On Reducing the Encoding Complexity of Random Linear Network Coding

Abstract

This paper presents a novel implementation of Random Linear Network Coding (RLNC) as an attempt to improve the encoding complexity by reducing the overall number of multiplicative operations during a transmission session. The idea is to allow the encoding intermediate nodes to decide on the usefulness of multiplying a received packet by a random coefficient. In this approach, the encoding operation used at a given encoding intermediate node v adapts to the previous encoding operations used to encode the received packets, in such a way that the used encoding operation is RLNC-equivalent with fewer multiplications.