Iterative Channel Estimation for OTFS Systems Based on Low-PAPR Hybrid Superimposed Pilots
Maowu Zhou, Fangjiong Chen, Hua Yu, Jianxian Lu
Abstract
Traditional superimposed single pilot (SSIP) and superimposed sparse pilot (SSPP) schemes require high-power pilot symbols for channel estimation, leading to a high peak-to-average power ratio (PAPR). To tackle this problem, we develop a new hybrid superimposed pilots (HSP) structure for orthogonal time frequency space modulation (OTFS) systems, which consist of a high-power center pilot and low-power edge pilots. Based on the HSP structure, a two-stage iterative channel estimation (TSICE) algorithm is developed. In the first stage, the center pilot is exploited to estimate the delay/Doppler taps in the delay-Doppler (DD) domain. In the second stage, the multi-path channel gains and transmitted data are calculated according to the estimated taps. The two stages are iteratively applied to refine the channel estimation. The proposed TSICE algorithm does not require a pre-set threshold, and outperforms other pilot deployment schemes in bit error ratio (BER) and PAPR performance. Simulation results give some illustrative examples.