Integrated Safe Motion Planning and Distributed Cyclic Delay Diversity

In this paper, we propose a safe motion planning protocol that integrates a distributed cyclic delay diversity (dCDD) system for indoor environments with static obstacles. In addition to collision avoidance, an additional goal of jointly minimizing energy consumption to control dynamic movements of an unmanned autonomous ground vehicle (AGV) and maximizing spectral efficiency (SE) achieved by a set of distributed remote radio heads is investigated in the framework of reinforcement learning (RL). There are several challenges, such as a lack of knowledge about the environment and nonexistent feasible mathematical analysis to utilize the distribution of the sum of the receive signal-to-noise ratios (SNRs) over the energy conscious motion planning. Thus, in this paper, we propose a model-free and off-policy soft actor critic (SAC) algorithm to learn and determine optimal actions for the AGV to reach its target with the following three objectives: i) achieving the safe motion planning that avoids collision with the static obstacles, ii) minimizing the control energy consumption, and iii) maximizing SE. Simulation results verify that these three objectives can be achieved efficiently and effectively by the proposed integrated SAC-based safe motion planning and dCDD system.