Counting De Bruijn sequences as perturbations of linear recursions
Don Coppersmith, Robert C. Rhoades, Jeffrey M. VanderKam
Abstract
Open-access reader
Don Coppersmith, Robert C. Rhoades, Jeffrey M. VanderKam
Abstract
Open-access reader
Every binary De~Bruijn sequence of order n satisfies a recursion 0=x_n+x_0+g(x_{n-1}, ..., x_1). Given a function f on (n-1) bits, let N(f; r) be the number of functions generating a De Bruijn sequence of order n which are obtained by changing r locations in the truth table of f. We prove a formula for the generating function \sum_r N(\ell; r) y^r when \ell is a linear function. The proof uses a weighted Matrix Tree Theorem and a description of the in-trees (or rooted trees) in the n-bit De Bruijn graph as perturbations of the Hamiltonian paths in the same graph.
OpenAlex reports 9 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Every binary De~Bruijn sequence of order n satisfies a recursion 0=x_n+x_0+g(x_{n-1}, ..., x_1). Given a function f on (n-1) bits, let N(f; r) be the number of functions generating a De Bruijn sequence of order n which are obtained by changing r locations in the truth table of f. We prove a formula for the generating function \sum_r N(\ell; r) y^r when \ell is a linear function. The proof uses a weighted Matrix Tree Theorem and a description of the in-trees (or rooted trees) in the n-bit De Bruijn graph as perturbations of the Hamiltonian paths in the same graph.
Key concepts: De Bruijn sequence, De Bruijn graph, Combinatorics, Recursion (computer science), Mathematics, Sequence (biology), Discrete mathematics, Graph