generalized CE-BEM (GCE-BEM) model to fit the high-speed channel. Then it uses the estimated Channel Impulse Response (CIR) obtained by traditional channel estimation algorithms as the original image. RNN is used to remove the noise error introduced in the GCE-BEM Least Squares (LS) estimation result, and the denoised image is the final estimated CIR. This scheme fully utilizes the data fitting function of Neural Networks (NNs). It reduces the inherent modeling error of Basis Expansion Model (BEM) by constructing a mapping relationship between channel estimation values and true values. The simulation results show that GCE-RNN exhibits better estimation performance at different Signal-to-Noise Ratios (SNR) and speeds.
