On the Computational Capacity of Parallel Communicating Finite Automata
Systems of parallel finite automata communicating by states are investigated. We consider deterministic and nondeterministic devices and distinguish four working modes. It is known that systems in the most general mode are as powerful as one-way multihead finite automata. Here we solve some open problems on the computational capacity of systems working in the remaining modes. In particular, it is shown that deterministic returning and non-returning devices are equivalent, and that there are languages which are accepted by deterministic returning and centralized systems but cannot be accepted by deterministic non-returning centralized systems. Furthermore, we show that nondeterministic centralized systems are strictly more powerful than their deterministic variants. Finally, incomparability with the class of (deterministic) (linear) context-free languages as well as the Church-Rosser languages is derived.
KeywordsTransition Function Finite Automaton Computational Capacity Input Symbol Query State
Unable to display preview. Download preview PDF.
- 4.Csuhaj-Varjú, E., Dassow, J., Kelemen, J., Păun, G.: Grammar Systems: A Grammatical Approach to Distribution and Cooperation. Gordon and Breach, Yverdon (1994)Google Scholar
- 5.Ďuriš, P., Jurdziński, T., Kutyłowski, M., Loryś, K.: Power of cooperation and multihead finite systems. In: Larsen, K.G., Skyum, S., Winskel, G. (eds.) ICALP 1998. LNCS, vol. 1443, pp. 896–907. Springer, Heidelberg (1998)Google Scholar
- 10.Klemm, R.: Systems of communicating finite state machines as a distributed alternative to finite state machines. Phd thesis, Pennsylvania State University (1996)Google Scholar
- 14.Wagner, K., Wechsung, G.: Computational Complexity. Reidel, Dordrecht (1986)Google Scholar