Workload balance-aware edge server placement in mobile edge computing

Abstract

With the rapid development of the Internet of Things and 5G networks, the number of mobile terminals is also rapidly increasing, and the demand for computing is increasing. However, mobile devices have limitations in computing resources and energy. Mobile edge computing can reduce network congestion and improve quality of service by sinking computing resources to the edge of the network. Edge server placement is an important issue in mobile edge computing, which is to place edge servers in candidate locations and process user requests. We propose a heuristic method that combines k-means and dung beetle optimization algorithm to minimize communication latency for mobile users while balancing the workload between edge servers as much as possible. We conducted experiments on the real dataset of the Shanghai base station and compared it with several other representative methods currently available. The experimental results show that this method can achieve a balance between communication delay and load balancing, and outperforms other methods in terms of overall performance.