A Scalable Testbed for Task Offloading and Deployment of Heterogeneous Edge Computing

Abstract

Heterogeneous edge computing has become a new computing paradigm with the aim of moving the computation away from cloud towards the edge of the network. With the rapid increase in communication data volume, changing the network structure to provide hard real-time services is one of the goals of edge computing. A proper testbed is essential for fully validating edge computing algorithms, especially for computation offloading. In this paper, we present a scalable testbed for task offloading and deployment of heterogeneous edge computing adopting cross-platform container technique to encapsulate the software environment and runtime dependencies of heterogeneous computing units. extensive experiments have demonstrated that containers do not bring extra time consumption or occupy extra computing resources.