Service Composition Orchestration, Service Caching, and Computation Reuse at the Edge

The ongoing unprecedented growth of interconnected devices requesting remote computation requires that novel network connectivity and management methods be designed to keep up with the demand. Moreover, contemporary applications in AR/VR, connected vehicles, Industry 4.0/5.0 and smart-buildings require composite services to be orchestrated and run with low latency in order to meet user Quality of Experience (QoE) performance. As more and more IoT devices need connectivity, some of the pitfalls of existing infrastructure are revealed, such as interoperability due to device and data heterogeneity, limited resources on edge servers, spatial and temporal dynamics due to mobility of users, scalability for deployment, and unknown task demand and variable task size. Enabling service caching and computation reuse within service composition can greatly reduce bandwidth utilization and the overall amount of data transferred between sub-services. In doing so, composition orchestration becomes very challenging. By leveraging Named Data Networking (NDN), a distributed methodology for service composition orchestration will be developed using in-network caching of services and intermediary results.