Theta2p-Completeness: A Classical Approach for New Results

  • Authors:
  • Holger Spakowski;Jörg Vogel

  • Affiliations:
  • -;-

  • Venue:
  • FST TCS 2000 Proceedings of the 20th Conference on Foundations of Software Technology and Theoretical Computer Science
  • Year:
  • 2000

Quantified Score

Hi-index 0.01

Visualization

Abstract

In this paper we present an approach for proving Θ2p- completeness. There are several papers in which different problems of logic, of combinatorics, and of approximation are stated to be complete for parallel access to NP, i.e. Θ2p-complete. There is a special acceptance concept for nondeterministic Turing machines which allows a characterization of Θ2p as a polynomial-time bounded class. This characterization is the starting point of this paper. It makes a master reduction from that type of Turing machines to suitable boolean formula problems possible. From the reductions we deduce a couple of conditions that are sufficient for proving Θ2p-hardness. These new conditions are applicable in a canonical way. Thus we are able to do the following: (i) we can prove the Θ2p-completeness for different combinatorial problems (e.g. max-card-clique compare) as well as for optimization problems (e.g. the Kemeny voting scheme), (ii) we can simplify known proofs for Θ2p-completeness (e.g. for the Dodgson voting scheme), and (iii) we can transfer this technique for proving Δ2p-completeness (e.g. TSPcompare).