Channel Estimation in Uplink Multi-User Scenario using OTFS Modulation
Yatish Pachigolla, Lorenzo Zaniboni, Mahdi Mahvari
TL;DR
This work addresses reliable uplink channel estimation for OTFS in multi-user, doubly selective channels by formulating the DD-domain channels as a sparse vector and solving the resulting problem with CS-based OMP and impulse-pilot methods, including an embedded pilot/data scheme. Formally, the system is r = Ψ h + w, where h is sparse, and the paper compares OMP and threshold-based estimators across user counts, analyzes OMP with side information (SCI), and demonstrates a bandwidth-efficient pilot/data placement enabling joint channel estimation and data detection. The main contributions are (i) a comparative analysis of OMP and threshold-based estimators for varying numbers of users, (ii) an investigation of OMP with SCI to reduce computational load, and (iii) a practical embedded pilot/data strategy that preserves spectral efficiency while supporting multi-user detection via LMMSE. The results indicate that OMP is advantageous for few users, impulse-pilot methods scale better with more users, SCI reduces complexity, and embedded pilot/data can approach perfect CSI performance at moderate SNR, offering actionable guidance for OTFS in multi-user uplinks.
Abstract
Channel estimation techniques for orthogonal time frequency space (OTFS) modulation scheme are investigated. The orthogonal matching pursuit algorithm is investigated with and without side channel information, and an efficient data placement is proposed alongside the pilot in the multi-user scenario based on impulse pilot-based estimation. Finally, the performance of the estimation techniques across different multi-user scenarios is evaluated and compared, highlighting the strengths and weaknesses of each method