Home // UBICOMM 2023, The Seventeenth International Conference on Mobile Ubiquitous Computing, Systems, Services and Technologies // View article

Threat Detection based on System Credibility by Logging Analysis and Visualization

Authors:
Wei Qiao
Youjun Bu
Yao Chen
Xiaoxiao Jiang
Bingbing Jiang

Keywords: cybersecurity, threat detection, credibility.

Abstract:
The novel theory of Endogenous Safety and Security was proposed from a system architecture perspective is striving to address the current complex cybersecurity threats dilemma. It utilizes multiple heterogeneous and functionally equivalent systems (called mimic systems) to detect threats because different implementations have different vulnerabilities and are dynamically scheduled on some feedback strategies, making it impossible for a single attack to simultaneously compromise all of these implementations. The threat detection heavily relies on the adjudication of outputs from multiple heterogeneous and functionally equivalent systems because it is possible to adjust the outcomes of the majority of compromised systems as correct. Therefore, the credibility of the adjudication should be evaluated for verifying the trustworthiness of the mimic system, but no research is currently available on the system credibility in the mimicry environment. In this paper, we propose a logging analysis algorithms to evaluate the credibility of the adjudication which is related to each single system's disturbance event history, disturbance factors, disturbance number and one-time runtime duration. The experiments also prove the positive performance of the proposed algorithm. The lower the credibility, the higher the possibility of the system being compromised.

Pages: 28 to 33

Copyright: Copyright (c) IARIA, 2023

Publication date: September 25, 2023

Published in: conference

ISSN: 2308-4278

ISBN: 978-1-68558-106-0

Location: Porto, Portugal

Dates: from September 25, 2023 to September 29, 2023