Publication

Abstract :

With the rapidly shifting landscape in cybersecurity, cyber threats have made the old perimeter-based models obscurate. With the increasing proliferation of cloud-native applications and microservice architectures into organizations, a sound security model shall be a basic necessity that shall withstand these threats. This paper introduces a Zero Trust Security Model for web applications in Kubernetes clusters, utilizing key technologies such as JSON Web Token (JWT) authentication and mutual Transport Layer Security (mTLS). Unlike traditional models that rely on predefined trust zones, this approach enforces continuous identity verification, dynamic access control, and fine-grained micro-segmentation to secure sensitive data. The model incorporates Zero Trust principles, including encrypted communications, strict user and device authentication, and real-time threat detection. It provides robust policy enforcement to protect APIs, containers, databases, and other critical resources in a microservices environment. The persistent observation of user activity, along with the implementation of machine learning algorithms for the identification of threats, improves the system's capability to recognize and address potential incidents in real-time. Our experimental validation shows that this model effectively minimizes lateral movement and unauthorized access, significantly reducing attack surfaces. The proposed Zero Trust architecture offers a scalable, secure, and resilient framework suited to modern containerized applications, ensuring data integrity and security.