Sichere, autonome und KI-basierte Systeme
Sichere, autonome und KI-basierte Systeme
Die digitale Transformation und der Ausbau cyber-physischer Systeme erfordern zunehmend kollaborative Lösungen und Mensch-Maschine-Interaktion. Cognitive Computing erhöht zugleich die Autonomie der Systeme (zum Beispiel autonome Fahrzeuge und Flugobjekte). Damit verbunden sind auch die Herausforderungen der Kommunikation und Schnittstellengestaltung zwischen den Komponenten und Systemen, die Datenerfassung und -analyse mittels Künstlicher Intelligenz (u.a. Big Data, Maschinelles Lernen) sowie die IT-Sicherheit.
Geforscht wird hierzu vor allem am Affective and Cognitive Institute (ACI), am Institute for Machine Learning and Analytics (IMLA), am Institute for Unmanned Aerial Systems (IUAS) und am Institut für verlässliche Embedded Systems und Kommunikationselektronik (ivESK).
| Titel | PAL SAaaS: Building Triangular Trust for Secure Cloud Auditing |
| Kurzname | PAL SAaaS |
| Kurzbeschreibung | The most relevant security threats in cloud computing are data breaches and data loss. Thus, to bring cloud computing to the next level, customers need to be ensured that a cloud provider securely and transparently offers the promised functionality. The currently discussed and probably most promising concept is 'Security Adit as a Service' (SAaaS) where a third party, called the auditor, supervises the cloud provider on behalf of the user. Therefore, the overall project is to develop and integrate cryptographic building blocks for secure and liable cloud audits. Our focus will be on PAL as the acronym for: Privacy: The process of auditing requires the collection and analysis of digital evidence about the user and/or the service provider. It is necessary that by doing so, the privacy of these parties must not be violated. This also includes aspects like confidentiality of outsourced data. Availability: A main concern with respect to the usage of cloud services is the availability of the outsourced data. Consequently, this needs to be also addressed by a cloud service audit. In addition, availability is necessary for the collected meta data and the results of computations as well. Liability: This property has been mostly overlooked so far. Especially when cloud services are used by entreprises, the question of liability needs to be settled. This covers both the service provider and the auditor. In cryptographic terms: each party needs to be able to convincingly prove that it did its job. |
| Jahr der Einwerbung | 2015 |
| Laufzeit Beginn | 01.11.2015 |
| Laufzeit Ende | 31.03.2019 |
| Projektleitung | Westhoff, Dirk, Prof. Dr. |
| Fakultät | M |
| Institut | ivESK |