Joint Computation Offloading and Multi-User Scheduling for Max-Latency Minimization in MEC-Aided Cell-Free Networks

Combining cell-free networks and mobile edge computing (MEC) is expected to meet the low latency and multi-user access requirements of the Industrial Internet of Things (IIoT). Firstly, this paper considers MEC-aided cell-free networks in multi-user access scenarios, our goal is minimizing the maximum user latency. Secondly, in order to mitigate the impact of both multi-user interference and pilot contamination on system performance, we divide the frequency band into multiple sub-channels and dynamically schedule users on sub-channels in the uplink. Subsequently, we jointly optimize computation offloading, multi-scheduling and computational resources allocation, resulting in a complex non-convex optimization problem. Finally, To address this problem, we introduce a Bisection Search Feasibility (BSF)-Proximal Policy Optimization (PPO) algorithm, PPO algorithm is utilized to determine user scheduling strategy, and then the remaining non-convex problem is converted into a convex opti-mization problem using BSF algorithm. The simulation results demonstrate that the suggested algorithm achieves rapid convergence and surpasses the performance of baseline approaches.