Research on 5G Signal Frequency Offset Estimation Method Based on Fully Connected Layer Network
Ouyang Xu, Junkai Ma, Dan‐Xia Xu, Jin Chen · Internet Technology Letters · 2025
ABSTRACT For the problem that carrier frequency offset in the 5G communication system leads to the inability of the demodulation end to correctly recover the original information sequence, a 5G signal frequency offset estimation method based on the fully connected layer network is proposed. The transmission signal adopts the 64QAM high‐order modulation mode, establishing the nonlinear mapping relationship between the IQ two‐channel signals and the frequency offset. The frequency offset value is estimated by training the model, and the frequency offset compensation is performed at the receiving end for demodulation. After simulation tests, it can achieve a 100% correct rate of frequency offset estimation under the condition of a 20 dB signal‐to‐noise ratio, and the estimation accuracy is 0.1 Hz.