TS-Hybrid-SB: A Two-Stage Hybrid Optimization Framework with Stackelberg–Bandit Pricing for Resource Scheduling in MEC-Enabled IoT Systems

Abstract

The highly intensive proliferation of Internet of Things (IoT) devices and latency-sensitive applications has compounded the need for efficient task offloading and resource management in Mobile Edge Computing (MEC) settings. Conventional heuristic and metaheuristic methods, including Genetic Algorithm (GA), Simulated Annealing (SA), and Particle Swarm Optimization (PSO), are unable to provide adaptability and equity in real-time and dynamic network environments. This paper suggests a new framework called Two-stage Hybrid Optimization and Stackelberg-bandit Pricing (TS-Hybrid-SB) to address these challenges. The suggested model unites Confidence-Aware PSO (CA-PSO) to explore the world, convex optimization to localize refinements, and the Stackelberg-Bandit pricing system to adapt the economic equilibrium between MEC servers (leaders) and IoT devices (followers). The simulation outcomes using MATLAB 2013a indicate that TS-Hybrid-SB achieves a minimum latency reduction of up to 40%, consumes 45% less energy, and has a 10% higher fairness index compared to base techniques. Moreover, it offers 14 percent better utility of MEC and reduces the convergence steps by half, which demonstrates its scalability, energy efficiency, and economic stability in MEC-enabled IoT systems.