An NP-Complete Number-Theoretic Problem
- 1 July 1979
- journal article
- Published by Association for Computing Machinery (ACM) in Journal of the ACM
- Vol. 26 (3), 567-581
- https://doi.org/10.1145/322139.322152
Abstract
Systems of nonlinear equations of the form D Aft = ~(x), where A is an m × n matrix of ratmnal constants and fi = (yl, , y.), 8(x) = (ol(x), , on(x)) are column vectors, are considered Each o,(x) is of the form r,(x) or lr,(x)), where r,(x) is a rational function ofx with raUonal coefficients It Is shown that the problem of determining for a given system D whether there exists a nonnegatlve integral solution (yh,, y., x) satisfying Dts deodable In fact, the problem is NP-complete when restricted to systems D m which the maximum degree of the polynomials defining the o,(x)'s is bounded by some fixed polynomial m the length of the representation of D Some recent results connecting Dmphantme equations and counter machines are briefly mentioned.Keywords
This publication has 7 references indexed in Scilit:
- Reversal-Bounded Multicounter Machines and Their Decision ProblemsJournal of the ACM, 1978
- Bounds on positive integral solutions of linear Diophantine equationsProceedings of the American Mathematical Society, 1976
- Hilbert’s tenth problem: Diophantine equations: positive aspects of a negative solutionPublished by American Mathematical Society (AMS) ,1976
- Reduction of an arbitrary diophantine equation to one in 13 unknownsActa Arithmetica, 1975
- Computationally Related ProblemsSIAM Journal on Computing, 1974
- There Cannot be any Algorithm for Integer Programming with Quadratic ConstraintsOperations Research, 1973
- Reducibility among Combinatorial ProblemsPublished by Springer Nature ,1972