A Frequency Offset Estimation Algorithm Based on Under-Sampling for THz Communication

The frequency offset estimation algorithm for terahertz (THz) communication is not only required to deal with estimation accuracy, SNR (signal-to-noise ratio) threshold and estimation range, but also needs to take high Doppler shift caused by high operating band, the complexity of real-signal processing and large hardware cost into consideration. A carrier frequency offset estimation algorithm based on under sampling for THz communication is proposed in this paper. This algorithm utilizes methods including narrow band filtering (the bandwidth of filtered signal is only about 0.1% of the original signal bandwidth), under-sampling based on coprime sampling and second time estimation. For signal processing, we use FFT (Fast Fourier Transform) to achieve correlation operation on the frequency domain, which effectively reduce the computational complexity. The method we proposed significantly reduce the sampling rate as well as improve the estimation accuracy and thus can be applied to THz communication. The simulation results show that the algorithm can estimate a large dynamic range of the frequency offset at a low SNR with a low sampling rate, which reduce the difficulty of signal processing and hardware design.