AI-Enabled Joint Pilot Assignment and Power Control for Short Packet Transmission in Cell-Free mMIMO System

Ultra-reliable and low-latency communications (URLLC) is crucial for Internet of Things (IoT) applications, while cell-free massive MIMO (mMIMO) is a promising paradigm for URLLC in IoT, as it can improve the reliability of cell-edge users. However, URLLC typically uses short packet transmission, requiring pilot reuse among users to reduce pilot overhead, which inevitably leads to pilot contamination. There-fore, efficient pilot assignment and power control are essential. In this paper, we consider a cell-free mMIMO system for a large-scale smart factory, and formulate a problem to maximize sum rate by jointly optimizing pilot assignment and power control. Meanwhile, the lower bound of the ergodic data rate under finite blocklength is derived to provide an explicit expression for subsequent algorithm design. Then, we transform the non-convex sum-rate maximization problem into a Markov decision process and design a multi-agent proximal policy optimization algorithm. Finally, simulation results demonstrate the superiority of the proposed algorithm over the baselines in both static and dynamic environments.