Enforcing Security on Autonomous Vehicle Searches Through the Quantification of Opacity
| dc.contributor.author | Schonewille, Bryony | en |
| dc.contributor.department | Electrical and Computer Engineering | en |
| dc.contributor.supervisor | Rudie, Karen | |
| dc.date.accessioned | 2022-01-05T18:39:55Z | |
| dc.date.available | 2022-01-05T18:39:55Z | |
| dc.degree.grantor | Queen's University at Kingston | en |
| dc.description.abstract | Recently, topics of security have been explored in the field of discrete-event systems (DES). By modelling these systems with DES, the evolution of the processes can be captured, allowing for different vulnerabilities to be noticed. The DES field also provides a different set of tools which can generate new strategies to tackle the security problems in these systems. The motivating problem that this work focuses on is a group of autonomous vehicles traversing some terrain while trying to cooperatively complete a task such as searching for a target. This work shows that strategies can be employed that remove the need for communication under certain conditions. When no such strategy can be developed, it can be useful to have a tool to classify a system’s security. In DES, this has traditionally been captured by the notion of opacity. Unfortunately, this framework has many limitations. Opacity can only be applied to systems with a specific secret and it cannot tell you how obscured a system is. To address these failings, this work develops the concept of degree of opacity. This framework can measure the degree to which a system is secure based on a supplied criterion. It also can be applied to systems as a whole eliminating the dependency on a specific secret. Degree of opacity is a more general version of opacity and can be handled in similar ways. Degree of opacity can be enforced using supervisory control to produce a minimally restrictive supervisor. Alternatively, what is communicated to a vehicle or observed by a vehicle can be controlled, rather than restricting the vehicle’s movements. Unfortunately, non-monotonicity of observability means that it cannot be easily used to enforce degree of opacity like controllability was. To combat this, a strategy called transition pairing is developed which allows monotonicity to be achieved. This strategy is applied to construct an algorithm to pick which communications can be communicated to achieve a target degree of opacity. | en |
| dc.description.degree | M.A.Sc. | en |
| dc.identifier.uri | http://hdl.handle.net/1974/29864 | |
| dc.language.iso | eng | en |
| dc.relation.ispartofseries | Canadian theses | en |
| dc.rights | Attribution 3.0 United States | * |
| dc.rights | Attribution 3.0 United States | |
| dc.rights.uri | http://creativecommons.org/licenses/by/3.0/us/ | * |
| dc.rights.uri | http://creativecommons.org/licenses/by/3.0/us/ | |
| dc.subject | Discrete Event Systems | en |
| dc.subject | Opacity | en |
| dc.subject | Security | en |
| dc.subject | Autonomous Vehicles | en |
| dc.title | Enforcing Security on Autonomous Vehicle Searches Through the Quantification of Opacity | en |
| dc.type | thesis | en |
