An Efficient Doppler Sensing for 5GNR System in Fast Time-Varying Communications Scenarios

In this paper, we considers an efficient sensing scheme for doppler frequency in fast time-varying channels. Specifically, we formulate the perception and estimation of doppler frequency with using only one OFDM symbol in 5GNR system, which will gradually reduce the training overhead compared with the classical scheme with using multiple OFDM symbols. There is no mature algorithm for frequency offset estimation within one symbol in frequency-domain. We propose two novel frequency offset estimation structures based on the adaptive filter and match filter. The doppler frequency offset in the channel can be perceived and estimated by using the estimated filter coefficients. Numerical simulation fully verifies the effectiveness of the two proposed algorithms. Meanwhile, compared with CP-based time-domain frequency offset estimation algorithm, the proposed algorithms have the ability to estimate arbitrary integer frequency offset and decimal frequency offset. The proposed scheme with adaptive filter not only has the ability of sensing and estimating frequency offset, but also has the ability of tracking, which can adapt to the rapid doppler variation scenario. In addition, the proposed scheme with match filter has strong performance toughness, which means that the amount of computation can be greatly reduced to implement low-complexity with a slight performance loss.