Abstract—The rapid growth of intelligent connected vehicles fosters the development of Internet-of-Vehicle (IoV) applications. Due to the complex traffic environments, integrated sensing, communication and computation (ISCC) technology would be seen as an efficient paradigm to support a plethora of IoV applications, which causes such a surge of computation-intensive tasks that the task computation with a single server node cannot keep energy-efficient under the strict delay constraints. Inspired by the idea of distributed computing, a scheme of distributed edge computing-based task offloading for ISCC is designed and various assisted edge nodes (AENs) are introduced in a stochastic geometry approach to alleviate the dilemma of energy consumption. After analyzing the proposed joint optimization problem for energy minimization, a double-iteration joint optimization algorithm (DIJOA) is developed to derive the solution. The results of the performance evaluation not only verify the plausibility of the proposed model, but also indicate that the introduction of AENs can significantly reduce system energy consumption and the proposed algorithm outperforms other schemes by over 2.5% in energy consumption, which corroborates the superiority of the proposed scheme through numerical simulations. 

Fig. 1. The architecture of distributed edge computing in vehicular networks.