AZ – TAC Adaptive Zero Trust Access Control with Continuous Authentication for Secured Cloud Environments

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Identity-centric access approaches are becoming fundamental to enable the management of highly dynamic users, services, and workloads in cloud computing environments. However, traditional access control systems are predominantly focused on static tokens, one-time authentication and coarse-grained authorization, which is not ideal to combat modern-day cloud attacks including leaking of credentials, abusing non-human identity, replay attack and lateral movement. In this paper we present an Adaptive Zero -Trust Cloud Access Security Protocol that is based on the concepts of continuous authentication and risk aware authorization applied to human and non-human identities in a cloud environment. The protocol leverages the Zero Trust model to ensure that each access request is cryptographically authenticated, corroborated with contextual security signals, and mathematically verified to present an acceptable risk. In this proposed work, a secure set of techniques for limiting disclosure of personal data so that access and identity are separate, while use of Proof-of-Possession tokens, Mutual Authentication and short-lived credentials mitigate impersonation and token misuse. This framework also ties together dynamic authorization with least-privilege enforcement, automatic credential revocation and optimized key rotation to minimize the effects of identity compromise. For accountability purposes all security-critical events are being recorded into an append-only audit log which is tamper-evident. Formal security of the protocol in the Random Oracle Model is presented, which shows that it can withstand impersonation, replay, man-in-the-middle and credential theft attacks. An informal security analysis also supports its resilience to insider threats and non-human identity violation. The theoretical foundations of strong cryptography combined with flexible risk evaluation and optimization driven approach allows the proposed approach to deliver improved cloud access security without sacrificing scalability and performance, thus being suitable even in large-scale and dynamic cloud settings.

Article activity feed