Software Engineering for Self-Adaptive Systems: Research Challenges in the Provision of Assurances

Abstract

The important concern for modern software systems is to become more cost-effective, while being versatile, flexible, resilient, dependable, energy-efficient, customisable, configurable and self-optimising when reacting to run-time changes that may occur within the system itself, its environment or requirements. One of the most promising approaches to achieving such properties is to equip software systems with self-managing capabilities using self-adaptation mechanisms. Despite recent advances in this area, one key aspect of self-adaptive systems that remains to be tackled in depth is the provision of assurances, i.e., the collection, analysis and synthesis of evidence that the system satisfies its stated functional and non-functional requirements during its operation in the presence of self-adaptation. The provision of assurances for selfadaptive systems is challenging since run-time changes introduce a high degree of uncertainty. This paper on research challenges complements previous roadmap papers on software engineering for self-adaptive systems covering a different set of topics, which are related to assurances, namely, perpetual assurances, composition and decomposition of assurances, and assurances obtained from control theory. This research challenges paper is one of the many results of the Dagstuhl Seminar 13511 on Software Engineering for Self-Adaptive Systems: Assurances which took place in December 2013.

Zone-based formal specification and timing analysis of real-time self-adaptive systems

“…The formula reported in (12) mans that does not exist a path, where the proposition φ eventually holds. Thus, (11) and (12) are semantically equivalent to (9) and (10), respectively.…”

“…• Support for feedback control loops: feedback loops for adaptation (such as the MAPE-K control loop [3]) are considered cornerstones for the design of self-adaptive systems [4,60,9], since it is generally agreed that feedback loops facilitate the identification of the core phenomena to control and the realization of compositional verification techniques that can be applied incrementally along the adaptation loop [60]. Therefore, a formal framework should allow their modeling and analysis explicitly.…”

“…9 We did not attempt an exhaustive list of features for reviewing existing approaches, but only those features we consider relevant for our research goal. Table 4 below summarizes the results of such a comparison by reporting the main features of the frameworks we considered in our work (see Subsection 7.1), including our proposal (the ZAFETY framework).…”

“…However, the survey in [8] and the Dagstuhl seminar [9] show that, although the attention for self-adaptive software systems is gradually increasing, the number of studies that employ mathematically-based specification and verification techniques from the area of formal methods remains low. In particular, specification and verification of selfadaptive systems are very difficult to carry out when involving time constraints.…”

Self-managing Internet of Things

“…In the past few years, research in this area has particularly been focussing on how to provide assurances for the adaptation goals of self-adaptive systems that operate under uncertain operating conditions [14,31]. This is particularly important for systems with strict quality goals.…”