Limiting Worst-Case End-to-End Latency When Traffic Increases in a Switched Avionics Network
M. Nam, E. Seo, L. Sha, K. Park, and K. Kang. Embedded and Real-Time Computing Systems and Applications (RTCSA), 2011 IEEE 17th International Conference on, 1, page 285-294. (August 2011)
DOI: 10.1109/RTCSA.2011.9
Abstract
New features are often added incrementally to avionics systems. This avoids redesign and recertification but still requires verifying the timing constraints of both new and existing applications. We introduce a new switch that facilitates this verification by bounding the latency of end-to-end communication across a network. Our clock-driven real-time switching algorithm is throughput-optimal with a bounded worst-case delay for all feasible traffic. Associated heuristics can verify whether the timing constraints of an avionics network are met, after new features have caused traffic to increase, and then search for alternative network configurations if necessary. We show how these heuristics cope with changes to an example environmental monitoring architecture within an avionics system that incorporates our switch. Our approach to analysis can be used to determine, quickly but rigorously, which system architecture meet timing constraints, and it allows the system architect to manage the cascading effects of component changes in a comprehensive manner.
Description
IEEE Xplore Abstract - Limiting Worst-Case End-to-End Latency When Traffic Increases in a Switched Avionics Network
%0 Conference Paper
%1 nam2011
%A Nam, Min-Young
%A Seo, Eunsoo
%A Sha, Lui
%A Park, Kyung-Joon
%A Kang, Kyungtae
%B Embedded and Real-Time Computing Systems and Applications (RTCSA), 2011 IEEE 17th International Conference on
%D 2011
%K case noc worst
%P 285-294
%R 10.1109/RTCSA.2011.9
%T Limiting Worst-Case End-to-End Latency When Traffic Increases in a Switched Avionics Network
%U http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6029839
%V 1
%X New features are often added incrementally to avionics systems. This avoids redesign and recertification but still requires verifying the timing constraints of both new and existing applications. We introduce a new switch that facilitates this verification by bounding the latency of end-to-end communication across a network. Our clock-driven real-time switching algorithm is throughput-optimal with a bounded worst-case delay for all feasible traffic. Associated heuristics can verify whether the timing constraints of an avionics network are met, after new features have caused traffic to increase, and then search for alternative network configurations if necessary. We show how these heuristics cope with changes to an example environmental monitoring architecture within an avionics system that incorporates our switch. Our approach to analysis can be used to determine, quickly but rigorously, which system architecture meet timing constraints, and it allows the system architect to manage the cascading effects of component changes in a comprehensive manner.
@inproceedings{nam2011,
abstract = {New features are often added incrementally to avionics systems. This avoids redesign and recertification but still requires verifying the timing constraints of both new and existing applications. We introduce a new switch that facilitates this verification by bounding the latency of end-to-end communication across a network. Our clock-driven real-time switching algorithm is throughput-optimal with a bounded worst-case delay for all feasible traffic. Associated heuristics can verify whether the timing constraints of an avionics network are met, after new features have caused traffic to increase, and then search for alternative network configurations if necessary. We show how these heuristics cope with changes to an example environmental monitoring architecture within an avionics system that incorporates our switch. Our approach to analysis can be used to determine, quickly but rigorously, which system architecture meet timing constraints, and it allows the system architect to manage the cascading effects of component changes in a comprehensive manner.},
added-at = {2014-08-12T11:12:57.000+0200},
author = {Nam, Min-Young and Seo, Eunsoo and Sha, Lui and Park, Kyung-Joon and Kang, Kyungtae},
biburl = {https://www.bibsonomy.org/bibtex/2274607424adc25dfdc28955efe6552e3/eberle18},
booktitle = {Embedded and Real-Time Computing Systems and Applications (RTCSA), 2011 IEEE 17th International Conference on},
description = {IEEE Xplore Abstract - Limiting Worst-Case End-to-End Latency When Traffic Increases in a Switched Avionics Network},
doi = {10.1109/RTCSA.2011.9},
interhash = {def0787dfed89017d5b8fb267bebf996},
intrahash = {274607424adc25dfdc28955efe6552e3},
issn = {1533-2306},
keywords = {case noc worst},
month = aug,
pages = {285-294},
timestamp = {2014-08-12T11:12:57.000+0200},
title = {Limiting Worst-Case End-to-End Latency When Traffic Increases in a Switched Avionics Network},
url = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6029839},
volume = 1,
year = 2011
}