In this paper, the inherent relationships between the running regulations and behavior characteristics of cellular automata are presented; an imprecise taxonomy of such systems is put forward; the three extreme cases of stable systems are discussed; and the illogicalness of evolutional strategies of cellular automata is analyzed. The result is suitable for the emulation and prediction of behavior of discrete dynamics systems; especially it can be taken as an important analysis means of dynamic performance of complex networks.