GPGPU Based Parallel Implementation of Spectral Correlation Density Function

Scott Marshall, Garrett Vanhoy, Ali Akoglu, Tamal Bose, Bo Ryu

Research output: Contribution to journalArticle

Abstract

In this study, the parallelization of a critical statistical feature of communication signals called the spectral correlation density (SCD) is investigated. The SCD is used for synchronization in OFDM-based systems such as LTE and Wi-Fi, but is also proposed for use in next-generation wireless systems where accurate signal classification is needed even under poor channel conditions. By leveraging cyclostationary theory and classification results, a method for reducing the computational complexity of estimating the SCD for classification purposes by 75% or more using the Quarter SCD (QSCD) is proposed. We parallelize the SCD and QSCD implementations by targeting general purpose graphics processing unit (GPU) through architecture specific optimization strategies. We present experimental evaluations on identifying the parallelization configuration for maximizing the efficiency of the program architecture in utilizing the threading power of the GPU architecture. We show that algorithmic and architecture specific optimization strategies result with improving the throughput of the state of the art GPU based SCD implementation from 120 signals/second to 3300 signals/second.

Original languageEnglish (US)
JournalJournal of Signal Processing Systems
DOIs
StatePublished - Jan 1 2019

Fingerprint

GPGPU
Parallel Implementation
Density Function
Probability density function
Correlation Function
Graphics Processing Unit
Wi-Fi
Parallelization
Orthogonal frequency division multiplexing
Computational complexity
Synchronization
Throughput
Optimization
Communication
Orthogonal Frequency Division multiplexing (OFDM)
Experimental Evaluation
Graphics processing unit
Computational Complexity
Configuration
Architecture

Keywords

  • FFT accumulation method
  • GPGPU
  • Signal classification
  • Spectral correlation density

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Theoretical Computer Science
  • Signal Processing
  • Information Systems
  • Modeling and Simulation
  • Hardware and Architecture

Cite this

GPGPU Based Parallel Implementation of Spectral Correlation Density Function. / Marshall, Scott; Vanhoy, Garrett; Akoglu, Ali; Bose, Tamal; Ryu, Bo.

In: Journal of Signal Processing Systems, 01.01.2019.

Research output: Contribution to journalArticle

@article{a9612dedb95a4a6eb1f6664f85b376e7,
title = "GPGPU Based Parallel Implementation of Spectral Correlation Density Function",
abstract = "In this study, the parallelization of a critical statistical feature of communication signals called the spectral correlation density (SCD) is investigated. The SCD is used for synchronization in OFDM-based systems such as LTE and Wi-Fi, but is also proposed for use in next-generation wireless systems where accurate signal classification is needed even under poor channel conditions. By leveraging cyclostationary theory and classification results, a method for reducing the computational complexity of estimating the SCD for classification purposes by 75{\%} or more using the Quarter SCD (QSCD) is proposed. We parallelize the SCD and QSCD implementations by targeting general purpose graphics processing unit (GPU) through architecture specific optimization strategies. We present experimental evaluations on identifying the parallelization configuration for maximizing the efficiency of the program architecture in utilizing the threading power of the GPU architecture. We show that algorithmic and architecture specific optimization strategies result with improving the throughput of the state of the art GPU based SCD implementation from 120 signals/second to 3300 signals/second.",
keywords = "FFT accumulation method, GPGPU, Signal classification, Spectral correlation density",
author = "Scott Marshall and Garrett Vanhoy and Ali Akoglu and Tamal Bose and Bo Ryu",
year = "2019",
month = "1",
day = "1",
doi = "10.1007/s11265-019-01448-7",
language = "English (US)",
journal = "Journal of Signal Processing Systems",
issn = "1939-8018",
publisher = "Springer New York",

}

TY - JOUR

T1 - GPGPU Based Parallel Implementation of Spectral Correlation Density Function

AU - Marshall, Scott

AU - Vanhoy, Garrett

AU - Akoglu, Ali

AU - Bose, Tamal

AU - Ryu, Bo

PY - 2019/1/1

Y1 - 2019/1/1

N2 - In this study, the parallelization of a critical statistical feature of communication signals called the spectral correlation density (SCD) is investigated. The SCD is used for synchronization in OFDM-based systems such as LTE and Wi-Fi, but is also proposed for use in next-generation wireless systems where accurate signal classification is needed even under poor channel conditions. By leveraging cyclostationary theory and classification results, a method for reducing the computational complexity of estimating the SCD for classification purposes by 75% or more using the Quarter SCD (QSCD) is proposed. We parallelize the SCD and QSCD implementations by targeting general purpose graphics processing unit (GPU) through architecture specific optimization strategies. We present experimental evaluations on identifying the parallelization configuration for maximizing the efficiency of the program architecture in utilizing the threading power of the GPU architecture. We show that algorithmic and architecture specific optimization strategies result with improving the throughput of the state of the art GPU based SCD implementation from 120 signals/second to 3300 signals/second.

AB - In this study, the parallelization of a critical statistical feature of communication signals called the spectral correlation density (SCD) is investigated. The SCD is used for synchronization in OFDM-based systems such as LTE and Wi-Fi, but is also proposed for use in next-generation wireless systems where accurate signal classification is needed even under poor channel conditions. By leveraging cyclostationary theory and classification results, a method for reducing the computational complexity of estimating the SCD for classification purposes by 75% or more using the Quarter SCD (QSCD) is proposed. We parallelize the SCD and QSCD implementations by targeting general purpose graphics processing unit (GPU) through architecture specific optimization strategies. We present experimental evaluations on identifying the parallelization configuration for maximizing the efficiency of the program architecture in utilizing the threading power of the GPU architecture. We show that algorithmic and architecture specific optimization strategies result with improving the throughput of the state of the art GPU based SCD implementation from 120 signals/second to 3300 signals/second.

KW - FFT accumulation method

KW - GPGPU

KW - Signal classification

KW - Spectral correlation density

UR - http://www.scopus.com/inward/record.url?scp=85064904105&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85064904105&partnerID=8YFLogxK

U2 - 10.1007/s11265-019-01448-7

DO - 10.1007/s11265-019-01448-7

M3 - Article

JO - Journal of Signal Processing Systems

JF - Journal of Signal Processing Systems

SN - 1939-8018

ER -