Application and Performance Evaluation of Resource Pool Architecture in Satellite Edge Computing

Abstract

Satellites will play a vital role in the future of the global Internet of Things (IoT); however, the resource shortage is the biggest limiting factor in the regional task of massiveequipment in the IoT for satellite service. Compared with the traditional isolated mode of satellite resources, the current research aims to realize resource sharing through satellite cooperation in satellite edge computing, to solve the problems of limited resources and low service quality of a single satellite. We propose a satellite resource pool architecture-oriented regional task in satellite edge computing. Different from fixed servers in ground systems, the satellite orbital motion brings challenges to the construction of the satellite resource pool. After the capacity planning of the satellite resource pool for regional tasks is given, an algorithm based on search matching is proposed to solve the dynamic satellite selection problem. A ground semi-physical simulation system is built to perform experiments and evaluate the performance of three modes of satellite resource sharing: isolated mode, cooperative mode, and pooled mode. The results show that the pooled mode, compared with the isolated mode, improves the task success rate by 19.52%, and at the same time increases network resources and energy consumption in the same scenario. Compared with the cooperation mode, the performance of task success rate and resource utilization rate is close to that of the pooled mode, but it has more advantages in response time and load balancing of satellite resources. This shows that in the IoT, the resource pool is of great benefit as it improves the task response time and improves the load balance of satellite resources without degrading the performance, which makes sense in task-demanding scenarios.