Year: 2024
Type: article
Abstract: The development of network measurement technologies has greatly increased the speed of network scans, but it also poses risks for the stability of the scanned networks. How to reduce probing traffic and enhance the effectiveness of probing has become a new research issue. In this paper, we utilize network measurement and machine learning techniques, leveraging public interfaces from network mapping platforms to construct a dataset with 44 feature dimensions. By combining the categorical boosting (CatBoost) model with the particle swarm optimization (PSO) algorithm for heuristic optimization, we propose a host port openness prediction model that integrates the PSO algorithm and the CatBoost model. Through comparisons with various machine learning models, the effectiveness of our proposed model was validated. Using this model in network scanning can save approximately 65% of bandwidth on average, effectively reducing the impact on the probed network.
Source: Electronics
Authors: Xiaoxuan Liu, Guozheng Yang, Yi Xie, Xuehu Yan
Institution: National University of Defense Technology
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,728
OpenAlex ID: W4403396449