Game-Theoretic Optimization for Scalable Task Offloading and Dynamic Pricing in Heterogeneous Edge Networks

The rapid proliferation of latency-sensitive applications, such as autonomous driving, augmented reality, and smart manufacturing, has exposed the limitations of traditional cloud computing, catalyzing a shift toward multi-access edge computing (MEC). While MEC mitigates latency by migrating computational tasks to edge servers (ESs) located closer to end-users, the finite nature of edge resources gives rise to complex policy interactions between mobile users and ESs. Game theory, particularly the Stackelberg model, is widely employed to model these interactions. However, existing methods are often insufficient for dynamic multi-user, multi-task scenarios where discrete offloading strategies and continuous pricing schemes are intricately coupled. To address this gap, we propose a joint task offloading and resource pricing optimization framework formulated as a Stackelberg game. In this framework, we develop an Optimal Pricing via Interior-point Optimization (OPIP) algorithm to efficiently determine ES pricing and a Hybrid Greedy with 2-Opt Offloading (HG2O) algorithm to manage the NP-hard offloading problem for users. Theoretical analysis confirms the existence and uniqueness of a Stackelberg equilibrium. Furthermore, simulation results demonstrate that our proposed approach significantly reduces user costs and enhances ES revenue compared to baseline algorithms, affirming the model's practical effectiveness, robustness, and scalability for MEC resource management.