Abstract-The software-defined network (SDN) advocates a centralized network control, where a controller manages a network from a global view of the network. Large SDN networks may consist of multiple controllers or controller domains that distribute the network management between them, where each controller has a logically centralized but physically distributed vision of the network. In this context, a key challenge faced by providers is to define a scalable control network that exploits the benefits of SDN when used in conjunction with efficient management strategies. Most of the control layer models proposed are not concerned with controller scalability, because they assume that commercial controllers are scalable in terms of capacity (quantity of flows processed per second). However, it has been demonstrated that overloads and long propagation delays among controllers and controllers-switches can lead to a long response time of the controllers, affecting their ability to respond to network events in a very short time and reducing the reliability of communication.In this work we define the principles for designing a scalable control layer for SDN, and show the desired control layer characteristics that optimize the management of the network. We address these principles from the perspective of the controller placement problem. For this purpose we improve and evaluate our previous approach, the algorithm called k-Critical. K-Critical discovers the minimum number of controllers and their location to create a robust control topology that deals robustly with failures and balances the load among the selected controllers. The results demonstrate the effectiveness of our solution by comparing it with other controller placement solutions.Index Terms-Software-defined network, controller scalability, control layer, controller placement problem.

This work proposes an algorithm called k-Critical\ud Node to solve the controller placement problem in Software\ud Defined Networks. K-Critical finds the minimum number of\ud controllers to satisfy a target communication delay between\ud controller and nodes, Dreq. In addition, the controller selected\ud create a management architecture that improves the subjacent\ud network performance. In this work we focus on the controllers\ud selection procedure, and show the desired management architecture\ud characteristics that optimize the control and management\ud of the network. The results show that our management\ud trees balance the load among them and reduce the data loss.Peer ReviewedPostprint (published version

Abstract-In Software Defined Networking (SDN), the network management is logically centralized but physically distributed among the controllers in order to improve the scalability compared with a completely centralized model. As an alternative to managing the network state in SDN in a distributed way, this work proposes a distributed protocol called SDN Resource Discovery Protocol (SDN-RDP). This protocol divides and distributes the network management among the controllers. In essence, each controller discovers a portion of the network topology creating a minimum-latency tree rooted at each controller, thus creating the control layer. Through the delay-constrained shortest paths, henceforth called control channels, the controllers collect network state information from nodes, and decide and distribute the forwarding decisions to them. This process is asynchronous since there is no global initialization process to activate the execution of the protocol, and knowledge about the network is not required. Simulation results show that the proposed protocol works efficiently on large networks in terms of time and load.

Abstract-Software-Defined Networking (SDN) and Network Functions Virtualization (NFV) are technologies for enabling innovative network architectures. Nevertheless, a fundamental problem in instantiation of Virtual Networks (VNs), performed by NFV, is an optimal allocation of resources offered by one or more SDN domain networks. The process of instantiation of VNs is performed in several phases, including splitting and mapping algorithms. For each one of these phases, researchers have developed algorithms, being possible to obtain different results combining them. This paper introduces a modular and flexible graphical discrete event simulation tool for solving the complete virtual resource allocation in SDN domain networks problem. A Java-based tool has been developed to integrate existing and future algorithms related to each phase of the process. The simulator is a test-bed in which researchers can select the appropriate algorithm in each phase and display the results in a graphical form, while obtaining a performance evaluation of the selected and proposed algorithms.Index Terms-Network functions virtualization, network modeling, optimization, software-defined networking. I. INTRODUCTIONTraditional network architectures are ill-suited to meet the requirements of today's enterprises, carriers, and end users. Thanks to a broad industry effort spearheaded by the Open Networking Foundation (ONF), SDN is transforming the network architecture. In the SDN architecture, the control and data planes are decoupled. Network intelligence is centralized in software-based SDN controllers, which maintain a global view of the network. As a result, enterprises and carriers gain unprecedented programmability, automation, and network control, enabling them to build highly scalable, flexible networks that readily adapt to changing business needs. The ONF is a non-profit industry consortium that is leading the advancement of SDN and standardizing critical elements of the SDN architecture such as the OpenFlow TM protocol, which structures communication between the control and data layers of supported network devices. OpenFlow is the first standard interface designed specifically for SDN, providing high-performance, granular traffic control across multiple vendors' network devices. OpenFlow-based SDN is currently being rolled out in a variety of networking devices and software, delivering substantial benefits to both enterprises Manuscript received December 10, 2013; revised February 12, 2014. This work has been supported by the Government of Spain through the project TEC2010-20527-C02-01 and through a predoctoral FPI scholarship.The authors are with Department of Telematics Engineering. Universitat Politè cnica de Catalunya. BarcelonaTech. Esteve Terradas, 7, 08860, Castelldefels, Barcelona, Spain (e-mail: aurelioj.garcia@entel.upc.edu, e-mail: cristina@entel.upc.edu, e-mail: yury.jimenez@entel.upc.edu). and carriers.NFV aims to transform the way that network operators architect networks by evolving standard Information Technology...