Open AccessLoad balancing edge server placement method with QoS requirements in wireless metropolitan area networks
Author(s) -
Li Xingcun,
Zeng Feng,
Fang Guanyun,
Huang Yinan,
Tao Xunlin
Publication year - 2020
Publication title -
iet communications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.355
H-Index - 62
eISSN - 1751-8636
pISSN - 1751-8628
DOI - 10.1049/iet-com.2020.0651
Subject(s) - server , computer science , load balancing (electrical power) , mobile edge computing , computer network , quality of service , greedy algorithm , enhanced data rates for gsm evolution , distributed computing , cloud computing , network load balancing services , workload , wireless , edge computing , algorithm , operating system , mathematics , artificial intelligence , geometry , grid
Mobile edge computing (MEC) is concerned with moving complex tasks from data sources to nearby computing resources, which can reduce computing latency and remote cloud workload. Although there has been significant research in the field of MEC, research on edge server placement in wireless metropolitan area networks (WMANs) is overlooked, and the load balancing problem of edge servers is seldom discussed. From a practical perspective, how to place edge servers efficiently in WMANs while considering load balancing between edge servers is studied. A greedy algorithm is proposed that can balance the workload of edge servers more effectively. However, the performance of the greedy algorithm as the number of servers placed increases is not ideal. Therefore, the authors combine the greedy algorithm with a genetic algorithm (GA) to minimise the number of edge servers while ensuring load balancing between edge servers and quality of service (QoS) requirements for mobile users. Finally, they conduct simulation experiments and compare the proposed algorithms with other algorithms. The improved GA proposed is superior to the greedy algorithm in terms of load balancing and the number of servers. The experimental results demonstrate that the algorithm has excellent performance.