New LP-based Upper Bounds in the Rate-vs.-Distance Problem for Linear Codes

Abstract: We develop a new family of linear programs, that yield upper bounds on the rate of binary linear codes of a given distance. Our bounds apply {\em only to linear codes.} Delsarte's LP is the weakest member of this family and our LP yields increasingly tighter upper bounds on the rate as its control parameter increases. Numerical experiments show significant improvement compared to Delsarte. These convincing numerical results, and the large variety of tools available for asymptotic analysis, give us hope that our work will lead to new and improved asymptotic upper bounds on the possible rate of linear codes. A concurrent work by Coregliano, Jeronimo, and Jones offers a closely related family of linear programs which converges to the true bound. Here we provide a new proof of convergence for the same LPs.