Multi-Task Offloading Based on Optimal Stopping Theory in Edge Computing Empowered Internet of Vehicles.

Area of Science:

• Computer Science
• Electrical Engineering
• Network Engineering

Background:

• Vehicular edge computing (VEC) is an emerging paradigm integrating edge computing into the Internet of Vehicles (IoV).
• VEC enables low-latency services by deploying resources closer to IoV users, such as roadside units (RSUs).
• Efficient sequential task offloading for mobile nodes to Mobile Edge Computing (MEC) servers presents a significant challenge.

Purpose of the Study:

• To propose novel time-optimized, multi-task-offloading models for VEC environments.
• To leverage Optimal Stopping Theory (OST) to maximize offloading to optimal MEC servers and minimize total offloading delay.

Main Methods:

• Development of two OST-based models for task offloading in VEC.
• One model maximizes the probability of selecting optimal MEC servers.
• A second model minimizes total offloading delay under uniform server utilization.
• Experimental evaluation using simulated and real-world datasets, including sensitivity analysis.

Main Results:

• The proposed OST-based offloading models are efficiently implementable on mobile nodes.
• Significant reduction in the total expected processing time for vehicular tasks.
• Demonstrated effectiveness compared to existing OST models in various scenarios.

Conclusions:

• The proposed OST models offer an effective solution for optimizing task offloading in VEC.
• These models contribute to improving the efficiency and performance of MEC services for IoV.
• The approach successfully addresses the challenge of sequential task offloading in dynamic vehicular environments.