Permanent errors may contribute to emergent behavior in one-dimensional cellular automata

This paper describes the possibility of increasing the complexity of behavior of one-dimensional cellular automata with two states. The mechanism is based on simulating permanent errors which may occur in hardware implementation of cellular automata employed e.g. in Artificial Life. Complete exploration of simple 3-neighborhood is conducted and the change of behavior is illustrated in changes of Wolfram's classification of said automata. Several 5-neighborhood examples of similar behavior are provided to show the consistency of complexity-enhancing behavior in different type of one-dimensional cellular automata.