Modular Quality-of-Service Analysis of Software Design Models for Cyber-Physical Systems

Abstract

Emerging applications such as collaborative and autonomous cyber-physical systems (CPS) seek for innovative techniques that support Quality-of-Service (QoS) analysis as key concern to be considered. The objective of this paper is to complement the software design models with an approach that provides a set of modules that are (i) representative of multiple QoS-based properties, and (ii) equipped with strategies aimed to establish rules of interaction among them in a feedback loop fashion. We propose a novel methodology that builds upon the specification of QoS-based modules and enables the generation of design alternatives as outcome of an internal intertwining of different QoS analysis results for CPS. The approach is applied to a collaborative and autonomous network of sensors, and experimental results show that software designers are supported in the selection of design alternatives by quantitative information. A comparison with an integrated model is performed to show the advantages of our novel modular QoS-based analysis.