This research applies an emerging field of mathematics known as Dynamic Equations on Time Scales to the problem of bandwidth utilization in real-time distributed control networks. Time scales theory allows for the seamless analysis of dynamic processes on continuous, discrete, or mixed time domains, including non-uniform discrete time domains. Recent advances in time scales show that periodic servo control processes on distributed networks can adapt their timing characteristics in the presence of aperiodic high-priority traffic bursts without losing control of their plants. The adaptive servo timing method results in an increase in the effective available bandwidth for aperiodic traffic and better integration of the overall plant/processor/network system. Study of the theory of time scales is leading to further advances toward the unification of continuous and discrete systems as well. This unique application of an area of mathematics previously unknown to the computer science and engineering communities has the potential to stimulate external research into related topics such as network scheduling and real-time control performance.
The project will pursue empirical validation of the new theory on a 40-node Controller Area Network of embedded processors developed at Baylor University. Successful research in this area has immediate utility and potential economic impact for a large industrial base employing distributed control networks, notably in the automotive, aerospace, and manufacturing industries. Baylor's relatively new graduate engineering programs will continue to support the educational needs of engineers employed by local industries in the underserved Central Texas area.
