Efficient and scalable cloudlet placement approaches for edge computing in next-generation networks

Date
2022
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Emerging applications with low-latency requirements such as real-time analytics, immersive media applications, and intelligent virtual assistants have rendered Edge Computing as a critical computing infrastructure. Since large data centers cannot be placed everywhere, placing cloudlets nearby the users provides an effective solution to these requirements. Existing studies have explored the cloudlet placement problem in a homogeneous scenario with different goals such as latency minimization, load balancing, energy efficiency, and placement cost minimization. However, placing cloudlets in a highly heterogeneous deployment scenario considering the next-generation networks and IoT applications is still an open challenge. ☐ The goal of this dissertation is to specifically tackle strategic placement of cloudlets to reduce costs associated with providing edge services (deployment cost, energy consumption) as well as improving metrics important to the users such as coverage, service stability, and access latency. These problems are significantly challenging because the cloudlets have limited capacities whereas users have dynamic demands. Moreover, both users and cloudlets can be heterogeneous. As we move into next-generation networks and IoT, the placements need to be done in even denser and highly heterogeneous scenarios. Another challenging aspect is the continuous user mobility. These several placement challenges in edge computing systems are hence addressed by mathematically modeling diverse placement scenarios based on real-life constraints. However, the complexities of these problems make the mathematical models inherently NP-hard, even with the simplifying assumptions. As a result, the core of the research lies at solving these problems by designing efficient and scalable solution approaches.
Description
Keywords
Cloudlet, Distributed algorithm, Edge computing, Next-generation network, Physics-inspired
Citation