This project will develop techniques to reduce information adversaries can gain from observing cyber physical systems. Using this the capability to develop new context privacy techniques for new systems and scenarios faster will be provided so novel context privacy threats can be rapidly addressed. Theoretical foundations will be used to support practical deployments.
In March 2017 the UK’s Office for Nuclear Regulation (ONR) introduced the Security Assessment Principles for the Civil Nuclear Industry (SyAPs). As these regulatory principles were a significant departure from previous regulatory approaches, the ONR committed to an independent review after 5 years. In this project the independent review was performed and confidential results were provided to ONR.
Resource-constrained IoT devices have typically been used to perform sensing and actuation, however, there is increasing interest in those devices performing decision making. However, if these tasks are computationally or memory intensive, then the IoT devices will not have insufficient resources to execute the tasks. One solution is to offload the tasks from resource-constrained IoT devices to resource-rich Edge nodes. For redundancy multiple Edge nodes should be provisioned, but this raises the question of which Edge node should be selected to perform a task. This project investigated building a middleware to perform task offloading based on a measure of behavioural trust with limited resources (e.g., 32 KiB of RAM).
Vehicular position, navigation, and timing (PNT) systems are of vital importance to current vehicles, future autonomous vehicles and infrastructure depending on time synchronisation. Existing GNSS infrastructure has limited mitigation to prevent a variety of attacks. This project performed a short feasibility study on practical attacks against a vehicle’s PNT system using a PNT attack emulator.
Access to space is becoming increasingly cheaper, meaning companies and organisations who were previously priced out of the market are now considering space-based deployments. These deployments may also include novel functionality such as debris collection. New entrants will lack the knowledge that well-established space organisations have about how to secure these systems and the new functionality will increase the ways in which these systems will be attacked. This project looked at ways in which these vulnerabilities could be identified and then formally proved to not be present.
With the vehicle-to-vehicle communications being used to facilitated new functionality, there is a need for new security mechanisms to protect this communication. Much work has already been undertaken to develop these mechanisms, but they are rarely tested in real-world environments. This project took existing security and privacy techniques and performed a deployment at three sites in the UK to test their efficacy.
As connected and autonomous vehicles are expected to become commonplace on roads nationwide, new featured are being integrated into vehicles. This includes functionality such as vehicle-to-vehicle communication and machine learning models. It is important that we investigate and address security issues posed by the larger attack surface before a wide deployment. This project designed a reference architecture which facilitated attack surface analysis of connected autonomous vehicles.
Wireless sensor networks are useful for monitoring events over large areas for example, tracking the location of endangered species. However, by deploying this network to obtain data for conservation, it also reveals context information to an adversary about where the animals are. My PhD involved developing routing algorithms to delay an adversary in their attempt to locate the source of messages in such a network.
