Logo des Repositoriums

Auflistung nach Autor:in "Ben Zid, Ines"

1 - 2 von 2
  • Konferenzbeitrag
    The application of Articial Intelligence for Cyber Security in Industry 4.0
    (INFORMATIK 2019: 50 Jahre Gesellschaft für Informatik – Informatik für Gesellschaft (Workshop-Beiträge), 2019) Ben Zid, Ines; Parekh, Mithil; Waedt, Karl; Lou, Xinxin
    The use of Artificial Intelligence (AI) in different domains is continuously growing. In particular for cybersecurity, we can see the implementations of AI solutions, e.g. machine learning, in a wide range of applications from various domains. While some consider this step as risk for cybersecurity, others agree that it is in fact a solution to many issues as well. This leads to a higher necessity of having a right understanding as well as handling of cybersecurity controls that enforce meeting domain, project and application specific security targets. This implies that more efforts and resources have to be focused and invested towards cybersecurity. One reason for this is that attackers (threat agents) may integrate AI based algorithms and AI based evaluation of data, which forces the security staff to respond at a similar level. Thus, we are considering AI as a potential solution for satisfying a set of rising needs and objectives. In this paper, we present the concept for merging and integration of these three major domains and applications. Also, we detail the relevant motivations, requirements and challenges to be considered when coming to such combination.
  • Konferenzbeitrag
    Operational Security Modeling and Analysis for IACS
    (INFORMATIK 2019: 50 Jahre Gesellschaft für Informatik – Informatik für Gesellschaft (Workshop-Beiträge), 2019) Gao, Yuan; Ben Zid, Ines; Lou, Xinxin; Parekh, Mithil
    Security Certifications based on international standards, like ISO 27000 and IEC 62443 series, are strongly favored by industrial manufactures and (critical) facility owners. However, comparing to mature safety certification procedures, there is only a small portion of security certifications available on the market for the booming Industry 4.0 solutions and IoT/IIoT products. The major challenge is how to define a practical working scope, which is compatible with frequent system updates as well as creations of new systems by coupling supplier services. Meanwhile, the potential security impacts should be quantitatively predictable since some of them are tolerable, which are different from most of safety constraints. Thus, in this paper, we proposed an operational security model, which intends to support monitoring and analysis on a dynamically running system. It was extended from the 3-domains security model we proposed in previous work by introducing run-time perspectives and procedures. In addition, cooperating with the security in design concept, the proposed operational procedures were developed following the guidance of the security standard series IEC 62443. For addressing the external threats, Open Source Intelligence (OSINT) were involved to query whether some confidential information, like user-credentials and system vulnerabilities are already collected and publicly known to adversaries. The introduction of OSINT can support more transparent risk assessment approaches. As the conclusion, with the operational security model, we proposed a hybrid approach which consists of security certifications and continuous monitoring/consulting to solve the current challenge.