HLS based hardware acceleration on the zynq SoC: A case study for fall detection system

A. A. S. Ali, M. Siupik, A. Amira, F. Bensaali, Juan Pablo Casaseca

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Fall detection is a major problem in healthcare systems, especially for elderly people who are the most vulnerable. It is important to design and implement not only an accurate fall detection system (FDS) but also a system with a real-time response. The achievement of high accuracy and fast response time together allows the development of a system that helps saving lives, time and money in healthcare industry. This paper presents the design, simulation and implementation of a novel FDS using the Shimmer wearable sensor. The discrete wavelet transform (DWT) is applied for preprocessing the data coming from the Shimmer platform, principal component analysis (PCA) is used for dimensionality reduction and feature extraction and finally, a binary decision tree (DT) is utilized for classification purpose. The system is simulated in MATLAB prior to the implementation on the Zynq system-on-chip (SoC) for hardware acceleration. DWT is executed on the processing system (PS) of the Zynq platform in a software manner while PCA and DT are both implemented on the programmable logic (PL) for hardware acceleration. PCA and DT are developed in C and synthesized in Vivado high level synthesis (HLS) tool to transform the C based designed into a register transfer level (RTL) implementation. Various optimization techniques are explored in Vivado HLS. The performance of the FDS in terms of accuracy of the classifier is 88.4% while the overall resources used in PL of the Zynq vary between 2% and 23% depending on the running frequency and optimization technique used.
Original languageEnglish
Title of host publicationIEEE/ACS 11th International Conference on Computer Systems and Applications (AICCSA), 2014
PublisherIEEE
Pages685-690
Number of pages6
ISBN (Print)9781479971008
DOIs
Publication statusPublished - 2014

Fingerprint

Decision trees
Principal component analysis
Discrete wavelet transforms
Hardware
MATLAB
Feature extraction
Classifiers
Processing
High level synthesis
System-on-chip
Industry

Keywords

  • assisted living
  • discrete wavelet transforms
  • feature extraction
  • geriatrics
  • high level synthesis
  • principal component analysis
  • system-on-chip
  • DWT
  • FDS
  • HLS based hardware acceleration
  • PCA
  • RTL implementation
  • Shimmer platform
  • Shimmer wearable sensor
  • Vivado HLS tool
  • Vivado high level synthesis tool
  • Zynq SoC
  • Zynq system-on-chip
  • binary DT
  • binary decision tree
  • dimensionality reduction
  • discrete wavelet transform
  • elderly people
  • fall detection system
  • healthcare industry
  • healthcare systems
  • programmable logic
  • real-time response system
  • register transfer level implementation
  • Acceleration
  • Decision trees
  • Hardware
  • Training
  • decision tree classifier

Cite this

Ali, A. A. S., Siupik, M., Amira, A., Bensaali, F., & Casaseca, J. P. (2014). HLS based hardware acceleration on the zynq SoC: A case study for fall detection system. In IEEE/ACS 11th International Conference on Computer Systems and Applications (AICCSA), 2014 (pp. 685-690). IEEE. https://doi.org/10.1109/AICCSA.2014.7073266
Ali, A. A. S. ; Siupik, M. ; Amira, A. ; Bensaali, F. ; Casaseca, Juan Pablo. / HLS based hardware acceleration on the zynq SoC : A case study for fall detection system. IEEE/ACS 11th International Conference on Computer Systems and Applications (AICCSA), 2014. IEEE, 2014. pp. 685-690
@inproceedings{0694e1882a3e4e5b8b2bea29e0bab920,
title = "HLS based hardware acceleration on the zynq SoC: A case study for fall detection system",
abstract = "Fall detection is a major problem in healthcare systems, especially for elderly people who are the most vulnerable. It is important to design and implement not only an accurate fall detection system (FDS) but also a system with a real-time response. The achievement of high accuracy and fast response time together allows the development of a system that helps saving lives, time and money in healthcare industry. This paper presents the design, simulation and implementation of a novel FDS using the Shimmer wearable sensor. The discrete wavelet transform (DWT) is applied for preprocessing the data coming from the Shimmer platform, principal component analysis (PCA) is used for dimensionality reduction and feature extraction and finally, a binary decision tree (DT) is utilized for classification purpose. The system is simulated in MATLAB prior to the implementation on the Zynq system-on-chip (SoC) for hardware acceleration. DWT is executed on the processing system (PS) of the Zynq platform in a software manner while PCA and DT are both implemented on the programmable logic (PL) for hardware acceleration. PCA and DT are developed in C and synthesized in Vivado high level synthesis (HLS) tool to transform the C based designed into a register transfer level (RTL) implementation. Various optimization techniques are explored in Vivado HLS. The performance of the FDS in terms of accuracy of the classifier is 88.4{\%} while the overall resources used in PL of the Zynq vary between 2{\%} and 23{\%} depending on the running frequency and optimization technique used.",
keywords = "assisted living, discrete wavelet transforms, feature extraction, geriatrics, high level synthesis, principal component analysis, system-on-chip, DWT, FDS, HLS based hardware acceleration, PCA, RTL implementation, Shimmer platform, Shimmer wearable sensor, Vivado HLS tool, Vivado high level synthesis tool, Zynq SoC, Zynq system-on-chip, binary DT, binary decision tree, dimensionality reduction, discrete wavelet transform, elderly people, fall detection system, healthcare industry, healthcare systems, programmable logic, real-time response system, register transfer level implementation, Acceleration, Decision trees, Hardware, Training, decision tree classifier",
author = "Ali, {A. A. S.} and M. Siupik and A. Amira and F. Bensaali and Casaseca, {Juan Pablo}",
year = "2014",
doi = "10.1109/AICCSA.2014.7073266",
language = "English",
isbn = "9781479971008",
pages = "685--690",
booktitle = "IEEE/ACS 11th International Conference on Computer Systems and Applications (AICCSA), 2014",
publisher = "IEEE",
address = "United States",

}

Ali, AAS, Siupik, M, Amira, A, Bensaali, F & Casaseca, JP 2014, HLS based hardware acceleration on the zynq SoC: A case study for fall detection system. in IEEE/ACS 11th International Conference on Computer Systems and Applications (AICCSA), 2014. IEEE, pp. 685-690. https://doi.org/10.1109/AICCSA.2014.7073266

HLS based hardware acceleration on the zynq SoC : A case study for fall detection system. / Ali, A. A. S.; Siupik, M.; Amira, A.; Bensaali, F.; Casaseca, Juan Pablo.

IEEE/ACS 11th International Conference on Computer Systems and Applications (AICCSA), 2014. IEEE, 2014. p. 685-690.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - HLS based hardware acceleration on the zynq SoC

T2 - A case study for fall detection system

AU - Ali, A. A. S.

AU - Siupik, M.

AU - Amira, A.

AU - Bensaali, F.

AU - Casaseca, Juan Pablo

PY - 2014

Y1 - 2014

N2 - Fall detection is a major problem in healthcare systems, especially for elderly people who are the most vulnerable. It is important to design and implement not only an accurate fall detection system (FDS) but also a system with a real-time response. The achievement of high accuracy and fast response time together allows the development of a system that helps saving lives, time and money in healthcare industry. This paper presents the design, simulation and implementation of a novel FDS using the Shimmer wearable sensor. The discrete wavelet transform (DWT) is applied for preprocessing the data coming from the Shimmer platform, principal component analysis (PCA) is used for dimensionality reduction and feature extraction and finally, a binary decision tree (DT) is utilized for classification purpose. The system is simulated in MATLAB prior to the implementation on the Zynq system-on-chip (SoC) for hardware acceleration. DWT is executed on the processing system (PS) of the Zynq platform in a software manner while PCA and DT are both implemented on the programmable logic (PL) for hardware acceleration. PCA and DT are developed in C and synthesized in Vivado high level synthesis (HLS) tool to transform the C based designed into a register transfer level (RTL) implementation. Various optimization techniques are explored in Vivado HLS. The performance of the FDS in terms of accuracy of the classifier is 88.4% while the overall resources used in PL of the Zynq vary between 2% and 23% depending on the running frequency and optimization technique used.

AB - Fall detection is a major problem in healthcare systems, especially for elderly people who are the most vulnerable. It is important to design and implement not only an accurate fall detection system (FDS) but also a system with a real-time response. The achievement of high accuracy and fast response time together allows the development of a system that helps saving lives, time and money in healthcare industry. This paper presents the design, simulation and implementation of a novel FDS using the Shimmer wearable sensor. The discrete wavelet transform (DWT) is applied for preprocessing the data coming from the Shimmer platform, principal component analysis (PCA) is used for dimensionality reduction and feature extraction and finally, a binary decision tree (DT) is utilized for classification purpose. The system is simulated in MATLAB prior to the implementation on the Zynq system-on-chip (SoC) for hardware acceleration. DWT is executed on the processing system (PS) of the Zynq platform in a software manner while PCA and DT are both implemented on the programmable logic (PL) for hardware acceleration. PCA and DT are developed in C and synthesized in Vivado high level synthesis (HLS) tool to transform the C based designed into a register transfer level (RTL) implementation. Various optimization techniques are explored in Vivado HLS. The performance of the FDS in terms of accuracy of the classifier is 88.4% while the overall resources used in PL of the Zynq vary between 2% and 23% depending on the running frequency and optimization technique used.

KW - assisted living

KW - discrete wavelet transforms

KW - feature extraction

KW - geriatrics

KW - high level synthesis

KW - principal component analysis

KW - system-on-chip

KW - DWT

KW - FDS

KW - HLS based hardware acceleration

KW - PCA

KW - RTL implementation

KW - Shimmer platform

KW - Shimmer wearable sensor

KW - Vivado HLS tool

KW - Vivado high level synthesis tool

KW - Zynq SoC

KW - Zynq system-on-chip

KW - binary DT

KW - binary decision tree

KW - dimensionality reduction

KW - discrete wavelet transform

KW - elderly people

KW - fall detection system

KW - healthcare industry

KW - healthcare systems

KW - programmable logic

KW - real-time response system

KW - register transfer level implementation

KW - Acceleration

KW - Decision trees

KW - Hardware

KW - Training

KW - decision tree classifier

U2 - 10.1109/AICCSA.2014.7073266

DO - 10.1109/AICCSA.2014.7073266

M3 - Conference contribution

SN - 9781479971008

SP - 685

EP - 690

BT - IEEE/ACS 11th International Conference on Computer Systems and Applications (AICCSA), 2014

PB - IEEE

ER -

Ali AAS, Siupik M, Amira A, Bensaali F, Casaseca JP. HLS based hardware acceleration on the zynq SoC: A case study for fall detection system. In IEEE/ACS 11th International Conference on Computer Systems and Applications (AICCSA), 2014. IEEE. 2014. p. 685-690 https://doi.org/10.1109/AICCSA.2014.7073266