Smart Infrastructure Deployment Strategy Based on Hybrid Networking: Toward City-Scale Vehicle-Cloud Computation

Vehicle-cloud computation (VCC) is vital for urban intelligent transportation systems (ITS). The realization of VCC relies on the proper deployment and application of intelligent infrastructures, such as Roadside Units (RSUs) for communication and Edge Computing Servers (ECSs) for computation. However, the time-varying mobility characteristics of individual vehicles and uncertain service demands pose challenges for effective infrastructure deployment. This paper proposes a two-stage deployment optimization model based on real vehicle trajectory data to enhance the connected vehicles-RSUs-ECSs chain for information propagation and task offloading. In the first stage, a hybrid communication networking model is developed to design an information propagation tree, leading to a tri-objective optimization problem for RSU deployment, solved using NSGA-II to generate Pareto frontier solutions. The second stage focuses on ECS deployment through a bi-objective optimization model, incorporating chance-constrained planning to handle variability of RSU computation demand caused by dynamic vehicle distributions. The <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\epsilon $</tex-math> </inline-formula>-constraint method is applied for hierarchical optimization, and an enhanced tabu search (ETS) algorithm is employed to efficiently determine optimal solutions. Subsequently, an online task reallocation and relaying mechanism based on sequential decision-making is developed to enhance system resilience under extreme scenarios. Validation on the Anhui Xuancheng dataset demonstrates that the proposed method significantly improves communication service capacity and balances computational loads. This research contributes to the continuous and optimal deployment of smart infrastructures to advance the application of ITS for sustainable urban mobility.