Dynamic Adaptive Network Edge Service Migration Method Based on a Docker Container
Xiangjun Zhang, Weiguo Wu, Chi Zhang, Wei Song · 2019
With the development and maturity of 5G technology, edge networks have become more suitable for moving many low-latency cloud services to the edges of networks, thus significantly improving user service quality. Since mobile terminal users move with their location, the edge computing platform needs to seamlessly migrate services using users' mobile devices. However, the network environment of an edge network is complex, and service migration faces many challenges, such as how to overcome network delays and reduce the service interruption time. In this paper, we propose a service migration method based on Docker base image and design a compression algorithm for an adaptive network environment to compress images and other data. This method balances the multi-core processor computing power and network bandwidth resources. We implement pipeline parallel processing of each process and build a model to handle the situation. The results show that our method is more adaptable to changing network environments, and the migration time is reduced by 45.4%, on average, compared with the traditional uncompressed service switching method.