Mapping TSN Traffic Scheduling and Shaping to FPGA-Based Architecture

Time-Sensitive Networking (TSN), which evolves from the Ethernet standards, has been developed to ensure deterministic transmission in data networks. Asynchronous Traffic Shaping (ATS) extends the conventional synchronized TSN with an asynchronous scheduler to guarantee a bounded transmitting delay....

Full description

Bibliographic Details
Main Authors: Zifan Zhou, Michael Stubert Berger, Ying Yan
Format: Article
Language:English
Published: IEEE 2020-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/9290032/
id doaj-631678b6d17444489d8e67094db9c311
record_format Article
spelling doaj-631678b6d17444489d8e67094db9c3112021-03-30T04:45:59ZengIEEEIEEE Access2169-35362020-01-01822150322151210.1109/ACCESS.2020.30438879290032Mapping TSN Traffic Scheduling and Shaping to FPGA-Based ArchitectureZifan Zhou0https://orcid.org/0000-0003-4251-4889Michael Stubert Berger1Ying Yan2https://orcid.org/0000-0003-0678-8831Department of Photonics Engineering, Technical University of Denmark, Kongens Lyngby, DenmarkDepartment of Photonics Engineering, Technical University of Denmark, Kongens Lyngby, DenmarkDepartment of Photonics Engineering, Technical University of Denmark, Kongens Lyngby, DenmarkTime-Sensitive Networking (TSN), which evolves from the Ethernet standards, has been developed to ensure deterministic transmission in data networks. Asynchronous Traffic Shaping (ATS) extends the conventional synchronized TSN with an asynchronous scheduler to guarantee a bounded transmitting delay. In this work, we present a Field Programmable Gate Arrays (FPGA) implementation of a TSN scheduling entity, which leverages ATS for the frame forwarding process. We explore the ATS design by function blocks and compare it with a benchmark design utilizing strict-priority scheduling. In terms of operating frequency, our results indicate that strict-priority scheduling performs 1.05% to 9.56% higher maximum frequency than ATS with the same configurations. Regarding resource utilization, ATS consumes 51% to 119% more logic blocks and 51% to 101% more registers than strict-priority scheduling. Based on the synthesis and fitting results from Register-Transfer Level (RTL) simulations, we provide a general vision of designing and implementing considerations of the ATS function. Specifically, we show the influences of the buffer and bus width configuration on the FPGA implementation scale and data rate.https://ieeexplore.ieee.org/document/9290032/Ethernet networksreal-time systemshardwarescheduling algorithmFPGA
collection DOAJ
language English
format Article
sources DOAJ
author Zifan Zhou
Michael Stubert Berger
Ying Yan
spellingShingle Zifan Zhou
Michael Stubert Berger
Ying Yan
Mapping TSN Traffic Scheduling and Shaping to FPGA-Based Architecture
IEEE Access
Ethernet networks
real-time systems
hardware
scheduling algorithm
FPGA
author_facet Zifan Zhou
Michael Stubert Berger
Ying Yan
author_sort Zifan Zhou
title Mapping TSN Traffic Scheduling and Shaping to FPGA-Based Architecture
title_short Mapping TSN Traffic Scheduling and Shaping to FPGA-Based Architecture
title_full Mapping TSN Traffic Scheduling and Shaping to FPGA-Based Architecture
title_fullStr Mapping TSN Traffic Scheduling and Shaping to FPGA-Based Architecture
title_full_unstemmed Mapping TSN Traffic Scheduling and Shaping to FPGA-Based Architecture
title_sort mapping tsn traffic scheduling and shaping to fpga-based architecture
publisher IEEE
series IEEE Access
issn 2169-3536
publishDate 2020-01-01
description Time-Sensitive Networking (TSN), which evolves from the Ethernet standards, has been developed to ensure deterministic transmission in data networks. Asynchronous Traffic Shaping (ATS) extends the conventional synchronized TSN with an asynchronous scheduler to guarantee a bounded transmitting delay. In this work, we present a Field Programmable Gate Arrays (FPGA) implementation of a TSN scheduling entity, which leverages ATS for the frame forwarding process. We explore the ATS design by function blocks and compare it with a benchmark design utilizing strict-priority scheduling. In terms of operating frequency, our results indicate that strict-priority scheduling performs 1.05% to 9.56% higher maximum frequency than ATS with the same configurations. Regarding resource utilization, ATS consumes 51% to 119% more logic blocks and 51% to 101% more registers than strict-priority scheduling. Based on the synthesis and fitting results from Register-Transfer Level (RTL) simulations, we provide a general vision of designing and implementing considerations of the ATS function. Specifically, we show the influences of the buffer and bus width configuration on the FPGA implementation scale and data rate.
topic Ethernet networks
real-time systems
hardware
scheduling algorithm
FPGA
url https://ieeexplore.ieee.org/document/9290032/
work_keys_str_mv AT zifanzhou mappingtsntrafficschedulingandshapingtofpgabasedarchitecture
AT michaelstubertberger mappingtsntrafficschedulingandshapingtofpgabasedarchitecture
AT yingyan mappingtsntrafficschedulingandshapingtofpgabasedarchitecture
_version_ 1724181222845841408