A Novel Resource Allocation Method for Network Slices in Power Grid

Abstract

Abstract To allocate the network slice optimally, a two-phase optimal allocation approach is proposed to assign the network slices with users. First, the network slice allocation system in the core cloud network is modeled as a Markov process, and then the deep reinforcement learn is adopted to allocate the network slices. In the second phase, we formulate the network slice allocation problem as Lagrange multipliers problem. The iteration method is designed to achieve the optimization problem. Numerical results demonstrate that the proposed approach can allocate the network slices at a high efficiency.