Research ® Resilience Benchmarking
I am currently very interesting in resilience benchmarking. Resilience benchmarks enable analysis and quantification of a system's ability to cope with and withstand faults while sustaining proper operation. More generally, the goal of resilience benchmarking is to provide generic methods for i) characterising and quantifying the system behaviour in the presence of faults, and ii) comparing the resilience of alternative solutions. Additionally, weak parts of the system can be identified and emphasis put on enhancing their ability to withstand faults.
While benchmarking is widely used to measure computer performance in reproducible ways, resilience benchmarking is still emerging. Some of the most comprehensive work reported to date on dependability benchmarking, was carried out in DBench (Dependability Benchmarking, IST-2000-25425), an FP5 project. The partners of DBench were mostly from academia: LAAS-CNRS (coordinator), Critical Software (SME), University of Coimbra, University of Erlangen and Technical University of Valencia. DBench has defined a framework for benchmarking the dependability of computer systems, with emphasis on so called software Off-the-Shelf components.
The aim of this research is to develop resilience benchmarks for benchmarking the resilience of execution supports specifically designed for personal mobile devices, like smartphones and PDAs, taking into consideration malicious faults in addition to accidental ones. The provision of methodologies and tools to evaluate mobile system platforms from the three-fold perspective of their performance, security and dependability is thus the main goal of this research.
For the time being, I am only at the beginning of this research. I am studying the problem and classifying available information on related topics. Hereafter a first attempt to this classification :
Validate links, HTML, stylesheet, accessiblity.