Adaptive Hierarchy Based Non-Cooperative Satellite Network Topology Detection

Detecting the topology of non-cooperative satellite networks is crucial for identifying enemy critical nodes and developing effective offensive and defensive strategies. Traditional methods for detecting the topology of non-cooperative satellite networks use a maximum likelihood-based topology estimation approach. This method prevents us from getting complete satellite network link connection information, making it difficult to detect enemy satellite networks. In this type of detection, each network internal node detected is a fork node (a node at which data can be transmitted downwards via at least two links). Other nodes may exist between adjacent forked nodes, which are referred to as hidden nodes. Maximum likelihood topology estimation methods cannot detect hidden nodes, and this method relies on using variance of network performance parameters during topology estimation. At relatively long distances, the number of satellite routing hops is fewer than that of ground routing, and the variance of measured values for some leaf nodes is very small, making it inconvenient to use maximum likelihood estimation calculations. To tackle these issues, we introduce a topology estimation algorithm based on hierarchical approach. Our approach calculates the number of layers from internal nodes to source nodes and supplements the hidden nodes of the network structure. Performance simulations show our algorithm achieves over 85% accuracy in topology detection and is robust under three different network loads.