Service Migration in Multi-domain Cellular Networks based on Machine Learning Approaches

Abstract

The number of mobile subscribers has increased drastically with the deployment of high-performance mobile cellular networks such as Long Term Evolution (4G) (LTE), and the upcoming 5th Generation Mobile Cellular Network (5G). Seamless connectivity is an important factor to provide better Quality-of-Service (QoS) as well as Quality-of-Experience (QoE) in cellular networks. In this regard, the utilization of Mobile Edge Computing (MEC) technology allows to bring required mobile services close to the end users to reduce latency, and increase service quality. However, switching between MECs nodes needs to be optimized in order to satisfy the expected quality of service as well as to guarantee service continuity SC. This paper addresses this problem by proposing a methodology and a prediction algorithm to boost SC in a MEC configuration.