Toward a Performance-Based Trustworthy Edge-Cloud Continuum

Abstract

The Edge-Cloud Continuum refers to the dynamic provisioning of distributed computing and network resources that can be scaled to support the creation of secure, resource-efficient, and decentralised digital ecosystems, which also support federated topologies for collaborating and sharing resources. Trusted interaction and orchestration of distributed edge-cloud resources are the fundamental principles of distributed network infrastructure and service provisioning. The zero trust architecture (ZTA) paradigm is gaining momentum based on being able to ensure trusted and secure interaction for edge-cloud networks. However, ZTA's strict authentication policy mandates devices to be authenticated for every session, leading to significant overhead for resource-constrained devices engaged in multiple sessions. To address this challenge, this paper proposes a ZTA that integrates a performance-based trust assessment mechanism, allowing a higher number of consecutive sessions without the need for costly authentication/authorisation while preserving system integrity. Reputation, viewed through the performance lens, is a metric to gauge a node's trustworthiness, considering its past behaviour and interactions. The proposed trust assessment mechanism is evaluated for its feasibility within our conceptualised ZTA for edge computing environments with limited resources, and simulation results demonstrate the practicality of utilising this technique in zero trust environments.