PETRAS National Centre of Excellence for IoT Systems Cybersecurity is a consortium of eleven leading UK universities which will work together over the next three years to explore critical issues in privacy, ethics, trust, reliability, acceptability, and security.
Knowledge-driven Cyber-Physical Security at the Edge
Context Aware Knowledge-driven Cyber-Physical Security at the Edge for Smart Homes
Smart devices are heterogeneous where each of them has a different set of capabilities in terms of sensing and actuation. To unlock the true potential of self-adaptive smart spaces, these devices should work and collaborate by sharing their capabilities to achieve a given goal. These smart IoT devices should automatically evolve, depending on the needs of users, and adapt to the new contexts/conditions. While smart spaces are advantageous and desirable in many ways, they may be hacked, exposing privacy and security, or rendering the entire area a hostile environment in which ordinary tasks are impossible to do. Therefore, securing smart spaces can be challenging due to device heterogeneity, continuous changes of context, and limited device resources.
This project aims to develop techniques that can dynamically configure a given smart space (i.e., self-adapting) to achieve a goal (i.e., ensuring security and safety of the cyber-physical system) without needing of cloud services (i.e., edge computing). To achieve this, we adopt Monitor-Analyze-Plan-Execute-Knowledge (MAPE-k) method. Some of the investigations we need to carry out are as follows. First, we need to capture information that MAPE-k requires. Some key pieces of static information are smart device capabilities, limitations. For example, devices such as smart vacuum cleaners can move. Another example is webcams which have the capability of taking images. Other important information needs to be continuously updated (e.g., device locations, weather, environment conditions, calendar information). Some updates could be simple as downloading a calendar, whereas others require data analytics (e.g., detect a window open by analysing temperature variations near the window). We expect this knowledge base to be modelled around well-known ontologies (e.g., W3C SSN, W3C BOT). Next, we aim to assess and select open-source frameworks that can analyse a given context and plan the right course of action to achieve the given goal. We aim to combine rule-based systems, e.g., Drools/OpenHAB-Rules and AI planning techniques, e.g., Optaplanner, to implement parts of MAPE-k. We aim to build the demonstrators using OpenHAB. Currently, smart home security solutions focus on network traffic analysis to detect cyber-physical threats using ML/DL technique. This project aims to demonstrate the utility of knowledge-bases systems towards smart home security.
- Conduct a literature review on knowledge-based techniques that are being developed and deployed within the smart home domain with a special focus on cyber-physical security
- Develop a knowledge model to capture all the relevant information required by Monitor-Analyze-Plan-Execute-Knowledge (MAPE-k) loop to enable self-adaptive cyber-physical security.
- Investigate, select and implement the best techniques for each phase within MAPE-k while utilising open-source APIs/frameworks as much as possible.
- Measure the trade-offs of competing techniques and make recommendations for their use
- Develop a series of demonstrators to showcase how knowledge-based self-adaptive systems work in the wild in the context of smart homes.
Team
Partners
PETRAS 2
Building Research Establishment (BRE)
The Building Research Establishment (BRE) is a centre of building science in the United Kingdom, owned by a charitable organisation, the BRE Trust. BRE provides research, advice, training, testing, certification and standards for public and private sector organisations in the UK and abroad.
Outcomes
Journal
Azhar Alsufyani, Omer Rana, and Charith Perera,
Knowledge-based Cyber Physical Security at Smart Home: A Review,
ACM Computing Surveys (ACM Comput. Surv.), Volume 57, Issue 3, Article No: 53, November 2024, pp. 1–36.
