SimCA*

Abstract

Self-adaptation provides a principled way to deal with software systems’ uncertainty during operation. Examples of such uncertainties are disturbances in the environment, variations in sensor readings, and changes in user requirements. As more systems with strict goals require self-adaptation, the need for formal guarantees in self-adaptive systems is becoming a high-priority concern. Designing self-adaptive software using principles from control theory has been identified as one of the approaches to provide guarantees. In general, self-adaptation covers a wide range of approaches to maintain system requirements under uncertainty, ranging from dynamic adaptation of system parameters to runtime architectural reconfiguration. Existing control-theoretic approaches have mainly focused on handling requirements in the form of setpoint values or as quantities to be optimized. Furthermore, existing research primarily focuses on handling uncertainty in the execution environment. This article presents SimCA*, which provides two contributions to the state-of-the-art in control-theoretic adaptation: (i) it supports requirements that keep a value above and below a required threshold, in addition to setpoint and optimization requirements; and (ii) it deals with uncertainty in system parameters, component interactions, system requirements, in addition to uncertainty in the environment. SimCA* provides guarantees for the three types of requirements of the system that is subject to different types of uncertainties. We evaluate SimCA* for two systems with strict requirements from different domains: an Unmanned Underwater Vehicle system used for oceanic surveillance and an Internet of Things application for monitoring a geographical area. The test results confirm that SimCA* can satisfy the three types of requirements in the presence of different types of uncertainty.

Effects of time and temperature during melanging on the volatile profile of dark chocolate

“…Volatile metabolite abundances were compared using two-way analysis of variance (ANOVA) via the aov function in the R statistical environment v. 3.5.1 27 , and false discovery rate adjustment was performed on the ANOVA p-values using the Benjamini–Hochberg algorithm 28 Principal Components Analysis (PCA), and Orthogonal Partial Least Squares Analysis (OPLS) were conducted on 88 annotated metabolites with SIMCA software v. 15 (Sartorius Stedim Biotech, Umea, Sweden). Scores and loadings values from OPLS models were conducted in SIMCA software 29 on data for time and temperature (y) and annotated, unit variance scaled volatile metabolites (x). Predictive power (Q2) was determined via cross-validation, by which the data was divided into seven parts and 1/7th of the data were removed, and the model was built on the remaining 6/7th of data, and the removed 1/7th of data are predicted from the model.…”

Quantitative verification with adaptive uncertainty reduction

“…The approach from [71] proposes a combination of adaptation and evolution of the software to make its behavior resilient to uncertainty, which in turn entails that the software system is sustainable, while [72] focuses on the uncertainty surrounding the execution of cyber-physical production systems. A different approach can be found in [73], where a control-theoretic approach is adopted to handle uncertainty in self-adaptive software systems. Furthermore, the need for software systems to operate well under the existing uncertainties is among the main waves that have advanced the research on self-adaptive systems [74], although a perpetual assurance of goal satisfaction in selfadaptive systems is still an open research challenge [75].…”

A Formal Model for Quality-Driven Decision Making in Self-Adaptive Systems

“…In [24], an approach called Simplex Control Adaptation (SimCA*) is presented, it allows building self-adaptive software systems that satisfy multiple STO-reqs -a combination of Sreqs (stakeholder requirements), T-reqs (threshold requirements), and O-reqs (optimization requirements) in the presence of different types of uncertainty. SimCA* has dedicated specific components to monitor changes in the underlying system or its environment and adjust the adaptation logic accordingly to deal with different types of uncertainty.…”

“…Therefore, self-adaptation mechanisms driven by quality modify system behavior dynamically. Albeit there has been extensive research to address uncertainty in SASs [24,19,21], there is no focus on proposing solutions to identify uncertainty at different levels of the decision-making process and considering it when modelling the SASs. Besides, the main focus has been on achieving adaptations without determining their side effects on the overall system qualities.…”