Performance Analysis of Noise Subspace-based Narrowband Direction-of-Arrival (DOA) Estimation Algorithms on CPU and GPU

TL;DR

This paper evaluates the performance of noise subspace-based DOA estimation algorithms on CPU and GPU, demonstrating significant speedups through various optimization strategies and providing publicly available source code.

Contribution

It introduces optimized GPU implementations of DOA algorithms and compares their performance with CPU versions, highlighting effective strategies for real-time array signal processing.

Findings

Up to 3.1x speedup on GPU over CPU implementations

Comparison of MATLAB, C/C++, and CUDA implementations

Public release of optimized source code

Abstract

High-performance computing of array signal processing problems is a critical task as real-time system performance is required for many applications. Noise subspace-based Direction-of-Arrival (DOA) estimation algorithms are popular in the literature since they provide higher angular resolution and higher robustness. In this study, we investigate various optimization strategies for high-performance DOA estimation on GPU and comparatively analyze alternative implementations (MATLAB, C/C++ and CUDA). Experiments show that up to 3.1x speedup can be achieved on GPU compared to the baseline multi-threaded CPU implementation. The source code is publicly available at the following link: https://github.com/erayhamza/NssDOACuda