Congestion-aware Stackelberg pricing game in urban Internet-of-Things networks: A case study

We consider a one-leader multi-follower Stackelberg pricing game in urban Internet-of-Things networks. The leader manages multiple services and sets service prices to maximize profit, while followers seek the most cost-effective services, taking into account service prices, service congestion, and individual locations. Existing approaches often simplify the game process by neglecting followers' locations, but they are not applicable to real-world complex spatial distributions. In this paper, we focus on the congestion-aware Stackelberg pricing game in urban Internet-of-Things networks, using electric vehicle charging as a case study. The game is reformulated as a mathematical program with equilibrium constraints (MPEC) that considers both congestion effects and followers' spatial distributions. To solve the MPEC, we introduce the Segmentation-based Pricing with ITERative Optimization (SPITER) algorithm, which converges to a local maximum. Additionally, optimization techniques are developed to improve the performance of solving SPITER in urban environments. We evaluate the performance of SPITER using extensive experiments with two real-life urban datasets, demonstrating the advantages of our model and illustrating SPITER's effectiveness and convergence.