• service continuity
• edge computing
• p4
• segment routing network programming

Edge Computing is a distributed-computing solution that provides proximity services to end users while reducing latency and bandwidth usage and ensuring high dependability, which is required for constrained applications such as those for mobile and IoT devices.
To reap the benefits of Edge Computing, it is necessary to track device mobility and have services "follow" them; to this end, services may be deployed in containers that can be quickly transferred from one location to another based on the linked mobile node position.
To make the service migration transparent to the client, network traffic must be properly managed. Many technologies and protocols can be used to route traffic and manage network device behavior: in this work, service continuity with workload migration is provided by exploiting the functionalities and flexibility given by Segment Routing and the P4 language for network device programmability. Within the core network, the solution enables dynamic flow allocation and migration management.
Several Segment Routing behaviors are defined to achieve this goal: each behavior is implemented by a P4 program loaded in programmable switches, which are configured centrally by an SDN controller.
The proposed solution is validated on a testbed comprised of Mininet networks with P4 BMv2 switches and an ONOS SDN controller, and a comparison between Segment Routing behaviors and against state-of-the-art alternatives is performed. The results obtained highlight the scalability and flexibility of the proposed solution compared to the state-of-the-art.