DRL-Based Noncoherent PPM Detection in Time-Varying Wireless Channels

초록

Pulse position modulation (PPM) is an energy-efficient modulation technique widely used in low-complexity wireless communication systems. However, conventional noncoherent PPM receivers, which rely on simple slot-wise envelope comparison, suffer significant performance degradation in time-varying Rayleigh fading channels. To address this limitation, we propose a deep reinforcement learning (DRL)-based detection framework for Q-ary noncoherent PPM. The proposed detector observes enriched envelope-domain features extracted from the received waveform and learns a slot-selection policy during training, while performing inference using only received observations without requiring channel state information. Simulation results show that the DRL-based receiver improves detection accuracy compared with conventional noncoherent detectors under time-varying fading conditions. © 1997-2012 IEEE.

키워드