A Novel Cost Optimization Framework for Multi-Cloudlet Environment over Optical Access Networks

In the post-4G era, "low latency" has become one of the most important network requirements along with support for ultra-high capacity and ultra-high reliability. This has led to the evolution of "cloudlets" to support similar services provided by legacy cloud technology viz., storage capacity and computational capability. In this paper, we propose a novel framework for cloudlet-empowered-cloud network design and planning, based on optical access infrastructures. Our focus is on network planning to optimize network infrastructure cost by formulating a nonlinear programming model to identify placement locations of cloudlet servers subjected to capacity and latency constraints. We demonstrate the feasibility of the proposed model against urban, suburban and rural scenarios, providing guidance on the installation and maintenance costs. Furthermore, we assess the percentage of incremental energy arising from the presence of cloudlets in the optical access network. The proposed framework is a first in yielding insights that will serve as a foundation for further cloudlet network planning strategies.