Analysis and Optimization of Hybrid Software-Defined Networks

Hybrid IP networks that use both control plane paradigms - distributed and centralized - promise the best of two worlds: programmability and flexible control of Software-Defined Networking (SDN), and at the same time the reliability and fault tolerance of distributed routing protocols like Open Shortest Path First (OSPF). Hybrid SDN/OSPF networks typically deploy OSPF to assure care-free operation of best effort traffic, while SDN can control prioritized traffic. This "ships-passing-in-the-night" approach, where both control planes are unaware of each other's configurations, only require hybrid SDN/OSPF routers that can participate in the domain-wide legacy routing protocol and additionally connect to a central SDN controller. This mode of operation is however known for a number of challenges in operational networks, including those related to network failures, size of forwarding tables, routing convergence time, and the increased complexity of network management. There are alternative modes of hybrid operation that provide a more holistic network control paradigm, either through an OSPF-enabled SDN controller, or a common network management system that allows the joint monitoring and configuration of both control planes, or via the partitioning of the legacy routing domain with SDN border nodes. The latter mode of operation offers to some extent to steer the working of the legacy routing protocol inside the sub-domains, which is new. The analysis, modeling, and evaluative comparison of this approach called SDN Partitioning with other modes of operation is the main contribution of this thesis. This thesis addresses important network planning tasks in hybrid SDN/OSPF networks and provides the according mathematical models to optimize network clustering, capacity planning, SDN node placement, and resource provisioning for a fault tolerant operation. It furthermore provides the mathematical models to optimize traffic engineering, failure recovery, reconfiguration scheduling, and traffic monitoring in hybrid SDN/OSPF networks, which are vital network operational tasks.

[1]  Albert G. Greenberg,et al.  Fast accurate computation of large-scale IP traffic matrices from link loads , 2003, SIGMETRICS '03.

[2]  Konstantina Papagiannaki,et al.  A distributed approach to measure IP traffic matrices , 2004, IMC '04.

[3]  Admela Jukan,et al.  A performance study of network migration to SDN-enabled Traffic Engineering , 2013, 2013 IEEE Global Communications Conference (GLOBECOM).

[4]  Biswanath Mukherjee,et al.  Optical WDM Networks (Optical Networks) , 2006 .

[5]  Olivier Bonaventure,et al.  On the co-existence of distributed and centralized routing control-planes , 2015, 2015 IEEE Conference on Computer Communications (INFOCOM).

[6]  S. Rajagopalan,et al.  Adoption timing of new equipment with another innovation anticipated , 1999 .

[7]  Olivier Bonaventure,et al.  Opportunities and research challenges of hybrid software defined networks , 2014, CCRV.

[8]  Christophe Diot,et al.  Reformulating the Monitor Placement Problem: Optimal Network-Wide Sampling , 2006 .

[9]  Laurent Vanbever,et al.  Central Control Over Distributed Routing , 2015, Comput. Commun. Rev..

[10]  Biswanath Mukherjee,et al.  Energy optimization in IP-over-WDM networks , 2011, Opt. Switch. Netw..

[11]  Dominic A. Schupke,et al.  Impact and Handling of Demand Uncertainty in Multiperiod Planned Networks , 2011, 2011 IEEE International Conference on Communications (ICC).

[12]  Anja Feldmann,et al.  Incremental SDN deployment in enterprise networks , 2013, Conference on Applications, Technologies, Architectures, and Protocols for Computer Communication.

[13]  Fernando A. Kuipers,et al.  Traffic uncertainty models in network planning , 2014, IEEE Communications Magazine.

[14]  Sheng Wang,et al.  Towards accurate online traffic matrix estimation in software-defined networks , 2015, SOSR.

[15]  Jörn Altmann,et al.  Dynamic netvalue analyzer - A pricing plan modeling tool for isps using actual network usage data , 2002, Proceedings Fourth IEEE International Workshop on Advanced Issues of E-Commerce and Web-Based Information Systems (WECWIS 2002).

[16]  Laurent Vanbever,et al.  Sweet Little Lies: Fake Topologies for Flexible Routing , 2014, HotNets.

[17]  Sofie Verbrugge.  Strategic planning of optical telecommunication networks in a dynamic and uncertain environment , 2007 .

[18]  John Moy,et al.  OSPF Version 2 , 1998, RFC.

[19]  Rob Sherwood,et al.  The controller placement problem , 2012, HotSDN@SIGCOMM.

[20]  Yashar Ganjali,et al.  On scalability of software-defined networking , 2013, IEEE Communications Magazine.

[21]  Chen-Nee Chuah,et al.  MeasuRouting: A Framework for Routing Assisted Traffic Monitoring , 2010, IEEE/ACM Transactions on Networking.

[22]  Christophe Diot,et al.  Design of IGP link weight changes for estimation of traffic matrices , 2004, IEEE INFOCOM 2004.

[23]  Y. Vardi,et al.  Network Tomography: Estimating Source-Destination Traffic Intensities from Link Data , 1996 .

[24]  Andrew B. Kahng,et al.  Recent directions in netlist partitioning: a survey , 1995, Integr..

[25]  Murali S. Kodialam,et al.  Traffic engineering in software defined networks , 2013, 2013 Proceedings IEEE INFOCOM.

[26]  Konstantina Papagiannaki,et al.  Traffic matrices: balancing measurements, inference and modeling , 2005, SIGMETRICS '05.

[27]  Prasant Mohapatra,et al.  Simultaneously Reducing Latency and Power Consumption in OpenFlow Switches , 2014, IEEE/ACM Transactions on Networking.

[28]  Ralf Lehnert,et al.  Network migration using ant colony optimization , 2010, 2010 9th Conference of Telecommunication, Media and Internet.

[29]  Olivier Bonaventure,et al.  Interdomain traffic engineering with BGP , 2003, IEEE Commun. Mag..

[30]  Jun Luo,et al.  Cracking network monitoring in DCNs with SDN , 2015, 2015 IEEE Conference on Computer Communications (INFOCOM).

[31]  Jennifer Widom,et al.  Teletraffic modeling for personal communications services , 1997 .

[32]  Egon Balas,et al.  The vertex separator problem: a polyhedral investigation , 2005, Math. Program..

[33]  Min Zhu,et al.  B4: experience with a globally-deployed software defined wan , 2013, SIGCOMM.

[34]  Emilio Leonardi,et al.  Estimating dynamic traffic matrices by using viable routing changes , 2007 .

[35]  Marco Listanti,et al.  The Power of SDN to Improve the Estimation of the ISP Traffic Matrix Through the Flow Spread Concept , 2016, IEEE Journal on Selected Areas in Communications.

[36]  Aiko Pras,et al.  Assessing the Quality of Flow Measurements from OpenFlow Devices , 2016, TMA.

[37]  Mikkel Thorup,et al.  Internet traffic engineering by optimizing OSPF weights , 2000, Proceedings IEEE INFOCOM 2000. Conference on Computer Communications. Nineteenth Annual Joint Conference of the IEEE Computer and Communications Societies (Cat. No.00CH37064).

[38]  Mohit Chamania,et al.  Achieving IP routing stability with optical bypass , 2009 .

[39]  David R. Oran,et al.  OSI IS-IS Intra-domain Routing Protocol , 1990, RFC.