Computation Offloading and Resource Allocation based on Cell-Free Radio Access Network
Zhan Chen, Xin Su · 2022 IEEE 6th Information Technology and Mechatronics Engineering Conference (ITOEC) · 2022
Aiming at the high latency problem of computing task processing in traditional cellular radio access network, a method of computing task offloading and resource allocation based on cell-free fog radio access networks was proposed. A large number of wireless access points (AP) provide services in a “user-centric” manner. Fog-access points (F-AP) and cloud servers provide edge and cloud computing and caching functions, and establish a three-level offloading model of local, edge and cloud for computing task. Through deep reinforcement learning Deep Reinforcement Learning (DRL) method to make the optimal offloading decision and resource allocation for computing tasks, thereby reducing the overall delay of the system. The simulation results show that the optimal unloading decision significantly reduces the system delay and improves the system performance to a certain extent.