Optimal valid path prediction method for inter-domain networks considering commercial relationships*

In the security research of large-scale inter-domain networks, it is often necessary to consider network state changes under the condition of failure or attack. In this case, the connection between ASes often switches between disconnection and recovery frequently, requiring frequent prediction of the AS-level path. Previous AS-level path prediction methods generally require detailed network information data and are complicated to calculate, which is difficult to meet the needs of such large-scale dynamic computation. In this paper, an optimal valid path prediction method for inter-domain networks considering commercial relations is proposed. This method predicts the optimal valid path according to the commercial relationships between ASes, which requires less network information data and low computational complexity. Through experiments, it is found that compared with the shortest path method commonly used in the security simulation of large-scale inter-domain networks, the prediction results are closer to the actual path of the network and can better reflect the actual path situation of the inter-domain network with lower computational complexity.