Verifying self-adaptive applications suffering uncertainty

Abstract

Self-adaptive applications address environmental dynamics systematically. They can be faulty and exhibit runtime errors when environmental dynamics are not considered adequately. It becomes more severe when uncertainty exists in their sensing and adaptation to environments. Existing work verifies self-adaptive applications, but does not explicitly consider environmental constraints or uncertainty. This gives rise to inaccurate verification results. In this paper, we address this problem by proposing a novel approach to verifying self-adaptive applications suffering uncertainty in their environmental interactions. It builds Interactive State Machine (ISM) models for such applications and verifies them with explicit consideration of environmental constraints and uncertainty. It then refines verification results by prioritizing counterexamples according to their probabilities. We experimentally evaluated our approach with real-life self-adaptive applications, and the experimental results confirmed its effectiveness. Our approach reported 200-660% more counterexamples than not considering uncertainty, and eliminated all false counterexamples caused by ignoring environmental constraints.

A survey on dependability improvement techniques for pervasive computing systems

“…Pervasive computing systems can suffer from faults that exist in traditional software artifacts. What is worse, since pervasive computing systems continually adapt to environmental changes in an autonomic way, their adaptation may be faulty when the complexity of modeling all environmental changes is beyond a developer's capability [7,47]. For example, RFID systems, a typical kind of pervasive computing systems, are very difficult to implement because developers have to carefully consider, predict, and handle various environmental factors such as cross-coupling of neighboring RFID tags.…”

“…Determinism requires a pervasive computing system's adaptation behavior to be predictable upon any situation, that is, absence of nondeterminism. For example, many pervasive computing systems are designed based on a set of adaptation rules [7,47,49], which specify the actions that the systems should execute when certain conditions are satisfied at certain system states. A system is considered to be deterministic if there is at most one rule that can be triggered at any situation.…”

“…However, despite such fascinating progress, developing and deploying dependable pervasive computing systems still remains as a challenging task [4,5]. In fact, existing studies disclosed that many real-world pervasive computing systems do not provide dependable services, causing poor user experiences or economic losses [6,7]. A pervasive computing system is context-aware, which means the system is cognizant of its environment, and able to make reaction based on this information [4].…”

“…Ideally, each reader should be able to read all tags in its vicinity, but the correct reading rate of a reader may fall below 70% in real-life RFID deployments, which means at least 30% of its contexts derived from RFID reads can be incomplete or imprecise [19]. This is one kind of uncertainty, and there also exist other kinds of uncertainties caused by unreliable sensing or flawed adaptation in pervasive computing systems [7,20].…”

CoMID: Context-Based Multiinvariant Detection for Monitoring Cyber-Physical Software

“…C YBER-PHYSICAL software programs (or cyberphysical programs) continually interact with their external physical environments to provide context-aware adaptive functionalities. Examples of such programs include those running on robot cars [1]- [3], unmanned aerial vehicles (UAVs) [4]- [6], and humanoid robots [7]- [9]. Cyber-physical programs keep sensing environmental changes, making decisions based on their pre-programmed logics, and then taking physical actions to cope with the sensed changes.…”