Scalability and Performance Evaluation of Edge Cloud Systems for Latency Constrained Applications

This paper presents an analysis of the scalability and performance of an edge cloud system designed to support latency-sensitive applications. A system model for geographically dispersed edge clouds is developed by considering an urban area such as Chicago and co-locating edge computing clusters with known Wi-Fi access point locations. The model also allows for provisioning of network bandwidth and processing resources with specified parameters in both edge and the cloud. The model can then be used to determine application response time (sum of network delay, compute queuing and compute processing time), as a function of offered load for different values of edge and core compute resources, and network bandwidth parameters. Numerical results are given for the city-scale scenario under consideration to show key system-level trade-offs between edge cloud and conventional cloud computing. Alternative strategies for routing service requests to edge vs. core cloud clusters are discussed and evaluated. Key conclusions from the study are: (a) the core cloud-only system outperforms the edge-only system having low inter-edge bandwidth, (b) a distributed edge cloud selection scheme can approach the global optimal assignment when the edge has sufficient compute resources and high inter-edge bandwidth, and (c) adding capacity to an existing edge network without increasing the inter-edge bandwidth contributes to network-wide congestion and can reduce system capacity.