Fair and Efficient Resource Allocation Optimization in the IoV Environments for Daily Activities and Emergency Scenarios

Abstract

The rapid development of the Internet of Vehicles (IoV) has resulted in many computation-intensive and delay-sensitive applications, particularly in the context of 5G networks. While edge computing, including the use of Extreme Edge Devices (EEDs), has emerged as a promising solution to address latency issues, existing studies primarily focus on optimizing task completion or minimizing delays from the perspectives of requesters or EEDs, usually ignoring the need for fairness between these stakeholders. This thesis proposes a novel resource allocation optimization models . The first model focuses on balancing fairness between requesters and the EED devices, which are called responders in our work in IoV environments. In our model, requesters' tasks are completed relatively quickly in terms of the number of completed tasks, response time, and cost. Furthermore, by striking a balance between profits and the quantity of CPU cycles left, our suggested model guarantees that devices are not overloaded. We aim to maximize the number of completed tasks while minimizing delays and preserving the fairness of requesters and devices. We perform detailed experiments on randomly generated data instances. The results in the first work show the model's effectiveness in achieving its objectives regarding various factors such as task execution time, response time, cost, and profit in IoV environments. In the second model, we further develop a hybrid approach to optimizing resource allocation and task offloading within IoV systems, utilizing the capabilities of edge nodes (EN). To provide a fair and effective allocation of computing tasks, our suggested solution integrates the Max Carrier-to-Interference Ratio (Max C/I) for optimal channel selection with a Multi-objective Binary Integer Linear Program (MBILP) for resource allocation. The system predicts high-demand moments from predictive data applications, improving system responsiveness and reliability. Through the predictive data application, the system is also able to anticipate high-demand moments, thus working better and more efficiently. As predicted by the simulation results using SUMO simulator, the proposed approach is feasible in emergency scenario by reducing the latency, raising the task completion Ratio and generally improving the performance of the network