Locally definable acceptance types — The three-valued case

  • Ulrich Hertrampf
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 583)


We consider polynomial time nondeterministic machines, and augment the nodes of the tree associated with a computation by functions from {0,...,m−1}r into {0,...,m−1}, where r is the number of successor nodes and m is some fixed positive integer. The leaves of the tree are labelled with constants (less than m), deterministic nodes take the unchanged value from their successor nodes. Thus from the leaves to the root we can compute a value for every node in the tree, and we say the tree accepts, if and only if the root value is 1. If all the functions appearing in nondeterministic nodes of some computation tree are from some set of functions F, then F is called an m-valued locally definable acceptance type. The class of languages recognizable by such machines with acceptance type F is denoted by (F)P.

The case m = 2 was (in some different context) investigated by Goldschlager and Parberry [GP86]. The only classes appearing as (F)P for 2-valued acceptance types F are P, NP, coNP, ⊕P, and PSPACE. m-valued acceptance types for m>2 were introduced in [He91a]. There it is proved that for all finite m-valued acceptance types F there exists an m′-valued acceptance type G (with possibly different constant m′), such that (F)P = (G)P, and G consists of only one binary function.

In the current paper we completely classify all classes (G)P where G is some 3-valued acceptance type consisting of only one binary function. It turns out that 20 classes are characterized by such acceptance types, trivially including the five classes appearing in the two-valued case, further the complete second level of the polynomial time hierarchy (Σ 2 p , Π 2 p , Δ 2 p , Θ 2 p ), some interesting very new classes (like ∇P and co-∇P). and some classes that are well known, but did not have such a closed machine characterization before (like PNP[1]). The proof technique for the main result is the following: first some criteria are given that a function refers to a trivial class (i.e. a class already appearing in the two-valued case), then a computer program is designed to pick out only the relevant cases, and finally the remaining 104 functions are investigated in detail.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [BGH90]
    R. Beigel, J. Gill, U. Hertrampf, Counting Classes: Thresholds, Parity, Mods, and Fewness, Proc. 7th Symposium on Theoretical Aspects of Computer Science, LNCS 415 (1990), pp. 49–57.Google Scholar
  2. [CKS81]
    A.K. Chandra, D.C. Kozen, L.J. Stockmeyer, Alternation, J. ACM 28 (1981), pp. 114–133.Google Scholar
  3. [GNW90]
    T. Gundermann, N.A. Nasser, G. Wechsung, A Survey on Counting Classes, 5th Structure in Complexity Theory (IEEE) (1990), pp. 140–153.Google Scholar
  4. [GP86]
    Leslie M. Goldschlager, I. Parberry, On the Construction of Parallel Computers from various Bases of Boolean Functions, Theoretical Computer Science 43 (1986), pp. 43–58.Google Scholar
  5. [He90]
    Ulrich Hertrampf, Relations among MOD-Classes, Theoretical Computer Science 74 (1990), pp. 325–328.Google Scholar
  6. [He91a]
    Ulrich Hertrampf, Locally Definable Acceptance Types for Polynomial Time Machines, Technical Report No. 28, Universität Würzburg, 1991.Google Scholar
  7. [He91b]
    Ulrich Hertrampf, Locally Definable Acceptance Types — The Three-Valued Case, Technical Report No. 29, Universität Würzburg, 1991.Google Scholar
  8. [JGK70]
    S.W. Jablonski, G.P. Gawrilow, W.B. Kudrjawzew, Boolesche Funktionen und Postsche Klassen, Akademie-Verlag Berlin, 1970.Google Scholar
  9. [Ja79]
    S.W. Jablonski, Einführung in die Theorie der Funktionen der k-wertigen Logik, in Diskrete Mathematik und mathematische Fragen der Kybernetik, Birkhäuser, 1990.Google Scholar
  10. [PZ83]
    C.H. Papadimitriou, S.K. Zachos, Two Remarks on the Complexity of Counting, Proc. 6th GI Conf. on Theoretical Computer Science, LNCS 145 (1983), pp. 269–276.Google Scholar
  11. [To91]
    S. Toda, PP is as Hard as the Polynomial-time Hierarchy, SIAM J. Comput. 20 (1991), pp. 865–877.Google Scholar
  12. [Wa88]
    Klaus W. Wagner, Bounded Query Computations, Proc. 3rd Structure in Complexity Theory (IEEE) (1988), pp. 260–277.Google Scholar
  13. [Wa91]
    Klaus W. Wagner, personal communication.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1992

Authors and Affiliations

  • Ulrich Hertrampf
    • 1
  1. 1.Institut für InformatikUniversität WürzburgWürzburgGermany

Personalised recommendations