Evolutionary Intrusion Detection Strategy Under Zero Trust Architecture

In today’s increasingly complex and dynamic cyber threat environment, Zero Trust Architecture (ZTA) has emerged as a promising solution to address the limitations of traditional intrusion detection methods. While Intrusion Detection Systems (IDS) are essential for safeguarding organizational information assets, traditional methods have the risk of exposing security policies by collecting and utilizing alarm data, potentially revealing attack paths to adversaries. To overcome this challenge, we propose a novel intrusion detection strategy based on ZTA, emphasizing the protection of alarm data privacy. Additionally, we introduce an adaptive sparse connective evolutionary neural architecture search (ASCE-NAS) framework, which enables the automatic evolution of intrusion detection model structures to enhance adaptability and performance in dynamic environments. Notably, ASCE-NAS can also be beneficial for integrated sensing and computing chips and systems, contributing to a more secure and efficient cybersecurity framework to effectively combat evolving threats and attack methods.