Research on Penetration Testing Path Planning Method Based on Improved Potential Field Ant Colony Algorithm

This study proposes a penetration testing path planning method based on an improved potential field ant colony algorithm. First, the network topology is modeled as a graph structure, with each communication link assigned an attack cost weight. To overcome the issue of traditional artificial potential fields easily falling into local optima, this paper introduces a repulsion function that dynamically adjusts the repulsion intensity, weakening it as nodes approach the target system, thereby enhancing global search capability. The ant colony optimization algorithm is then employed, which iteratively optimizes path selection by simulating the pheromone release mechanism of ants. By combining the local search capability of the artificial potential field with the global search capability of the ant colony algorithm, the proposed improved algorithm demonstrates superior performance in terms of computation time, convergence speed, and path planning success rate. Experimental results in different network environments show that the improved algorithm’s performance increasingly surpasses that of other algorithms as network size grows, offering higher path planning success rates and faster convergence speeds. This study provides an efficient and reliable method for path planning in network security penetration testing.