Sparse-Recovery-Based Channel Estimation in Orthogonal Time-Frequency Space Modulation for High-Mobility Scenarios in 5G and Beyond

Abstract

Orthogonal time frequency space (OTFS) modulation is a breakthrough waveform that significantly outperforms conventional modulation schemes in high-mobility scenarios. Unlike traditional approaches that operate in the time-frequency domain, OTFS exploits the delay-Doppler domain, transforming a time-varying channel into a nearly time-invariant one. This paper focuses on the critical challenge of channel estimation (CE) in OTFS-based downlink communication. Recognizing the inherent sparsity of the delay-Doppler domain, the CE problem is formulated as a sparse recovery task, enabling the use of advanced compressed sensing techniques. A robust greedy algorithm, namely multipath matching pursuit (MMP), is aided for OTFS to enhance estimation accuracy. The effectiveness of MMP is benchmarked against orthogonal matching pursuit (OMP) and conventional impulse-based estimation methods. Simulation results demonstrate that the proposed MMP-based CE technique significantly improves channel state information acquisition and achieves superior normalized mean square error performance, making it a promising solution for high-mobility 5G and beyond communication systems.

Keywords