A Network Attack Path Discovery Method Based on the Fish Swarm Algorithm

The modern cyber security situation indicates that attacks against enterprises or organizations have a wide range, multiple sources, and unpredictable targets. Compared to other forms of attacks, they are more complex. However, most existing attack path discovery methods have a clear purpose from the beginning: to start from a fixed source point and find attack paths targeting some fixed endpoints. This is inconsistent with the situation where businesses and organizations face arbitrary attacks from unknown sources. In order to better study the impact of random unknown source attacks on enterprise networks, corresponding work in the field of penetration testing should be proposed to simulate and emulate the relevant attack processes, thereby helping defenders find ways to develop forward-looking defense strategies in advance. This article will propose a network attack path discovery method based on fish swarm algorithm. In this method, the defender's network environment is modeled and a Step-length Improvement Algorithm Based on Evaluation Function is introduced. By implementing these measures, the performance of the traditional fish swarm algorithm has been improved. It utilizes the global optimization capability of algorithms, greatly improving the efficiency of discovering feasible solutions for attack paths. Subsequently, by comprehensively considering various factors in path evaluation, it is easier to obtain the optimal solution for the attack path. Experimental verification and analysis indicate that this method is more suitable for research on discovering attack paths with unclear starting and ending nodes.