Joint Management of Communicating and Computing Resources in Sliced 5G Networks

In the fifth generation of mobile cellular network (5G), based on the allocating of communicating and computing resources, multiple slices are formed to serve various use cases with different quality of service (QoS) requirements. Since the communicating and computing resources are limited, it is important to share them among slices efficiently. However, the requirements on communicating and computing resources are usually coupled together considering mobile edge computing (MEC) in sliced 5G networks. To this end, this paper proposes the tandem queues to represent and analyze the coupling relationship between communicating and computing resources for slices. Then, based the derived relationship, the communicating and computing resources are allocated among slices when there is no burst computing task and there is burst computing task, which are formulated by two optimal problems. For the optimal problem without considering the burst computing task, the coupling relationship of the resources is introduced to the objective function to share the resources among the slices efficiently. Then considering the burst computing task, the probability that the enough resources are required by the slices is added to the constraint conditions. And two algorithms are designed to solve the two problems. By simulating, it can be seen that the system utility can be improved since more slices can be served by the system with the proposed algorithms. Therefore, the efficiency of managing communicating and computing resources jointly is proved.