Modeling and Analysis of Cascading Failures in LEO Satellite Networks
Le Zhang, Ye Du, Zhengyang Sun
Abstract
LEO satellite networks are prone to trigger cascading failures and cause global network outage after being cyber attacked. This article introduces the cascading failure theory into the field of LEO satellite network for the first time, which provides a new perspective for enhancing the network invulnerability in the future. In the proposed model, satellites and gateways have different working and failure states, and jointly determine the operating state of the network. The buffer capacity and maximum processing ability of nodes are dependent on the initial load present in hotspots. The real-time load is affected by the actual routing scheme. Too much unprocessed load will cause the degradation of processing ability. A cascading algorithm to implement the model is proposed and the feasibility is analyzed. Case constellation analysis reveals that satellites in hotspots are more susceptible to triggering cascading failures. The congestion-sensitive routing can enhance network invulnerability. The false routing information attacks will cause more serious cascading processes than black hole attacks.