Buffer-state Aware Task Offloading in Edge Networks With Task Splitting for IoV
Abbas Yekanlou, Ahmed I. Salameh, Jun Cai · 2023
The expansion of internet of vehicles (IoV) to host new applications for end users (EUs), such as extended reality (XR) and in-vehicle streaming services, is restrained by the computing capacity available at the EUs locally. To accelerate IoV expansion and abate the computation capacity limitation at the EU side, edge computing (EC) has exploded in recent years. In this work, we devise a workframe to optimize task offloading (TO) and results caching from an EU to an edge network made of two EC nodes with the objective of queuing delay risk assessment to minimize task dropping and maximize earned task credit by the primary EC node. We first formulate an integer non-linear programming (INLP) problem. Then, an algorithm based on the genetics algorithm (GA) is proposed to solve the problem. Our results show that the proposed algorithm achieves the best performance in terms of average task execution time, dropping rate, and earned credit by the primary EC node compared to traditional task offloading methods.