Dingo-optimization-based task-offloading algorithm in multihop V2V/V2I-enabled networks.

In the Vehicle-to-Vehicle- (V2V) and Vehicle-to-Infrastructure- (V2I) enabled edge computing networks, the task vehicle may fail to offload tasks to the nodes outside the one-hop communication with high computing resources, resulting in longer offloading time. Therefore, in this paper, we first construct a multihop V2V/V2I communications-enabled edge computing architecture, which innovatively adopts vehicles and roadside units (RSUs) task offloading jointly to expand the communication resources of the task vehicle. Then, for the offloading node selection strategy, we create a vehicle adjacency table and propose a quickly depth-first search-based scheme to search the optimal multihop path. Finally, we develop an optimal offloading scheme based on the discrete dingo optimization algorithm (DDOA) to solve the task-processing-time minimization problem, which can converge quickly and achieve lower task offloading latency. Each dingo in DDOA is designed to choose one of the group attack, persecution, and scavenger strategies to search the solution space and accelerate the convergence of the DDOA by a survival rule. The numerical experimental results reveal that our proposed algorithm can outperform the other algorithms and reduce the delay time by 80% compared with the local scheme.