Pricing
On the Placement of Edge Server for Mobile Edge Computing
This paper investigates how to efficiently place the edge servers (or racks) under the constraints of QoS latency, computation demands and site rental costs and proposes an efficient heuristic algorithm, namely, Offloading-Factor based edge Rack Placement (OFRP), which significantly outperforms the greedy-based algorithm.
Abstract
As an emerging technique, Mobil Edge Computing (MEC) brings high computational capacity close to the mobile users. In MEC, edge servers accept the computation-intensive demands offloaded from the mobile users. Given the projected computation demands or site rental costs, it is challenging to effectively place the edge servers such that the offloaded demands can be completed in time. In this paper, we investigate how to efficiently place the edge servers (or racks) under the constraints of QoS latency, computation demands and site rental costs. We formulate the placement problem by utilizing the technique of Integer Linear Programming (ILP). We also propose an efficient heuristic algorithm, namely, Offloading-Factor based edge Rack Placement (OFRP). OFRP guarantees an ln(n)+1-approximation boundary, where n is the size of the mobile edge network. Our simulations and experimental results demonstrate that the proposed algorithm significantly outperforms the greedy-based algorithm.