Implementation of random linear network coding using NVIDIA'S CUDA toolkit

Abstract

In this paper we describe an efficient GPU-based implementation of random linear network coding using NVIDIA's CUDA toolkit. The implementation takes advantage of the highly parallel nature of modern GPUs. The paper reports speed ups of 500% for encoding and 90% for decoding in comparison with a standard CPU-based implementation. © Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering 2010.