Context-awareness is an important feature in Ambient Intelligence environments including in pervasive middleware. In addition, there is a growing trend and demand on self-management capabilities for a pervasive middleware in order to provide high-level dependability for services. In this chapter, we propose to make use of context-awareness features to facilitate self-management. To achieve self-management, dynamic contexts for example device and service statuses, are critical to take self-management actions. Therefore, we consider dynamic contexts in context modeling, specifically as a set of OWL/SWRL ontologies, called the Self-Management for Pervasive Services (SeMaPS) ontologies. Self-management rules can be developed based on the SeMaPS ontologies to achieve self-management goals. Our approach is demonstrated within the LinkSmart pervasive middleware. Finally, our experiments with performance, extensibility, and scalability in the context of LinkSmart show that the SeMaPS-based self-management approach is effective.
Handbook of Research on Ambient Intelligence and Smart Environments Trends and Perspectives, 2011, p. 248-271