Pilot Design via Cyclic Algorithm for Channel Estimation in Zak-OTFS Communication Systems

In this paper we present an algorithm for generating pilot sequences with a good auto-ambiguity function, which can be employed for channel estimation in Zak-OTFS modulation-based wireless communication systems. The sequences designed by the proposed algorithm have an auto-ambiguity function that approaches the ideal thumbtack structure within the fundamental delay-Doppler period of the Zak-OTFS system. The proposed algorithm is rooted in the cyclic algorithmic framework, which enjoys several useful properties: monotonic convergence, the ability to design sequences of arbitrary dimension, and the handling of information symbol constraints like unimodularity. To validate our approach, we perform numerical simulations in which we estimate the channel in a Zak-OTFS system using the proposed pilot sequences. Our results show that, compared to existing analytical sequences, the proposed sequences achieve lower normalized mean square error (NMSE) in channel estimation and superior bit-error-rate (BER) performance.