A Novel Tabu Search Detector for OTFS

Orthogonal Time Frequency Space (OTFS) has demonstrated to be a promising technique in high-mobility scenarios. However, designing a low-complexity OTFS equalizer that achieves full diversity in transmission scenarios with short frame remains challenging. In this letter, we first develop an efficient Reduced Neighborhood Reactive Tabu Search (RN-RTS) detector for OTFS by leveraging the circulant property of the channel matrix block and deriving the recursive expressions for certain RN-RTS components. Moreover, the maximal ratio combining (MRC) method is utilized to optimize the initial solution, thus further enhancing its performance. Simulation results reveal that the proposed algorithm achieves a good tradeoff between the performance and the complexity.