Visible to the public Policy-Based Security Modelling and Enforcement Approach for Emerging Embedded Architectures

TitlePolicy-Based Security Modelling and Enforcement Approach for Emerging Embedded Architectures
Publication TypeConference Paper
Year of Publication2018
AuthorsHagan, Matthew, Siddiqui, Fahad, Sezer, Sakir
Conference Name2018 31st IEEE International System-on-Chip Conference (SOCC)
KeywordsAccess Control, application threat analysis, Biological system modeling, configurable policy engine, connected vehicles, device life-cycle, DREAD risk assessment model, embedded architectures, embedded security, Embedded systems, Guidelines, Hardware, OEMs, policy, policy update, policy-based governance, policy-based security enforcement, policy-based security modelling, pubcrawl, risk management, Secure by design, Security Modelling, security of data, Security Policies Analysis, Software, STRIDE, STRIDE threat modelling, threat modelling
AbstractComplex embedded systems often contain hard to find vulnerabilities which, when exploited, have potential to cause severe damage to the operating environment and the user. Given that threats and vulnerabilities can exist within any layer of the complex eco-system, OEMs face a major challenge to ensure security throughout the device life-cycle To lower the potential risk and damage that vulnerabilities may cause, OEMs typically perform application threat analysis and security modelling. This process typically provides a high level guideline to solving security problems which can then be implemented during design and development. However, this concept presents issues where new threats or unknown vulnerability has been discovered. To address this issue, we propose a policy-based security modelling approach, which utilises a configurable policy engine to apply new policies that counter serious threats. By utilising this approach, the traditional security modelling approaches can be enhanced and the consequences of a new threat greatly reduced. We present a realistic use case of connected car, applying several attack scenarios. By utilising STRIDE threat modelling and DREAD risk assessment model, adequate policies are derived to protect the car assets. This approach poses advantages over the standard approach, allowing a policy update to counter a new threat, which may have otherwise required a product redesign to alleviate the issue under the traditional approach.
DOI10.1109/SOCC.2018.8618544
Citation Keyhagan_policy-based_2018