Energy Efficient Joint Computing and Communication Resource Management in Satellite-Terrestrial Integrated Networks

Multi-access edge computing (MEC) assisted satellite-terrestrial integrated network (STIN) is promising in offering ubiquitous network coverage and swift computing services for Earth observation-based applications. Given the constraints of limited onboard resources and the substantial energy costs associated with complex computations and transmissions, the imperative of minimizing energy consumption emerges as a pivotal challenge in MEC-assisted STIN. This paper investigates the energy minimization problem, subject to the resource and latency constraints of terrestrial base stations and Low-Earth-Orbit satellites. To solve the formulated NP-hard mixed-integer nonlinear programming problem efficiently, we first conduct problem analysis to decompose the original problem into two coupling sub-problems. Then we propose a genetic algorithm based-joint computing and communication resource allocation (GA-CCRA) scheme to solve the sub-problems. Our simulation results show the superiority and effectiveness of GA-CCRA in terms of energy efficiency in the MEC-assisted STIN.