With the development of computing capability onboard, satellite-terrestrial integrated network presents characteristics such as large scale, complex topology, and heterogeneous and diverse resources, making it difficult to accurately adapt to the resource requirements of different services. This article analyses the diverse and heterogeneous computing and networking resources in satellite-terrestrial integrated network, and presents an intelligent resource management and control architecture. This architecture aims to achieve intelligent collaborative management of multi-dimensional and multi-domain computing and networking resources in space and ground, and to provide resources for diverse scenarios as needed. Based on the proposed architecture, a knowledge space-based resource management mechanism is proposed to achieve deep integration of computing, data, algorithms, and networks, and to improve the effectiveness of resource orchestration. Meanwhile, a highly dynamic end-to-end network slicing management mechanism is presented to adapt to the differentiated needs and to provide on-demand network services for users. Finally, the experimental results showed that a dynamic deployment policy for the satellite-terrestrial integrated network can be generated in just 1.3 seconds using the greedy algorithm to meet the computing and networking resource requirements of the service scenario on a scale of 20 satellites, which verified the rationality and effectiveness of the intelligent resource management architecture.
|