Towards efficient offloading in fog/edge computing by approximating effect of externalities
Vladimir Marbukh · 2018
Fog/Edge computing is an emerging architecture, which extends the Cloud computing paradigm to the edge of the network, enabling new applications and services, including Internet of Things (IoT). End devices have certain computation tasks, which can be completed either locally on the end device or remotely on the cloud via computation offloading. Due to mobility and highly dynamic nature of end users, the access is wireless with severe physical limitations on the access network capacity to sustain stringent latency requirements for streaming and real-time applications for large number of end users over wide-spread geographical area. Assuming that access network has cellular architecture and is interference limited, we quantify system performance by the aggregate utility. Optimization of the system computing and communication resources requires accounting for externalities due to interference created by the individual offloading transmissions. While each base station can account for the intracell externalities, exact accounting for the intercell externalities in a large-scale network is not feasible due to prohibitive exchange of the “microscopic” information on the intercell interference by each end user. We propose a “macroscopic” approximation for the intercell externalities, which significantly reduces information exchange, and thus allows for decentralized near-maximization of the aggregate system utility.