Optimizing Task Processing Efficiency in MEC Networks Through Cooperative Offloading and Resource Allocation

As Internet of Things (IoT) devices continue to proliferate and users' demands for low-latency and compute-intensive services grow, mobile edge computing (MEC) can effectively reduce task processing delay and energy consumption by tasks offloading to the MEC server. In the context of an MEC network comprising IoT mobile device (IMD), neighboring IMDs, and MEC servers, we propose a joint cooperative partial offloading and resource allocation scheme that selects a neighboring IMD and an MEC server as offloading assistants to assist the IMD in processing tasks in MEC systems. Further, the allocation of IMD computing resources should also be taken into consideration. By jointly optimizing the offloading decision and computational resource allocation, we minimize the weighted sum cost of task processing delay and energy consumption of all IMDs as an optimization problem and solve the non-convex problem by a deep Q-network (DQN) algorithm. Extensive experimental results show that the proposed scheme can effectively reduce the cost compared to other baseline schemes.