A hybrid autoencoder and index modulation framework for OTFS modulation

Abstract

This paper presents an innovative approach to orthogonal time frequency space (OTFS) modulation by integrating autoencoder-based enhanced (AEE) joint delay-Doppler index modulation (JDDIM) techniques. The proposed AEE-JDDIM-OTFS framework leverages deep learning to optimize the mapping and demapping processes, significantly improving spectral and energy efficiency in high-mobility communication scenarios. The system's performance is further enhanced by the introduction of a low-complexity greedy detector that maintains robust detection accuracy, even under imperfect channel state information (CSI) conditions. Extensive simulation results demonstrate that the proposed scheme achieves superior bit error rate (BER) performance compared to conventional OTFS and other OTFS-based modulation schemes, even in imperfect channel state information situations. The findings suggest that the AEE-JDDIM-OTFS framework offers a practical, low-complexity solution with promising potential for next-generation wireless communication systems.