Resource Management and Computing Offloading to Minimize Power Consumption Cost in Slice-MEC System.

With the advancement of 5G technology, user demand grows, resources become scarce and differentiated services place greater requirements on performance in slice-MEC system. However, optimizing works in the vast majority of existing research focus only on slice selection, resource orchestration, utility improvement from the perspective of users, without considering whole system's power consumption cost. Firstly, to get over this dilemma, a two-tier architectural model for slice-MEC system is proposed, which is mainly divided into the slice instance layer and infrastructure layer. Next, we establish the optimization problem of minimizing weighted power consumption cost while guaranteeing requirements of delay and rate, taking into account both communication and computational resources. Moreover, to solve the problem above, the HPCA-based resource management and computing offloading algorithm is designed. Finally, simulation results show that the proposed algorithm in this paper outperforms than benchmark algorithms and we also analyse the performance of the system by varying different parameter conditions.