Fast GPU Implementation of Sparse Signal Recovery from Random Projections

TL;DR

This paper presents a GPU-accelerated implementation of the Matching Pursuit algorithm for sparse signal recovery, achieving up to 31 times faster performance than CPU versions using CUDA and CUBLAS libraries.

Contribution

The paper introduces a highly optimized GPU implementation of the greedy Matching Pursuit algorithm for sparse signal recovery, significantly improving computational speed.

Findings

GPU implementation is up to 31 times faster than CPU version.

The approach leverages NVIDIA CUDA and CUBLAS libraries.

The method enables faster sparse signal recovery from random projections.

Abstract

We consider the problem of sparse signal recovery from a small number of random projections (measurements). This is a well known NP-hard to solve combinatorial optimization problem. A frequently used approach is based on greedy iterative procedures, such as the Matching Pursuit (MP) algorithm. Here, we discuss a fast GPU implementation of the MP algorithm, based on the recently released NVIDIA CUDA API and CUBLAS library. The results show that the GPU version is substantially faster (up to 31 times) than the highly optimized CPU version based on CBLAS (GNU Scientific Library).