Abstract:In the multi-layer heterogeneous satellite Internet, time-varying network topology, dynamic link quality, unstable Satellite nodes and the massive emergence of high-dimensional data bring huge challenges to network operation, maintenance and management. Artificial intelligence technology is considered an effective solution to address the above challenges due to its ability to learn from massive historical data or learn from interactive feedback with dynamic environments. Focus on investigating the current research status of artificial intelligence technology in the field of satellite network operation, maintenance and management in three aspects: network resource management, business traffic management and network status management. On this basis, we will conduct an in-depth analysis of the shortcomings of existing methods and discuss the issues worthy of attention in the future. research direction.
Keywords:Satellite Internet; artificial intelligence; network operation, maintenance and management; resource allocation; anomaly detection
1 Introduction
Satellite Internet has the advantage of seamless global coverage and has been included in the scope of new infrastructure in my country. The operation, maintenance and management system of the satellite Internet is responsible for monitoring, adjusting and controlling its operation process. It is the core of the unified management of global resources and network operating conditions, and directly determines whether the satellite Internet can operate stably and meet user business needs. Traditional satellite Internet operation and maintenance systems are mostly based on rules, statistics or modeling optimization. These methods are usually based on expert knowledge or accurate modeling of the network environment. They will encounter the following challenges in satellite Internet.
(1) Complex and heterogeneous models: Space, space and ground integrated networks have become the development trend of the satellite Internet. The network structure is becoming increasingly complex, satellites and user terminals are becoming larger and larger, and business needs have expanded from mobile communication services to broadband Internet and Internet of Things services, giving the network Deterministic modeling brings challenges.
(2) Dynamic time-varying environment: High-speed moving multi-layer space-based platforms, complex spatial frequency interference and time-varying weather factors such as rainfall lead to dynamic changes in link performance. The trend of large-scale miniaturization of satellites leads to an increase in their failure probability. The increase in uncertainty causes the workload of operation and maintenance personnel to locate faults to increase exponentially. At the same time, satellite failures and changes in channel quality lead to dynamic changes in limited network resources, and it is difficult for traditional methods to efficiently adjust resource allocation plans, resulting in