Measuring and Evaluating the Performance of Self-Organization Mechanisms Within Collective Adaptive Systems

Benedikt Eberhardinger, Hella Ponsar, Dominik Klumpp, and Wolfgang Reif

By restructuring and reconfiguring itself at run-time, a collective adaptive system (CAS) is able to fulfill its requirements under uncertain, ever-changing environmental conditions. Indeed, this process of self-organization (SO) is of utmost importance for the ability of the CAS to perform. However, it is hard to design high-performing SO mechanisms, because the environmental conditions are partially unpredictable at design time. Thus, a crucial aid for the development of SO mechanisms is a tool set enabling performance evaluations at design time in order to select the best-fitting mechanism and parametrize it. We present a metric for measuring the performance of an SO mechanism as well as a framework that enables evaluation of this metric. The proposed metric is evaluated for different kinds of SO mechanisms in two case studies: a smart energy management system and a self-organizing production cell.
published 05.11.2018 In book: Leveraging Applications of Formal Methods, Verification and Validation. Distributed Systems/ISoLA Symposium 2018

For questions regarding the publication, please contact!