Edge-Boss: A Resource Optimization Framework at the Edge

Document Type

Conference Proceeding

Publication Title

IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC

Abstract

Recently, the benefits of moving computation to the edge were researched and proven. The paradigm of Mobile Edge Computing (MEC) provides a clear view of how offloading core network burden leads to the shortening of data access latencies, which is primal for supporting services at the edge. Increasing the density of edge data centers for supporting such requests is uneconomical. However, it is important to have a clear resource allocation strategy in such a way that user scheduling is considered. Therefore, in this paper, we study and prove how resource scheduling at the edge can substantially affect network performance. To this end, we focus on the stochastic nature of the wireless links and reduce this problem to a stochastic game. We design a software edge controller called Edge-Boss that resides at one base station and controls the wireless links available at each base station. In addition, the model ensures that Edge-Boss achieves an equilibrium that enables the Base stations (BS) to provide a higher performance (as this equilibrium acts as an incentive for the BS). Using simulations we show that Edge-Boss achieves latency reductions and a high network throughput.

DOI

10.1109/PIMRC56721.2023.10293785

Publication Date

1-1-2023

Keywords

Wireless communication, Base stations, Multi-access edge computing, Stochastic processes, Games, Throughput, Software

Comments

IR conditions: non-described

Link to Full Text

Share

COinS