The surge in computing demands of onboard devices in vehicles has necessitated the adoption of mobile edge computing (MEC) to cater to their computational and storage needs. This paper presents a task offloading strategy for mobile edge computing based on collaborative roadside parking cooperation, leveraging idle computing resources in roadside vehicles. The proposed method establishes resource sharing and mutual utilization among roadside vehicles, roadside units (RSUs), and cloud servers, transforming the computing task offloading problem into a constrained optimization challenge. To address the complexity of this optimization problem, a novel Hybrid Algorithm based on the Hill-Climbing and Genetic Algorithm (HHGA) is proposed, combined with the powerful Simulated Annealing (SA) algorithm. The HHGA-SA Algorithm integrates the advantages of both HHGA and SA to efficiently explore the solution space and optimize task execution with reduced delay and energy consumption. The HHGA component of the algorithm utilizes the strengths of Genetic Algorithm and Hill-Climbing. The Genetic Algorithm enables global exploration, identifying potential optimal solutions, while Hill-Climbing refines the solutions locally to improve their quality. By harnessing the synergies between these techniques, the HHGA-SA Algorithm navigates the multi-constraint landscape effectively, producing robust and high-quality solutions for task offloading. To evaluate the efficacy of the proposed approach, extensive simulations are conducted in a realistic roadside parking cooperation-based Mobile Edge Computing scenario. Comparative analyses with standard Genetic Algorithms and Hill-Climbing demonstrate the superiority of the HHGA-SA Algorithm, showcasing substantial enhancements in task execution efficiency and energy utilization. The study highlights the potential of leveraging idle computing resources in roadside parking vehicles to enhance Mobile Edge Computing capabilities. The collaborative approach facilitated by the HHGA-SA Algorithm fosters efficient task offloading among roadside vehicles, RSUs, and cloud servers, elevating overall system performance.

车辆车载设备的计算需求激增，需要采用移动边缘计算 (MEC) 来满足其计算和存储需求。本文提出了一种基于协同路边停车协作的移动边缘计算任务卸载策略，利用路边车辆的闲置计算资源。该方法在路边车辆、路边单元（RSU）和云服务器之间建立资源共享和相互利用，将计算任务卸载问题转化为约束优化挑战。为了解决该优化问题的复杂性，提出了一种基于爬山算法和遗传算法（HHGA）并结合强大的模拟退火（SA）算法的新型混合算法。HHGA-SA算法融合了HHGA和SA的优点，可以有效地探索解决方案空间并优化任务执行，同时减少延迟和能耗。该算法的 HHGA 组件利用了遗传算法和爬山算法的优点。遗传算法能够进行全局探索，识别潜在的最佳解决方案，而爬山算法则在本地改进解决方案以提高其质量。通过利用这些技术之间的协同作用，HHGA-SA 算法可以有效地应对多约束环境，为任务卸载提供强大且高质量的解决方案。为了评估所提出方法的有效性，在基于现实路边停车合作的移动边缘计算场景中进行了广泛的模拟。与标准遗传算法和爬山算法的比较分析证明了 HHGA-SA 算法的优越性，展示了任务执行效率和能量利用率的显着增强。该研究强调了利用路边停车车辆的闲置计算资源来增强移动边缘计算能力的潜力。HHGA-SA 算法促进的协作方法促进了路边车辆、RSU 和云服务器之间的高效任务卸载，从而提高了整体系统性能。