Edge Computing-Enabled Networks: Strategic MEC Deployment and Dynamic Task Scheduling

Abstract

Telecommunication operators are increasingly adopting Network Function Virtualization (NFV) and Multi-access Edge Computing (MEC) to meet the ultra-low latency and high reliability demands of 5G/6G services. These technologies enable the delivery of services through Service Function Chains (SFCs) composed of Virtual Network Functions (VNFs), utilizing computing resources near end users. A key challenge in this architecture is the efficient allocation of resources and the strategic placement of MEC sites to host VNFs. This thesis introduces a novel approach to efficiently determine MEC site locations to enhance dynamic network performance. Instead of conducting exhaustive and time-consuming simulations to evaluate each potential selection of MEC sites, we propose a precomputed load balance metric approach. By leveraging the Jain Fairness Index (JFI), promising site selections can be quickly identified. Our research demonstrates a statistically significant negative monotonic relationship between precomputed JFI and blocking probability when SFCs are dynamically established and released. This method allows network operators to prioritize detailed simulations only for the most promising MEC site combinations, significantly reducing the computational burden while improving network planning and operation.

Beyond telecommunication networks, the principles of edge computing are also highly relevant for Industry 4.0. Efficient resource allocation and task scheduling are crucial in industrial extreme-edge computing, particularly for Automated Guided Vehicles (AGVs). These mobile, resource-limited devices leverage embedded edge computing to minimize latency in dynamic operations. In this context, this thesis also introduces a Queue-Aware Scheduling and Deadlock Mitigation Strategy (QASDMS) to enhance multi-AGV coordination by enabling concurrent movement and data processing. Simulation results show that QASDMS improves resource utilization and reduces operation time. This research bridges edge computing with industrial automation, offering scalable and intelligent optimization strategies for next-generation infrastructures.