RLRTA: Reinforcement Learning based Real-Time Task Offloading Algorithm for IoV Applications

The growing number of individual vehicles and intelligent transportation systems has accelerated the development of Internet of Vehicles (IoV). Intelligent vehicles, like autonomous vehicles, have become increasingly popular. The various sensors they are equipped with grant vehicles powerful computing, communication, and storage capabilities. However, with the continuous development of in-vehicle applications, the computing power of a single vehicle is no longer sufficient to meet the growing demands. How to efficiently and reasonably allocate tasks has become an urgent problem to be solved. In IoV, vehicles are dynamic, therefore, as agents, vehicles need to make decisions in a constantly changing environment to optimally complete the assigned tasks. To address this issue, this paper proposes a task offloading strategy that combines reinforcement learning with vehicular edge computing. By providing computing resources at the network edge and optimizing the task assignment strategy, the computing power and response speed of the IoV are improved, and the delay and energy consumption are reduced, thereby meeting the needs of diverse applications.