Trustworthy Decentralised Adaptive Systems

The trustworthiness of the upcoming generation software-intensive systems will be of crucial importance to our society; an example of such systems is a smart area surveillance system. Our focus of trustworthiness is on guaranteeing reliability, performance and efficiency goals. Uncertainties in the operating context and inherent decentralisation make the engineering of these systems very complex. Existing approaches based on runtime adaptation are not adequate to realise coordinated adaptation in an efficient and trustworthy manner. We propose a novel approach to engineer trustworthy decentralised adaptive systems centred on: (i) a novel formally founded language to specify interacting feedback loops that can directly be executed at runtime, and (ii) statistical model checking at runtime to guarantee goals efficiently with required confidence. We validate the research results for an unmanned underwater vehicle system and two IoT systems.

Keywords:Software engineering, Self-adaptation, Trustworthiness, Executable modelling language, Verification

Disciplines:Applied mathematics in specific fields, Computer architecture and networks, Distributed computing, Information sciences, Information systems, Programming languages, Scientific computing, Theoretical computer science, Visual computing, Other information and computing sciences