A distributed computing model based on delegation of serverless microservices in a cloud-to-thing environment
Antonio Silva, Paulo Mendes, Denis Lima do Rosário, Eduardo Cerqueira, João Paulo C. L. da Costa, Edison Pignaton de Freitas · Journal of Internet Services and Applications · 2024
This article explores serverless cloud-to-thing architecture and the virtualization of functions across the network as a solution for implementing services across the cloud-to-thing continuum. It addresses the challenges posed by the emergence of 5G and 6G networks, where they require high bandwidth, low latency, and real-time computing capabilities. The proposed architecture leverages edge computing principles, orchestrating computing functions across edge/fog infrastructure. A serverless approach with microservices offers flexibility and scalability for deploying services on heterogeneous devices. The proof-of-concept implementation demonstrates the architecture's suitability for cloud-to-thing solutions and highlights possible research directions for improving its efficiency and reliability in dynamic network environments. In Scenario A, simulation results indicated a 23% rise in throughput at the cloud level within the first 2 seconds. At the 12-second mark, the throughput became uniformly distributed, indicating a significant offloading of computational tasks. Scenario B showed an almost linear throughput distribution between the cloud and the satellite starting at 8 seconds, highlighting the framework's capacity for dynamic reallocation of computing functions in real-time.