Towards Embedded Artificial Intelligence Based Security for Computer Systems.
ABSTRACT This paper presents experiments using Artificial Intelligence (AI) algorithms for online monitoring of integrated computer systems, including System-on-Chip based embedded systems. This new framework introduces an AI-lead infrastructure that is intended to operate in parallel with conventional monitoring and diagnosis techniques. Specifically, an initial application is presented, where each of the systempsilas software tasks are characterised online during their execution by a combination of novel hardware monitoring circuits and background software. These characteristics then stimulate a Self-Organising Map based classifier which is used to detect abnormal system behaviour, as caused by failure and malicious tampering including viruses. The approach provides a system-level perspective and is shown to detect subtle anomalies.
- [Show abstract] [Hide abstract]
ABSTRACT: This paper presents a system level framework for system-on-chip (SoC) based embedded devices that may include adaptive and reconfigurable elements. Current development support and debugging solutions are highly dependant on off-line post-mortem style inspection, and even those that utilise tracing for real-time and schedule-critical systems rely on external development tools and environments. This new framework introduces an AI-lead infrastructure that has the potential to reduce much of the development effort while complementing existing debugging circuits. Specifically this paper investigates how to use a Kohonen self-organising map (SOM) as a classifier, and shows a preliminary investigation into how to determine the quality of a map after training. This classifier is a first step in diagnosing failure, degradation and anomalies (i.e. provides condition monitoring) in an embedded system from a system level point of view, and in the larger task of self-diagnosis of an embedded system.Adaptive Hardware and Systems, 2008. AHS '08. NASA/ESA Conference on; 07/2008
Conference Paper: Adaptive Online Profiling Hardware for ICmetrics Based Security[Show abstract] [Hide abstract]
ABSTRACT: Monitoring circuitry is presented that extracts properties and features from a complex system based on a system-on-chip based device to support ICmetrics, a novel security concept that aims to uniquely identify and secure an embedded system based on its own behavioural identity. The circuits utilise a novel approach to profiling the instruction fetches and data accesses associated with each of the systempsilas component software tasks so that a representation of their address distributions can be generated online when required. By using profiling circuits with adaptive allocation of counters and by exploiting existing debugging infrastructure circuits, the overall resource requirement is kept sufficiently low for real-world applications. Experimental results are provided to illustrate the nature of features that can be obtained.Adaptive Hardware and Systems, 2008. AHS '08. NASA/ESA Conference on; 07/2008
- [Show abstract] [Hide abstract]
ABSTRACT: Advanced on-chip debug support can help overcome the challenges of developing real-time embedded systems driven by complex SoCs, making development support a decisive selection factor. One novel approach supports the on-time development of high-quality complex systems without greatly increasing the required design and production resourcesComputer 05/2007; · 1.44 Impact Factor