Computation Offloading with Bus-Mounted Edge Servers Based on Dung Beetle Optimization in Internet of Vehicles

In Internet of Vehicles (IoV), edge computing alleviates congestion in the backbone network and significantly reduces the delay by deploying servers close to users. However, the existing fixed-site and fixed-capacity edge server deployment cannot effectively cope with the spatiotemporal dynamics of IoV end users, i.e., vehicles on road. Therefore, this paper proposes leveraging the inherent mobility of public buses to carry edge servers, and cooperate with fixed-site edge servers to provision elastic joint task offloading service for 1oV end users. We first build a multi-objective optimization model with fixed-site and bus-mounted servers to minimize the user delay, energy consumption, and payment cost. Then, a computation offloading strategy based on the Dung Beetle optimization algorithm is proposed. Taking into account the varying tolerance for delay among different tasks, a preprocessing method is adopted to set the offloading ratio for each task, thereby reducing the search space and accelerating the convergence speed. Furthermore, the maximum acceptable delay constraint is transformed into an unconstrained problem through the penalty function method. Experimental results verify that the proposed computation offloading algorithm outperforms various classical intelligent optimization algorithms in terms of the overall cost.