Multi-user Cooperative Computation Offloading in Mobile Edge Computing

Mobile edge computing (MEC) is a promising technique to reduce users' response latency by deploying computation and storage resources at the network edge. In a multi-user MEC system, the MEC server may not be able to process the tasks offloaded from all users in its coverage area since it has limited computation resources as compared to the remote cloud. Device-to-device (D2D) assisted computation offloading may help to relieve the pressure on the MEC server, however, D2D communication is not always possible due to its short communication range. To overcome the limitation of D2D-assisted computation offloading, in this paper, we study the computation offloading of computationintensive tasks in a single base station multi-user MEC system and propose a cooperative computation offloading scheme to fully utilize the computation resources of the idle devices. The key idea is that the busy devices partially offload tasks to the MEC server, and the server can further offload tasks that cannot be accommodated to the idle devices. We model the computation offloading process as a sequential game and prove that the process can reach Nash equilibrium. Thus, each busy device can obtain the optimal decision for task offloading. A multi-user cooperative offloading algorithm is also proposed to solve the problem. Through extensive simulations, we verify the effectiveness of the proposed scheme and demonstrate that the proposed scheme can reduce the average task execution delay, as compared with other three schemes where cooperative computation offloading is not allowed.