Fast CRCs (Extended Version)

TL;DR

This paper introduces a class of fast cyclic redundancy checks (CRCs) that can be efficiently implemented in software by reducing polynomial division steps, improving error detection performance.

Contribution

The paper presents novel fast CRCs and an effective implementation technique that minimizes polynomial division steps without relying on table lookups.

Findings

Fast CRCs can be implemented with fewer polynomial division steps.

Implementation techniques significantly improve software performance.

Potential for broader adoption in error detection systems.

Abstract

CRCs have desirable properties for effective error detection. But their software implementation, which relies on many steps of the polynomial division, is typically slower than other codes such as weaker checksums. A relevant question is whether there are some particular CRCs that have fast implementation. In this paper, we introduce such fast CRCs as well as an effective technique to implement them. For these fast CRCs, even without using table lookup, it is possible either to eliminate or to greatly reduce many steps of the polynomial division during their computation.