Delay-Doppler Channel Estimation in OTFS Systems Using DoA Estimation Techniques

In orthogonal time frequency space (OTFS) modulation, information symbols are multiplexed at the transmitter and coherently detected at the receiver in the delay-Doppler (DD) domain. The receiver requires DD domain channel gains for coherent detection and to enable channel estimation, a pilot symbol is transmitted embedded among the information symbols with sufficient guard intervals along the delay and Doppler axes. We show that the DD domain channel estimation problem can he transformed into a direction-of-arrival (DoA) estimation problem using a uniform linear array. This allows us to use state-of-the-art algorithms for DoA estimation such as off-grid sparse Bayesian inference (OG-SBI) and multiple signal classification (MUSIC) algorithms. We show that OG-SBI achieves lower mean square error and symbol error probability compared to the existing least-squares, SBI, and correlation algorithms. MUSIC is seen to work well when the pilot power is high. These are also the case when interference due to reduced guard interval is present.