Аннотация
This research study introduces a new concept of Penetration Testing-Aware Intrusion Detection System (IDS) for wireless networks in urban digital environments. By integrating edge computing and machine learning with the proposed framework, it improves the security controls specific to the threats in urban wireless environments. An analysis was conducted using the UNSW-NB15 dataset for testing and validating two machine learning models of logistic regression and gradient boosting to assess their accuracy in real-time network intrusion detection. Comparing the results obtained in applying the gradient boosting and logistic regression, the first model described an accuracy of 95%, a precision-recall value of 0.97, and an AUC score of 0.94, whereas the second model described an accuracy of 88%, a precision-recall value of 0.92, and an AUC score of 0.81. The IDS also proactively passed a thorough vulnerability analysis and penetration testing and proved the possibility of the identification of a high percentage of attack scenarios with low false positives. The results of the present research highlight the importance of using machine learning approaches coupled with active vulnerability scanning to enhance the protection of wireless networks in urban digital ecosystems.