Server deployment planning and performance analysis in mobile edge computing with β-Ginibre point process

摘要

Mobile Edge Computing (MEC) has been included in the architecture of 5G network, which is expected to give users access to approximate edge servers and reduce latency of computation-intensive applications. The deployment of MEC servers is significantly different with that of wireless cellular network base stations, since it needs to take transmission quality, coverage of the station/server and computational resource provisioning into account comprehensively. β-Ginibre point process comes from a recently studied stochastic geometry theory, which enables simulation of base station locations that could approximate better real station distribution. In this paper, an application of β-Ginibre point process in the deployment and location choice of MEC servers is presented, we then analyze the coverage and latency performance based on β-Ginibre point process. Extensive simulations verify that the proposed server deployment planning could increase coverage on mobile users and reduce the latency of MEC services.