TY - JOUR
T1 - Unicast QoS Routing Algorithms for SDN
T2 - A Comprehensive Survey and Performance Evaluation
AU - Guck, Jochen W.
AU - Van Bemten, Amaury
AU - Reisslein, Martin
AU - Kellerer, Wolfgang
N1 - Funding Information:
Manuscript received March 7, 2017; revised July 24, 2017; accepted August 29, 2017. Date of publication September 7, 2017; date of current version February 26, 2018. This work was supported in part by the European Union’s Horizon 2020 Research and Innovation Programme under Grant 671648 (VirtuWind), in part by ERC Grant 647158 (FlexNets), in part by the Alexander von Humboldt Foundation through a Friedrich Wilhelm Bessel Research Award, and in part by the U.S. National Science Foundation under Grant #1716121. (Corresponding author: Wolfgang Kellerer.) J. W. Guck, A. Van Bemten, and W. Kellerer are with the Lehrstuhl für Kommunikationsnetze, Technical University of Munich, 80290 Munich, Germany (e-mail: guck@tum.de; amaury.van-bemten@tum.de; wolfgang.kellerer@tum.de).
Publisher Copyright:
© 2017 IEEE.
PY - 2018/1/1
Y1 - 2018/1/1
N2 - A variety of communication networks, such as industrial communication systems, have to provide strict delay guarantees to the carried flows. Fast and close to optimal quality of service (QoS) routing algorithms, e.g., delay-constrained least-cost (DCLC) routing algorithms, are required for routing flows in such networks with strict delay requirements. The emerging software-defined networking (SDN) paradigm centralizes the network control in SDN controllers that can centrally execute QoS routing algorithms. A wide range of QoS routing algorithms have been proposed in the literature and examined in individual studies. However, a comprehensive evaluation framework and quantitative comparison of QoS routing algorithms that can serve as a basis for selecting and further advancing QoS routing in SDN networks is missing in the literature. This makes it difficult to select the most appropriate QoS routing algorithm for a particular use case, e.g., for SDN controlled industrial communications. We close this gap in the literature by conducting a comprehensive up-to-date survey of centralized QoS routing algorithms. We introduce a novel four-dimensional (4D) evaluation framework for QoS routing algorithms, whereby the 4D correspond to the type of topology, two forms of scalability of a topology, and the tightness of the delay constraint. We implemented 26 selected DCLC algorithms and compared their runtime and cost inefficiency within the 4D evaluation framework. While the main conclusion of this evaluation is that the best algorithm depends on the specific sub-space of the 4D space that is targeted, we identify two algorithms, namely Lagrange relaxation-based aggregated cost (LARAC) and search space reduction delay-cost-constrained routing (SSR+DCCR), that perform very well in most of the 4D evaluation space.
AB - A variety of communication networks, such as industrial communication systems, have to provide strict delay guarantees to the carried flows. Fast and close to optimal quality of service (QoS) routing algorithms, e.g., delay-constrained least-cost (DCLC) routing algorithms, are required for routing flows in such networks with strict delay requirements. The emerging software-defined networking (SDN) paradigm centralizes the network control in SDN controllers that can centrally execute QoS routing algorithms. A wide range of QoS routing algorithms have been proposed in the literature and examined in individual studies. However, a comprehensive evaluation framework and quantitative comparison of QoS routing algorithms that can serve as a basis for selecting and further advancing QoS routing in SDN networks is missing in the literature. This makes it difficult to select the most appropriate QoS routing algorithm for a particular use case, e.g., for SDN controlled industrial communications. We close this gap in the literature by conducting a comprehensive up-to-date survey of centralized QoS routing algorithms. We introduce a novel four-dimensional (4D) evaluation framework for QoS routing algorithms, whereby the 4D correspond to the type of topology, two forms of scalability of a topology, and the tightness of the delay constraint. We implemented 26 selected DCLC algorithms and compared their runtime and cost inefficiency within the 4D evaluation framework. While the main conclusion of this evaluation is that the best algorithm depends on the specific sub-space of the 4D space that is targeted, we identify two algorithms, namely Lagrange relaxation-based aggregated cost (LARAC) and search space reduction delay-cost-constrained routing (SSR+DCCR), that perform very well in most of the 4D evaluation space.
KW - Delay-constrained least-cost (DCLC) routing
KW - performance evaluation framework
KW - quality of service (QoS)
KW - scalability
KW - software-defined networking (SDN)
UR - http://www.scopus.com/inward/record.url?scp=85029184651&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85029184651&partnerID=8YFLogxK
U2 - 10.1109/COMST.2017.2749760
DO - 10.1109/COMST.2017.2749760
M3 - Article
AN - SCOPUS:85029184651
SN - 1553-877X
VL - 20
SP - 388
EP - 418
JO - IEEE Communications Surveys and Tutorials
JF - IEEE Communications Surveys and Tutorials
IS - 1
ER -