Latency-Sensitive Service Chaining with Isolation Constraints

Multi-access Edge Computing (MEC) is one of the key enablers in 6G. By offering computing capabilities near the end users, it enables lower environmental impact, higher end-to-end latency, higher resiliency, among others. It is also a critical part in the Cloud-Edge-IoT continuum. However, the deployment of services in the edge poses some security challenges. This work aims at solving orchestration issues, such as how to deal with service deployment while meeting service-specific requirements, while at the same time meeting isolation requirements, in a multi-cluster environment. Our algorithm is estimated to be around 20% better than the baseline, while at the same time offering guarantees on the communication latency and isolation requirements, which the baseline cannot offer. We were able to build an orchestration engine, with an advanced algorithm to compute optimal placement of microservices, integrated in a state-of-art emulation platform, with industry best practice. This demonstrates the feasibility in practice and efficiency of our approach.