QoS middleware support for pervasive computing applications
37th Annual Hawaii International Conference on System Sciences, 2004. Proceedings of the
Today, pervasive computing technologies are being developed to provide automated, real-time, continual, and unobtrusive user services in dynamic heterogeneous environments such as telemedicine, manufacturing, space endeavors, crisis management, and military. However, the full potential of pervasive computing cannot be realized without enabling middleware technologies -henceforth referred to as "middleware services"; i.e., they provide services for high level applications. In this paper we
... uce middleware services that are required to address the challenges related to adapting to dynamically changing situations, meeting communication QoS requirements, and achieving scalability in large-scale pervasive computing applications over heterogeneous network infrastructures. Our approach in developing middleware support for pervasive computing is based on the community computing concept, which provides us with a unified method to dynamically integrate the middleware services with each other and with high level applications on a just-in-time basis. The proposed middleware services for pervasive computing are highly modular, lightweight, and easily deployable to meet the needs of pervasive computing applications in networked environments. In this paper we will describe our proposed QoS algorithms and techniques for synergistic integration of middleware services into pervasive computing applications. In addition, we propose new performance metrics and a benchmark suite approach to evaluate middleware services in pervasive computing. Finally, we present some early prototype results from our proof of concept implementation.