TY - GEN
T1 - Advanced Modeling and Analysis of Individual and Combined TSN Shapers in OMNeT++
AU - Debnath, Rubi
AU - Hortig, Philipp
AU - Zhao, Luxi
AU - Steinhorst, Sebastian
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - The selection of a Time-Sensitive Networking (TSN) shaping mechanism is a crucial design decision that impacts the Quality of Service (QoS) of applications and configuration complexity. However, current research has mainly evaluated TSN shapers individually, despite them being designed to work together in an egress port. Hence, the lack of investigation of the mixed TSN shaping mechanism is a major limitation of the current state of the art. Combined TSN traffic shaping provides greater flexibility to improve QoS than individual shapers, making it particularly beneficial for real-time applications. This paper aims to bridge this research gap by implementing the Asynchronous Traffic Shaper (ATS) in a plug-and-play manner, enabling its use individually or in combination with other TSN shapers. We propose various models of mixed TSN shaper architectures and implement the frozen and non-frozen credit behavior of the Time Aware Shaper (TAS) + Credit Based Shaper (CBS) during the guard band (GB) using OMNeT++. Furthermore, we compare the simulation results of ATS and CBS with the Network Calculus (NC) upper bounds. Our results indicate that the simulated delays (SMDs) were significantly lower than the theoretical worst-case delays (WCDs) obtained from the NC, indicating the need for tighter theoretical upper bounds, particularly for higher network loads. To the best of our knowledge, we are the first to provide simulation-based performance analysis of the combined TAS+ATS+CBS and TAS+ATS+Strict Priority (SP) architecture. Overall, this paper highlights the benefits of combining TSN shapers and encourages further research into the potential advantages of utilizing multiple shapers simultaneously to decrease reliance on TAS and CBS.
AB - The selection of a Time-Sensitive Networking (TSN) shaping mechanism is a crucial design decision that impacts the Quality of Service (QoS) of applications and configuration complexity. However, current research has mainly evaluated TSN shapers individually, despite them being designed to work together in an egress port. Hence, the lack of investigation of the mixed TSN shaping mechanism is a major limitation of the current state of the art. Combined TSN traffic shaping provides greater flexibility to improve QoS than individual shapers, making it particularly beneficial for real-time applications. This paper aims to bridge this research gap by implementing the Asynchronous Traffic Shaper (ATS) in a plug-and-play manner, enabling its use individually or in combination with other TSN shapers. We propose various models of mixed TSN shaper architectures and implement the frozen and non-frozen credit behavior of the Time Aware Shaper (TAS) + Credit Based Shaper (CBS) during the guard band (GB) using OMNeT++. Furthermore, we compare the simulation results of ATS and CBS with the Network Calculus (NC) upper bounds. Our results indicate that the simulated delays (SMDs) were significantly lower than the theoretical worst-case delays (WCDs) obtained from the NC, indicating the need for tighter theoretical upper bounds, particularly for higher network loads. To the best of our knowledge, we are the first to provide simulation-based performance analysis of the combined TAS+ATS+CBS and TAS+ATS+Strict Priority (SP) architecture. Overall, this paper highlights the benefits of combining TSN shapers and encourages further research into the potential advantages of utilizing multiple shapers simultaneously to decrease reliance on TAS and CBS.
KW - OMNeT++
KW - asynchronous traffic shaper
KW - audio-video bridging
KW - credit based shaper
KW - performance analysis
KW - real-time networks
KW - time aware shaper
KW - time sensitive network
UR - http://www.scopus.com/inward/record.url?scp=85178029250&partnerID=8YFLogxK
U2 - 10.1109/RTCSA58653.2023.00029
DO - 10.1109/RTCSA58653.2023.00029
M3 - Conference contribution
AN - SCOPUS:85178029250
T3 - Proceedings - 2023 IEEE 29th International Conference on Embedded and Real-Time Computing Systems and Applications, RTCSA 2023
SP - 176
EP - 185
BT - Proceedings - 2023 IEEE 29th International Conference on Embedded and Real-Time Computing Systems and Applications, RTCSA 2023
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 29th IEEE International Conference on Embedded and Real-Time Computing Systems and Applications, RTCSA 2023
Y2 - 30 August 2023 through 1 September 2023
ER -