Date of Award

5-1-2013

Document Type

Thesis (Undergraduate)

Department or Program

Department of Computer Science

First Advisor

Amit Chakrabarti

Abstract

Boolean functions are at the heart of all computations, and all Boolean functions can be reduced to a sum of pattern-matching functions, called minterm-cyclic functions. In this thesis, we examine properties of polynomials representing minterm-cyclic Boolean functions. We use the term "saturated" to represent a polynomial with degree equal to input size n for all n; this indicates the intuitive notion that such functions are in some way complex, or difficult to compute. We present three main results. Firstly, there exist an infinite number of monotone minterm-cyclic functions that are not saturated. Secondly, for a specific class of minterms called self-avoiding minterms, we prove that the associated pattern-matching functions are not saturated; specifically, they can only have non-zero degree-n coefficients for n a multiple of the size of the minterm. Thirdly, for self-avoiding minterms that contain some '*', the degree-n coefficients are always zero. These results may have implications in the fields of algebraic cryptographic attacks or efficiency of error-correcting codes.

Comments

Originally posted in the Dartmouth College Computer Science Technical Report Series, number TR2013-740.

Download
COinS