Scheduling of network access for feedback-based embedded systems

Technological advances are delivering devices that can be ubiquitously embedded in the physical world and that have sensing and communication capabilities., Examples range from smart dust embedded in building materials to networks of appliances in the home. Embedded devices will be deployed in unprecedented numbers, will enable pervasive distributed computing, and will radically change the way people interact with the surrounding environment.(1) The paper targets embedded systems and their real-time (RT) communication requirements. RT requirements arise from the "feedback loops" among distributed sensing and actuation elements plus the physical characteristics of their surrounding environment. A natural choice for modeling the feedback requirements is control theory and engineering.(2) The major thrust of this paper is to investigate the implications of control theory on network QoS. A control-theoretical analysis of the physical system dynamics yields error bounds functions, which, in turn, can be minimized by achieving novel QoS objectives in the network infrastructure. We aim at RT control-theoretical objectives in inexpensive LAN's, such as those that can be installed in houses or in small industrial plants. Such QoS objectives are achieved by a middleware that regulates network access through a polling scheme. The paper focuses on the middleware scheduling of network access. The middleware can also provide flexible support for fault-tolerance, error-recovery, and bandwidth monitoring. The advantages of our approach are its scalability, flexibility, evolvability, co-design of control systems and middleware, portability, provisioning for peak or catastrophic occurrences, high bandwidth utilization, self-clocking, link sharing across application, and reuse of existing plug-and-play solutions.