On the Causality of Mixed-Signal and Hybrid Models

Jie Liu and Edward A. Lee

6th Annual International Workshop on Hybrid Systems: Computation and Control (HSCC'03), April 3-5, 2003, Prague, Czech..

Prepublished version
Published version

ABSTRACT

This paper extends the application of the Cantor metric as a mathematical tool for defining causalities from pure discrete models to mixed-signal and hybrid models. Using the Cantor metric, which maps timed signals, continuous or discrete, into a metric space, we define causality as contractive properties of processes operating on these signals. Thus, the Banach fixed point theorem can be applies to establish conditions for the existence, uniqueness, and liveness of the behaviors for mixed-signal and hybrid systems. The results also provide theoretical foundations for the simulation technologies for such systems, including the time-marching strategy, evaluation of feedback loops, and the necessity of supporting rollback.