A distributed simulation technique for multiple input multiple output systems

Distributed computing concepts have been attempted in the realm of real-time process control for a long time. This started with the systolic arrays of processors and an enormous amount of research has been done in this field. This paper proposes a methodology by which a multiple input multiple output system representing a plant and defined in terms of connected blocks of transfer functions can be decomposed into a set of parallel simulation processes in a message passing environment viz. a network of work stations. The paper also analyzes the scope of optimization of the processing time associated with each task. The methodology includes online reconfiguration of the system.