Survivability Analysis of LEO Satellite Networks Based on Network Utility

The LEO satellite network is threatened by accidental faults, malicious attacks and other factors during its operation. This paper proposed a model of survivability evaluation based on network utility. First, calculate the communication performance indicators of delay, congestion rate and throughput of LEO satellite networks by different types of queue theory birth and death model; then, combined these performance indicators with reliability and cost planning to construct network utility function which is used for evaluating the survivability of LEO satellite networks under natural failure, random attack, deliberate attack and saturation attack modes. The case study shows that: the greater the arrival rate of natural faults, the shorter the time required to descend to the network utility threshold, the network is more vulnerable to intentional attacks; the survivability can't be enhanced by adding buffers; the importance of nodes and links is related to the proportion of real-time and non-real-time services and the traffic of different satellite coverage areas.