DICES: A dynamic adaptive service-driven SDN architecture

The SDNs promise is to provide flexibility, programmability and scalability, while reducing the infrastructure costs to Networks Operators (NO) by centralizing and softwarizing the control plane. Nonetheless, this centralization process goes against the distribution of the control that prevailed, until now, in telecommunication networks. Despite the hopes it raises, it brings new issues mainly related to the scalability and the reliability of this kind of architectures. Drawing on SDN's concepts, this paper's purpose is to build an innovative network architecture that is: programmable, flexible, scalable and aware of the services' SLA. To achieve this goal, we proposed a new architecture called DICES, which includes several concepts. Firstly, a basic controller is integrated as a generic function on the highest layer of all the network elements. Secondly, the controllers instances are activated and deployed thanks to a network orchestrator entity according to the requested SLA. Thirdly, the network services are composed and modeled using Petri nets, and validated before their deployment and execution on the controller level.

[1]  Ali Moeini,et al.  Simulating Turing Machines Using Colored Petri Nets with Priority Transitions , 2013 .

[2]  François Vernadat,et al.  Time Petri Nets Analysis with TINA , 2006, Third International Conference on the Quantitative Evaluation of Systems - (QEST'06).

[3]  Anja Feldmann,et al.  Logically centralized?: state distribution trade-offs in software defined networks , 2012, HotSDN '12.

[4]  Paul Müller Software Defined Networking: Bridging the gap between distributed-systems and networked-systems research , 2013, DFN-Forum Kommunikationstechnologien.

[5]  I. Pavesio,et al.  Using modular Petri nets for developing telecommunication software , 1994, Second Workshop on Parallel and Distributed Real-Time Systems.

[6]  Baek-Young Choi,et al.  Reliability and Scalability Issues in Software Defined Network Frameworks , 2013, 2013 Second GENI Research and Educational Experiment Workshop.

[7]  Rob Sherwood,et al.  FlowVisor: A Network Virtualization Layer , 2009 .

[8]  R.P. Moreno,et al.  Adaptive Petri Nets implementation. The execution time controller , 2008, 2008 9th International Workshop on Discrete Event Systems.

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

[10]  Yashar Ganjali,et al.  Kandoo: a framework for efficient and scalable offloading of control applications , 2012, HotSDN '12.

[11]  Matthew Roughan,et al.  The Internet Topology Zoo , 2011, IEEE Journal on Selected Areas in Communications.

[12]  Fang Hao,et al.  Towards an elastic distributed SDN controller , 2013, HotSDN '13.