Centralized Control Plane Research Articles

Jointly optimized QoS-aware virtualization and routing in software defined networks

Software Defined Networks (SDNs) have been recognized as the next-generation networking paradigm that decouples the network control plane from the data forwarding plane. A logically centralized controller is responsible for all the control decisions and communication among the forwarding switches. However, current traffic engineering techniques and state-of-the-art routing algorithms do not effectively use the merits of SDNs, such as global centralized visibility, control and data plane decoupling, network management simplification and great computation capability. In this paper, a multi-tenancy management framework is proposed to enable the jointly optimized design of quality-of-services (QoSs)-aware virtualization and routing by tenant isolation and prioritization as well as flow allocation, fulfilling QoS requirements of tenants’ applications. Specifically, a fine-grained network virtualization is first proposed to isolate and prioritize tenants through the design of network and switch hypervisors. Furthermore, a QoS-aware dynamic flow allocation is proposed to enable optimal flow routes selection upon the given network slicing with QoS provisioning. Finally, an adaptive feedback management tool, called QoS-aware Virtualization-enabled Routing (QVR), is proposed to combine virtualization with flow allocation and supports reliable and efficient transmissions with regards of time-varying QoS requirements, network topologies, and traffic statistics. Simulations confirm that QVR achieves much less shared edges, congestion latency, and traffic delay for multiple tenants, thus facilitating virtualization-enabled traffic engineering for multi-tenancy SDNs.

Baseband unit cloud interconnection enabled by flexible grid optical networks with software defined elasticity

The evolution toward 5G mobile networks is characterized by supporting higher data rate, excellent end-to-end performance and ubiquitous user-coverage with lower latency, power consumption, and cost. To support this, the RANs are evolving in two important aspects. One aspect is “cloudification,” which is to pool baseband units to be centralized for statistical multiplexing gain. The other aspect is to use advanced optical technologies for digital and analog signal transmission in a cloud-based RAN. In this article, we focus on BBU cloud interconnection with optical layer technologies. Flexible grid optical networks with the enabling technologies are introduced to provide elastic, transparent, and virtualized optical paths between the BBU pools. To improve the elasticity and intelligence of C-RAN, we propose a software defined centralized control plane to coordinate heterogeneous resources from three domains: the BBU domain, radio domain, and optical domain. We report an experimental demonstration of elastic lightpath provision for cloud radio-over-flexible grid optical networks in a software-defined-networking-based testbed.

Models and algorithms for centralized control planes to optimize control traffic overhead

Carrier Ethernet has rapidly advanced to become an important technology in metro transport. However, the overlapping of control plane and data plane in contemporary Carrier Ethernet (CE) networks leads to complex and unmanageable networks. CE networks that use packet technologies ought to be more manageable, scalable and robust. Recently proposed network architecture, Omnipresent Ethernet (OE) recommends that the control and management plane in CE networks be decoupled from the forwarding and routing plane to overcome the problem mentioned earlier. In view of this, we investigate the possibility of a centralized control plane from OE perspective. A centralized control plane is a networking paradigm that abstracts and centralizes the control information of the network from the underlying distributed data-forwarding infrastructure. In this paper, we have focused on engineering and architectural issues related to the design of a centralized control plane for the networks built on OE Networking paradigm. Furthermore, the problem of control traffic overhead in managed networks is analyzed using an appropriate simulation model. A scheme to divide the network into smaller sub-networks is proposed so that the total control traffic is always below some threshold. An Integer Linear Program (ILP) for the controllers' placement in the partitioned network is presented. The ILP attempts to minimize the total control traffic, the total controllers' implementation cost and the overall response time in the network. Since the ILP solves the problem optimally for networks with a limited number of nodes, a heuristic approach is developed and presented for larger and real-time service-provider networks.

An openflow architecture for energy-aware traffic engineering in mobile networks

To cope with the growing traffic demand, future mobile networks will be denser and integrate heterogeneous technologies. However, if not properly engineered, such networks may incur huge energy waste when there is little traffic, and may suffer from an unbearable management burden caused by the variety of technologies integrated. In this article, we propose and implement a novel management architecture for mobile networks based on OpenFlow, which supports resource-ondemand provisioning in a centralized control plane, and hides the technology specifics from the controller through the use of abstractions. The feasibility of the approach is demonstrated by a real-life prototype based on commercial off-the-shelf devices.

Future RAN Architecture: SD-RAN Through a General-Purpose Processing Platform

In this article, we identify and study the potential of an integrated deployment solution for energy-efficient cellular networks combining the strengths of two very active current research themes: 1) software-defined radio access networks (SD-RANs) and 2) decoupled signaling and data transmissions, or beyond cellular green generation (BCG2) architecture, for enhanced energy efficiency. While SD-RAN envisions a decoupled centralized control plane and data-forwarding plane for flexible control, the BCG2 architecture calls for decoupling coverage from the capacity and coverage provided through an always-on low-power signaling node for a larger geographical area; the capacity is catered by various on-demand data nodes for maximum energy efficiency. In this article, we show that a combined approach that brings both specifications together can not only achieve greater benefits but also facilitate faster realization of both technologies. We propose the idea and design of a signaling controller that acts as a signaling node to provide always-on coverage, consuming low power, and at the same time host the control plane functions for the SDRAN through a general-purpose processing platform. The phantom cell concept is also a similar idea where a normal macrocell provides interference control to densely deployed small cells, although our initial results show that the integrated architecture has a much greater potential for energy savings than phantom cells.

Service‐oriented 5G network architecture: an end‐to‐end software defining approach

SummaryWith the ever‐increasing mobile demands and proliferation of mobile services, mobile Internet has penetrated into every aspect of human life. Although the 4G mobile communication system is now being deployed worldwide, simply evolving or incrementally improving the current mobile networks can no longer keep the pace with the proliferation of mobile services. Against this background, aiming to achieve service‐oriented 5G mobile networks, this article proposes an end‐to‐end software defining architecture, which introduces a logically centralized control plane and dramatically simplifies the data‐plane. The control plane decomposes the diversified mobile service requirements and, correspondingly, controls the functions and behaviors of data‐plane devices. Consequently, the network directly orients towards services, and the devices are dynamically operated according to the service requirements. Therefore, the proposed architecture efficiently guarantees the end‐to‐end QoS and quality of experience. The challenges and key technologies of our architecture are also discussed in this article. Real traces‐based simulations validate the performance advantages of proposed architecture, including energy efficiency and the whole performance. Copyright © 2015 John Wiley & Sons, Ltd.

Security in Software Defined Networks: A Survey

Software defined networking (SDN) decouples the network control and data planes. The network intelligence and state are logically centralized and the underlying network infrastructure is abstracted from applications. SDN enhances network security by means of global visibility of the network state where a conflict can be easily resolved from the logically centralized control plane. Hence, the SDN architecture empowers networks to actively monitor traffic and diagnose threats to facilitates network forensics, security policy alteration, and security service insertion. The separation of the control and data planes, however, opens security challenges, such as man-in-the middle attacks, denial of service (DoS) attacks, and saturation attacks. In this paper, we analyze security threats to application, control, and data planes of SDN. The security platforms that secure each of the planes are described followed by various security approaches for network-wide security in SDN. SDN security is analyzed according to security dimensions of the ITU-T recommendation, as well as, by the costs of security solutions. In a nutshell, this paper highlights the present and future security challenges in SDN and future directions for secure SDN.

Energy efficiency with QoS control in dynamic optical networks with SDN enabled integrated control plane

The paper presents energy efficient routing algorithms based on a novel integrated control plane platform. The centralized control plane structure enables the use of flexible heuristic algorithms for route selection in optical networks. Differentiated routing for various traffic types is used in our previous work. The work presented in this paper further optimizes the energy performance in the whole network by utilizing a multi-objective evolutionary algorithm for route selection. The trade-off between energy optimization and QoS for high priority traffic is examined and results show an overall improvement in energy performance whilst maintaining satisfactory QoS. Energy savings are obtained on the low priority traffic whilst the QoS for the high priority traffic is not degraded.

Byzantine-Resilient Secure Software-Defined Networks with Multiple Controllers in Cloud

Software-defined network (SDN) is the next generation of networking architecture that is dynamic, manageable, cost-effective, and adaptable, making it ideal for the high-bandwidth, dynamic nature of today’s applications. In SDN, network management is facilitated through software rather than low-level device configurations. However, the centralized control plane introduced by SDN imposes a great challenge for the network security. In this paper, we present a secure SDN structure, in which each device is managed by multiple controllers, not just a single as in a traditional manner, with the dynamic and isolated instance provided by the cloud. It can resist Byzantine attacks on controllers and the communication links between controllers and SDN switches. Furthermore, we study a controller minimization problem with security requirement and propose a cost-efficient controller assignment algorithm with a constant approximation ratio. From the experiment result, the secure SDN structure has little impact on the network latency, provide better security than general distributed controller, and the proposed algorithm performs higher efficiency than random assignment.

PON versus AON: Which is the best solution to offload core network by peer-to-peer traffic localization

Video streaming and video-on-demand are gaining popularity nowadays which dictates a need of bandwidth upgrade for Internet users. Many next generation optical access network architectures have been proposed to meet high capacity requirement on a per-user basis. However, the capacity upgrade in access networks, may lead to a huge traffic growth in the aggregation/core network. One way to avoid this problem is to keep the traffic locally (i.e., inside the access network area) as much as possible. It can be obtained by using locality-aware peer-to-peer (P2P) applications for content distribution and has the potential to offload the core segment. However, various optical access network architectures accommodate the P2P traffic in different ways. Thus, it is important to study these differences in order to identify the best architecture option for capacity offloading in the core network, energy efficiency and network resource utilization. By deploying a proper architecture in the access segment along with an efficient traffic locality aware strategy, the extra investment and capacity upgrade of the expensive core network resources needed to support the future traffic expansion can be minimized. However, to the best of our knowledge this kind of assessment is so far not available.Therefore, in this paper, we analyze the efficiency of supporting locality-aware P2P video distribution algorithm in three main types of optical access network architectures, i.e., active optical network (AON), wavelength division multiplexing passive optical network (WDM PON) and time/wavelength division multiplexing PON (TWDM PON). Our goal is to provide important design guidelines for the next generation broadband access architectures, while minimizing the need for the core network upgrade. We obtain this objective by utilizing the unique characteristics of each access network architecture in accommodating P2P video delivery applications. We have done an extensive literature study and for the first time we have compared performance of these architectures with respect to the amount of the traffic on the links in different aggregation levels, power consumption taking into account sleep mode functionality at the user premises, and required switching capacity in the nodes. Our results reveal that both active and passive architectures have good ability to localize P2P traffic, whereas they show distinct performance with respect to the other aforementioned aspects. This is caused by the different number of aggregation levels, link capacity, and resource allocation protocols. Considering the overall performance evaluation, it is shown that TWDM PON is the most promising option for the future broadband access, where locality-aware P2P video distribution is applied, thanks to its low energy consumption and required switching capacity of the network equipment needed to deliver this service. This conclusion is against the general intuition because of the PON׳s centralized control plane and passive infrastructure without switching capability in the field. Our unexpected conclusion can be of particular interest to operators as it is perfectly aligned with next generation optical access architecture identified by Full Service Access Networks (FSAN).

A dormant multi-controller model for software defined networking

In order to improve the scalability and reliability of Software Defined Networking (SDN), many studies use multiple controllers to constitute logically centralized control plane to provide load balancing and fail over. In this paper, we develop a fexible dormant multi-controller model based on the centralized multi-controller architecture. The dormant multi-controller model allows part of controllers to enter the dormant state under light traffc condition for saving system cost. Meanwhile, through queueing analysis, various performance measures of the system can be obtained. Moreover, we analyze the real traffic of China Education Network and use the results as the parameters of computer simulation and verify the effects of parameters on the system characteristics. Finally, a total expected cost function is established, and genetic algorithm is employed to find the optimal values of various parameters to minimize system cost for the deployment decision making.

Combining OpenFlow and sFlow for an effective and scalable anomaly detection and mitigation mechanism on SDN environments

Software Defined Networks (SDNs) based on the OpenFlow (OF) protocol export control-plane programmability of switched substrates. As a result, rich functionality in traffic management, load balancing, routing, firewall configuration, etc. that may pertain to specific flows they control, may be easily developed. In this paper we extend these functionalities with an efficient and scalable mechanism for performing anomaly detection and mitigation in SDN architectures. Flow statistics may reveal anomalies triggered by large scale malicious events (typically massive Distributed Denial of Service attacks) and subsequently assist networked resource owners/operators to raise mitigation policies against these threats. First, we demonstrate that OF statistics collection and processing overloads the centralized control plane, introducing scalability issues. Second, we propose a modular architecture for the separation of the data collection process from the SDN control plane with the employment of sFlow monitoring data. We then report experimental results that compare its performance against native OF approaches that use standard flow table statistics. Both alternatives are evaluated using an entropy-based method on high volume real network traffic data collected from a university campus network. The packet traces were fed to hardware and software OF devices in order to assess flow-based data-gathering and related anomaly detection options. We subsequently present experimental results that demonstrate the effectiveness of the proposed sFlow-based mechanism compared to the native OF approach, in terms of overhead imposed on usage of system resources. Finally, we conclude by demonstrating that once a network anomaly is detected and identified, the OF protocol can effectively mitigate it via flow table modifications.

Demonstration of a Dynamic Transparent Optical Network Employing Flexible Transmitters/Receivers Controlled by an OpenFlow–Stateless PCE Integrated Control Plane [Invited

OpenFlow, which allows operators to control the network using software running on a network operating system within an external controller, has recently been proposed and experimentally validated as a promising intelligent control plane technique. To mitigate the potential scalability issue of an OpenFlow-based centralized control plane and to leverage the mature, well-defined, and feature-complete path computation element (PCE) communication protocol, the complex path computation function required in optical networks can be formally decoupled from the OpenFlow controller so the controller can off-load the task to one or more dedicated PCEs. In addition to the control plane intelligence, future optical networks also feature data plane intelligence such as the introduction of flexible transmitters and receivers, which can dynamically change the modulation format and transmission rate of the optical signal without hardware modifications. In this paper, for the first time, we successfully demonstrate a dynamic transparent wavelength-switched optical network employing flexible transmitters and receivers controlled by an OpenFlow–stateless PCE integrated control plane. Our designed flexible transmitter is implemented by a cascade of a dual-drive Mach–Zehnder modulator (MZM) and a dual-parallel MZM. By adjusting the electrical binary drive signals, the flexible transmitter is able to flexibly switch the symbol rate and the modulation format, including binary phase shift keying, quadrature phase shift keying, 8-ary quadrature amplitude modulation (8QAM), and 16QAM. The flexible receiver is able to automatically detect different modulation formats and symbol rates and measure the bit-error rate. All the network elements, including optical switching nodes and flexible transmitters and receivers, are extended with OpenFlow interfaces, which can be intelligently controlled by the OpenFlow–stateless PCE integrated control plane with significant protocol extensions. On an actual network testbed with real hardware, we successfully validate dynamic and seamless interworking operations between the OpenFlow controller, a stateless PCE, and all the data plane hardware. The overall feasibility and efficiency of the proposed solutions are verified, and dynamic end-to-end path provisioning and lightpath restoration in such a new network scenario are quantitatively evaluated. We also tested the scalability of the proposed control plane; the experiment results indicated that dynamic path provisioning and restoration can be achieved within hundreds of milliseconds by using the proposed approach, and the overall architecture scales well with a batch of requests.

Control and Management of Flexi-grid Optical Networks With an Integrated Stateful Path Computation Element and OpenFlow Controller [Invited

A path computation element (PCE) is briefly defined as a control plane functional component (physical or logical) that is able to perform constrained path computation on a graph representing (a subset of) a network. A stateful PCE is a PCE that is able to consider the set of active connections, and its development is motivated by the fact that such knowledge enables the deployment of improved, more efficient algorithms. Additionally, a stateful PCE is said to be active if it is also able to affect (modify or suggest the modification of) the state of such connections. A stateful active PCE is thus able not only to use the knowledge of the active connections as available information during the computation, but also to reroute existing ones, resulting in a more efficient use of resources and the ability to dynamically arrange and reoptimize the network. An OpenFlow controller is a logically centralized entity that implements a control plane and configures the forwarding plane of the underlying network devices using the OpenFlow protocol. From a control plane perspective, an OpenFlow controller and the aforementioned stateful PCE have several functions in common, for example, in what concerns network topology or connection management. That said, both entities also complement each other, since a PCE is responsible mainly for path computation accessible via an open, standard, and flexible protocol, and the OpenFlow controller assumes the task of the actual data plane forwarding provisioning. In other words, the stateful PCE becomes active by virtue of relying on an OpenFlow controller for the establishment of connections. In this framework, the integration of both entities presents an opportunity allowing a return on investment, reduction of operational expenses, and reduction of time to market, resulting in an efficient approach to operate transport networks. In this paper, we detail the design, implementation, and experimental evaluation of a centralized control plane based on a stateful PCE, acting as an OpenFlow controller, targeting the control and management of optical networks. We detail the extensions to both the OpenFlow and the PCE communication protocol (PCEP), addressing the requirements of elastic optical networks as well as the system performance, obtained when deployed in a laboratory trial.

PCE-based centralized control plane for filterless networks

In this article, we present the concept of the filterless optical network from a control plane perspective, defining the network and showing to what extent a control plane could be adapted to it. Then we propose a control plane architecture and study its performance. Filterless networks act as passive broadcast and select networks, in which active switching elements, such as the wavelength selection switch, are replaced by passive optical splitters and combiners, and network agility is provided by tunable transceivers. We introduce the concept of unfiltered channels and that of backward and forward ports, and show how they impact topological performance, and, in particular, the fact that wavelength usage increases greatly as the protection ratio shrinks.

Extending software defined network principles to include optical transport

Software defined networks are based on the principle of a centralized control plane separated from the network forwarding or switching plane that it controls. The switching plane can be heterogeneous, composed of network elements from multiple vendors, and it can provide distinct services with different characteristics, configurations, and control at the packet and/or optical layers. Abstracting the control plane from the network elements allows network-platform- specific characteristics and differences that do not affect services to be hidden. In addition, software defined networking (SDN) is based on the principle that applications can request needed resources from the network via interfaces to the control plane. Through these interfaces, applications can dynamically request network resources or network information that may span disparate technologies. For instance, the application layer can dynamically request and obtain network resources at the packet flow, circuit, or even optical level based on application layer requirements. Current SDN implementations focus on Ethernet switching primarily for data center resource optimization. This article reviews the benefits and challenges of extending SDN concepts to various transport network architectures that include optical wavelength and fiber switches, circuit switches, and sub-wavelength optical burst switches. Control plane implementations for optical networks are more complex since they must account for physical constraints including optical signal reachability, bandwidth availability and granularity, light path routing, and light path reconfiguration speed. The longterm goal is to apply SDN concepts across multi-layer multivendor networks in order to support a unified control structure.

Design and Experimental Validation of a GMPLS/PCE Control Plane for Elastic CO-OFDM Optical Networks

ITU-T Recommendation G.694.1 defines normative DWDM frequency grids, each being a reference set of values that correspond to allowed nominal central frequencies, obtained using a fixed channel spacing (e.g., 12.5 GHz, 25 GHz, 50 GHz or 100 GHz). This rigid, grid-based approach does not seem well adapted for data rates beyond 100 Gb/s, is particularly inefficient when a whole wavelength is assigned to a lower rate optical path, and is not flexible enough for multi-rate systems . Consequently, the next generation of optical networks will require a flexible, highly efficient and adaptive management of the optical spectrum, along with advanced optical modulation schemes that efficiently use allocated spectrum slots, and recent progress on optical network technology justifies research on both new optical transmission systems as well as the applicability of control and management frameworks to such networks. We design and deploy a GMPLS control plane for flexible optical networks with coherent optical orthogonal frequency-division multiplexing (O-OFDM) transmission; we detail its functional architecture, which combines a centralized entity that performs path routing and modulation assignment, with a distributed spectrum allocation. The centralized entity (i.e., a path computation element) uses pre-configured static path characterizations, based on exhaustive OFDM transmission simulations, when performing dynamic path computation in line with GMPLS constrained shortest path mechanisms. The distributed spectrum allocation assigns frequency ranges (slots) to connection requests, by using dynamic signaling procedures and applying slot assignment policies. We summarize the control plane protocol extensions involved in the main functional aspects: routing and topology dissemination, path computation, signaling and resource reservation. We experimentally validate and evaluate the integrated centralized PCE and GMPLS control plane in a control plane testbed, obtaining key performance indicators such as path setup latency and blocking probability for different frequency slot assignment policies.

On the placement of controllers in software-defined networks

The software-defined network (SDN) approach assumes a logically centralized control plane, which is physically decoupled from the data plane. Since this approach simplifies network management and speeds up network innovations, these benefits have led not just to production-intent prototypes, but real SDN deployments. For wide-area SDN deployments, multiple controllers are often required, and the placement of these controllers influences every aspect of an SDN. Since optimizing every metric is generally an non-deterministic polynomial (NP)-hard problem, it is of great importance to find an efficient controller placement algorithm. In this paper, we develop several placement algorithms to make informed placement decisions, which can be used to maximize the reliability of SDN, since network failures could easily cause disconnections between the control and forwarding planes. These algorithms are evaluated using real topologies. Simulation results show that the controller placement strategies are crucial to SDN performance, and a greedy algorithm provides placement solutions that are close to optimal.

Performance Evaluation of Service-Aware Optical Transport System

We propose and experimentally demonstrate a service-aware optical transport system. The proposed service-aware optical transport system makes a flow based on service type and priority of traffic. The generated flow is mapped to a corresponding sub-λ for transport over an optical network. Using sub-λ provided by the centralized control plane, we could effectively provide quality-of-service guaranteed Ethernet service and best-effort service simultaneously in a single link. The committed information rate (CIR) traffic and best-effort traffic are assigned to different sub-λs. The bandwidth of the CIR traffic is guaranteed without being affected by violation traffic because the bandwidth is managed per each sub-λ. The failure detection time and restoration time from a link failure is measured to be about 60 µs and 22 ms, respectively, in the ring network. The measured restoration time is much smaller than the 50 ms industry requirement for real-time services. The fast restoration time allows the proposed service-aware optical transport system to offer high availability and reliability which is a requirement for transport networks.

Layer 1 Virtual Private Networks in Multidomain Next-Generation Networks

High-capacity e-science and consumer applications require transport networks that are dynamically provisioned. Rapid advances in next-generation SONET/SPH and optical switching along with GMPLS control have enabled many new services' provisioning capabilities. In particular, a key paradigm is the new layer 1 virtual private network framework, which allows clients to directly provision their own services without deploying expensive infrastructures. This article presents an overview of L1-VPN and describes a resource management scheme that will enable transport network virtualization across a multidomain network infrastructure. The scheme is implemented in both centralized and distributed control frameworks, and allows for dynamic sharing of transport resources. A case study of performance analysis results is presented showing a distributed control plane in a multidomain network architecture achieving higher VPN carrying capacity than a centralized control plane