On System Cost Optimization for Satellite and HAP Assisted Internet of Things

This paper proposes a satellite and high-altitude platform (HAP) assisted Internet of Things network architecture, where ground users can either compute tasks locally or offload them to HAP or their access satellites for computation. Notably, the access satellite can collaborate with satellites in the candidate set to reduce the task overflow rate. Accordingly, an optimization problem is formulated to minimize the system cost of average energy consumption, computation resource price, and task overflow rate under latency constraints. Given the mixed-integer nonlinear nature of the problem, we decompose it into three subproblems. Specifically, an alternative iteration algorithm is employed for communication and computation resource allocation optimization, while a heuristic algorithm is designed to optimize task offloading decisions. Numerical results demonstrate that the proposed algorithm effectively reduces the system cost compared to benchmark algorithms.