Bioinspired Trust Score Weight Optimization for an Improved Zero Trust Architecture

Abstract

The current digital landscape is highly prone to different cyberattacks that exploit vulnerabilities of an organization. It has posed a significant challenge to traditional security measures, which relied on implicit trust in entities. It resulted in various attacks, including insider attacks. To design a proactive digital posture for organizations, Zero Trust (ZT) has emerged as a better alternative. It is a set of guiding principles and ideologies that frame the way organizations think about trust and access. The literature review reveals that a gap between the conceptual proposals and their actual deployment exists. Existing studies were observed to use a limited number of parameters to evaluate trust. In addition, they use predefined weights for parameters of the trustscore makes the approach unscalable and globally unacceptable. To alleviate this issue, a ZT Architecture (ZTA) is proposed with a detailed incorporation strategy. A comprehensive list of parameters based on six categories is presented to evaluate the trust of the device. This investigation contributes by integrating trustscore and optimizing its weights associated using different metaheuristic optimization techniques. This makes the approach scalable and organization-agnostic. Various performance measures are used to evaluate the efficacy of the optimization techniques. The accumulation of parameters and the final score are published as a publicly available dataset. The differential evolution was found to be the best optimizer for the current scenario. This work aims to make a meaningful contribution to ZT, benefiting both the technological and societal domains.