Dependable System Design Methodologies

Contact person: Cristiana Bolchini

System dependability and reliability issues are investigated with the aim of defining methodologies and tools able to support the designer in the introduction of fault detection, fault tolerance and recovery properties in digital systems. More precisely, the research covers reliability-related aspects related to both design and analysis of embedded systems.
The adopted fault model covers both transient and permanent failures, at different levels of abstraction. The proposed design methodologies act on both hardware and software, working at the architectural level as well as on the application being executed; the goal is the exploration of the design space, by considering figures of merit that take into account not only area overhead, but also flexibility, possibility to use Components-Off-The-Shelf (COTS), performance degradation, … Different implementation platforms are taken into account; Field Programmable Gate Arrays (FPGAs) are receiving a particular attention to exploit their flexibility in terms of reconfigurability to pursue system dependability properties.
Most recent research moves towards (self-)adaptable systems able to autonomously and dynamically cope with the occurrence of failures, offering a “tunable” degree of reliability.

Recent selected publications:

[bibtex cite=DATE2014]
[bibtex cite=TC2013]
[bibtex cite=JETTA2013]
[bibtex cite=TC2011BMS]