Network-Flow Algorithms for Virtual Service Placements in Mobile Edge Networks
Huan Tang, Yang Wang, Xi Chen · 2021
The rise of mobile edge computing has made cloud computing resources more fully utilized. Some traditional scheduling methods can no longer meet the needs of delay-sensitive applications. With the development of service migration technology, this problem has been alleviated to a certain extent. However, the existing service migration algorithms have certain deficiencies in scalability, balance, and timeliness. For this reason, this paper proposes a scheduling framework based on the minimum-cost maximum-flow(MCMF) algorithm. We build a MCMF model according to the corresponding network conditions. Considering the impact of edge node load conditions on service quality, we have made corresponding improvements to the model to prevent services from being piled up on edge nodes. In this way, the possibility of a series of problems caused by resource competition is reduced. The experimental results show that compared with the existing results, our algorithm not only significantly reduces the total service cost, but also achieves load balancing with high timeliness.