Federated Deep Reinforcement Learning for RIS-Assisted Indoor Multi-Robot Communication Systems

Abstract

Indoor multi-robot communications face two key challenges: one is the severe\nsignal strength degradation caused by blockages (e.g., walls) and the other is\nthe dynamic environment caused by robot mobility. To address these issues, we\nconsider the reconfigurable intelligent surface (RIS) to overcome the signal\nblockage and assist the trajectory design among multiple robots. Meanwhile, the\nnon-orthogonal multiple access (NOMA) is adopted to cope with the scarcity of\nspectrum and enhance the connectivity of robots. Considering the limited\nbattery capacity of robots, we aim to maximize the energy efficiency by jointly\noptimizing the transmit power of the access point (AP), the phase shifts of the\nRIS, and the trajectory of robots. A novel federated deep reinforcement\nlearning (F-DRL) approach is developed to solve this challenging problem with\none dynamic long-term objective. Through each robot planning its path and\ndownlink power, the AP only needs to determine the phase shifts of the RIS,\nwhich can significantly save the computation overhead due to the reduced\ntraining dimension. Simulation results reveal the following findings: I) the\nproposed F-DRL can reduce at least 86% convergence time compared to the\ncentralized DRL; II) the designed algorithm can adapt to the increasing number\nof robots; III) compared to traditional OMA-based benchmarks, NOMA-enhanced\nschemes can achieve higher energy efficiency.\n