Abstract

The need for faster and higher-capacity networks that can sustain modern, high-demanding applications has driven the development of 5G technology. Moreover, support for low-latency communication (1ms -10ms) is one of the main requirements of 5G systems. Multi-access Edge Computing (MEC) has been seen as a key component to attain the 5G objectives, since it allows hosting and executing critical services at the vicinity of users, thus reducing the latency to its minimum. Motivated by the evolution of real-time applications, we propose and evaluate two different mechanisms to improve the end-user experience by leveraging container-based live migration technologies. The first solution is aware of the users’ mobility patterns, while the other is oblivious to the users’ paths. Our results show approximately 50 percent reduction in downtime, which demonstrates the efficiency of the proposed solutions compared to prior works using similar underlying technology, i.e., LXC or Docker.