AI-Powered Intrusion Detection for Microservice-Based Architectures

• Author(s): ONI Samuel Boluwatife
• Paper ID: 1707665
• Page: 1064-1071
• Published Date: 31-03-2025
• Published In: Iconic Research And Engineering Journals
• Publisher: IRE Journals
• e-ISSN: 2456-8880
• Volume/Issue: Volume 8 Issue 6 December-2024

Abstract

The software architecture in microservices has revolutionized the software development procedure as we have known it by providing enormous benefits of scalability, flexibility, and durability. This is so because despite the numerous benefits offered by microservices architectures, it brings about the challenge of security challenges since the microservices applications are disperse and interconnects. Signature and anomaly-based methods are known to be impractical in identifying attacks specific to the microservice architecture due to evolving patterns of attack. To overcome with these limitations, this research expands a model of AI based IDS that uses techniques of advanced machine learning and deep learning for the improved and efficient identification of threats. As a result, the proposed system contains several modules: The first is a data collection module which collects network traffic and system logs; the second is a feature extraction module which finds out significant features of attacks; The third is AI-based detection module which classifies the normal behavior and malicious behavior. Supervised learning and unsupervised learning algorithms together with the deep learning models, for instance, CNN and RNNs are used to identify patterns and monitor and detect anomalies in real time [9]. To estimate the performance of the proposed system, benchmark datasets are used for performance evaluation. Evaluation of AI-based approaches using correct classification rate, precision, recall, and f-measure provide insights into the fact that AI-based techniques are far superior in detection when compared to conventional methods for detecting zero-day attacks and advanced persistent threats. Furthermore, the feature extraction seeks to be implemented by deep learning by eliminating the necessity of following conventional rules. Based on the results of this research, it becomes possible to conclude about the effectiveness of AI-based solutions in protecting microservice-based architectures from new generation threats. In a practical way, the study brings value to the field of cybersecurity, outlining an adaptable and sustainable model of detecting intrusion which is fit for cloud-only networks [11]. As the future work, the authors will aim to enhance the model interpretability, decrease the false alarms occurrence, and test the applicability of the proposed method in the real-world use cases related to microservices.

Keywords

AI-powered intrusion detection, Microservice security, Machine learning for cybersecurity, Deep learning-based threat detection, Anomaly detection in cloud computing

Citations

IRE Journals:
ONI Samuel Boluwatife "AI-Powered Intrusion Detection for Microservice-Based Architectures" Iconic Research And Engineering Journals Volume 8 Issue 6 2024 Page 1064-1071

IEEE:
ONI Samuel Boluwatife "AI-Powered Intrusion Detection for Microservice-Based Architectures" Iconic Research And Engineering Journals, 8(6)