ConsciousCache: Improving SDN controller scalability using flows similarities

Considering the problem that the SDN controller can not fully resolve the virtual network function during the Software Defined Network (SDN) and Network Function Virtualization (NVF) collaborative deployment in the satellite network, a SDN/NFV coordination based on the three-layer satellite network is designed. The framework extends the SDN controller function and proposes the SDN controller function extension method (SDN-FE) based on LightGBM algorithm. SDN-FE has extended the functions of the original SDN controller in three aspects. One is based on LightGBM to establish a classification model for intelligent forwarding of SDN controller data packets. The second is to learn and predict VNF modification of data packets through logistic regression. The third is to design an idle state monitoring mechanism. The above method effectively enhances the control capability of the SDN controller to the network. In terms of SDN packet forwarding classification, SDN-FE is compared with existing well-known machine learning classification algorithms such as SVM, random forest, LSTM. Simulation experiments show that the SDN packet forwarding classification problem is predictable. The average accuracy of the LightGBM model in multi-classification problems is over 69%, which is better than machine learning models such as SVM and random forest. The next step is to conduct an in-depth study of the packet rule learning algorithm in VNF.

Software-Defined Networking (SDN) is now a well-established approach in 5G, Internet of Things (IoT) and Cloud Computing. The primary idea behind its immense popularity is the separation of its underlying intelligence from the data-carrying components like routers and switches. The intelligence of the SDN-based networks lies in the central point, popularly known as the SDN controller. It is the central control hub of the SDN-based network, which has full privileges and a global view over the entire network. Providing security to SDN controllers is one such important task. Whenever one wishes to implement SDN into their data center or network, they are required to provide the website to SDN controllers. Several attacks are becoming a hurdle in the exponential growth of SDN, and among all one such attack is a Distributed Denial of Service (DDoS) attack. In a couple of years, several new SDN controllers will be available. Among many, Open Networking Operating System (ONOS) and OpenDayLight (ODL) are two popular SDN controllers laying the foundation for many other controllers. These SDN controllers are now being used by numerous businesses, including Cisco, Juniper, IBM, Google, etc. In this paper, vulnerability analysis is carried out against DDoS attacks on the latest released versions of both ODL and ONOS SDN controllers in real-time cloud data centers. For this, we have considered distributed SDN controllers (located at different locations) on two different clouds (AWS and Azure). These controllers are connected through the Internet and work on different networks. DDoS attacks are bombarded on the distributed SDN controllers, and vulnerability is analyzed. It was observed with experimentation that, under five different scenarios (malicious traffic generated), ODL-3 node cluster controller had performed better than ONOS. In these five different scenarios, the amount of malicious traffic was incregradually increased. It also observed that, in terms of disk utilization, memory utilization, and CPU utilization, the ODL 3-node cluster was way ahead of the SDN controller.

Software Defined Networking (SDN) is a rapidly emerging technology in the commercial market, and is a technology that should be embraced by the defence and emergency services to enable more flexible and adaptable communication networks in the future. Common deployment of SDN, especially in data centres, involves building a separate network for the control channel between forwarding devices and the SDN controllers. For geographically distributed networks typically found in defence and emergency services, it is often not feasible to establish a separate dedicated network for the control channel. In such networks, traffic between SDN controllers and forwarding devices therefore needs to be in-band in the same network that the SDN controllers should control. This paper describes the steps necessary for forwarding devices to both discover and set up the connections with SDN controllers when in-band control channels must be used. The work expands on earlier published work on setting up in-band control channels between SDN controllers and the forwarding devices. Especially, our solution that uses the Intent frameworks of major SDN controllers improves on network resilience in disconnected, intermittent and limited networks and the solution will allow for implementation of enhanced security mechanisms compared to previous work.

Software Defined Network (SDN) Technology made tremendous changes in networking field by controlling the hardware with remote software by decoupling data plane and control plane. The focus here is on the controller which is part of control plane. A controller is a central part of SDN; it is an intermediary between Data plane and Forwarding plane. It takes prime place in SDN architecture like Operating System in Computers. SDN Controller is a software which directs the packets using flow rules and also contains statistical information of network. The controller orchestrates the networking devices such as switches and routers by channelizing the packets. SDN controller has different flavor based on the commercial requirement achieved by including new pluggable module. This paper discusses some of the SDN Controller with their feature; the controllers are POX, NOX, Floodlight, ODL, Open Contrail. This paper also compares the characteristic of centralized and distributed controller by focusing on performance, bandwidth and round-trip time of Floodlight and Open Daylight (ODL) controller. The simulation is conducted using Mininet.

Software Defined Networking (SDN) and Network Function Virtualization (NFV) are transforming Data Center (DC), Telecom, and enterprise networking. The programmability offered by P4 enables SDN to be more protocol-independent and flexible. Data Centers are increasingly adopting SmartNICs (sNICs) to accelerate packet processing that can be leveraged to support packet processing pipelines and custom Network Functions (NFs). However, there are several challenges in integrating and deploying P4 based SDN control as well as host and sNIC-based programmable NFs. These include configuration and management of the data plane components (Host and sNIC P4 switches) for the SDN control plane and effective utilization of data plane resources. P4NFV addresses these concerns and provides a unified P4 switch abstraction framework to simplify the SDN control plane, reducing management complexities, and leveraging a host-local SDN Agent to improve the overall resource utilization. The SDN agent considers the network-wide, host, and sNIC specific capabilities and constraints. Based on workload and traffic characteristics, P4NFV determines the partitioning of the P4 tables and optimal placement of NFs (P4 actions) to minimize the overall delay and maximize resource utilization. P4NFV uses Mixed Integer Linear Programming (MILP) based optimization formulation and achieves up to 2. 5X increase in system capacity while minimizing the delay experienced by flows. P4NFV considers the number of packet exchanges, flow size, and state dependency to minimize the delay imposed by data transmission over PCI Express interface.

New and emerging use cases, such as the interconnection of geographically distributed data centers (DCs), are drawing attention to the requirement for dynamic end-to-end service provisioning, spanning multiple and heterogeneous optical network domains. This heterogeneity is, not only due to the diverse data transmission and switching technologies, but also due to the different options of control plane techniques. In light of this, the problem of heterogeneous control plane interworking needs to be solved, and in particular, the solution must address the specific issues of multi-domain networks, such as limited domain topology visibility, given the scalability and confidentiality constraints. In this article, some of the recent activities regarding the Software-Defined Networking (SDN) orchestration are reviewed to address such a multi-domain control plane interworking problem. Specifically, three different models, including the single SDN controller model, multiple SDN controllers in mesh, and multiple SDN controllers in a hierarchical setting, are presented for the DC interconnection network with multiple SDN/ OpenFlow domains or multiple OpenFlow/ Generalized Multi-Protocol Label Switching (GMPLS) heterogeneous domains. In addition, two concrete implementations of the orchestration architectures are detailed, showing the overall feasibility and procedures of SDN orchestration for the end-to-end service provisioning in multi-domain data center optical networks.

Due to the need to agilely cope with more complex and virtualized applications, integrating Software-Defined Networking (SDN) switches and OpenStack software platform into the data center has becomes an important enabling technology. However, the SDN controller and OpenStack have their own management interface, such as Horizon, which may not be able to meet the user's requirement on managing both platforms in one data center. For example, although the deployment of switches of a data center usually follows a layering hierarchy, the Graphical User Interface (GUI) of management system of the SDN controller, such as Ryu, shows the topology of switches in a random and messy manner. Furthermore, tightly integration of SDN and OpenStack is required for agile and real-time management. Thus, in this paper, we design and implement a management system which tightly integrates SDN and OpenStack. In the proposed management system, the topology displays SDN switches, Open vSwitches (OVS), as well as virtual machines in hierarchy of multi-rooted trees. In order to display millions of virtual machines and thousands of switches, the GUI of the proposed management system provides a scalable view for users to zoom in or zoom out the topology. The proposed management system adopts modular design and provides web-based user interface and RESTful API for interoperability. The prototype of the proposed management system is shown to demonstrate the usability and scalability of the proposed system.

Interest in OpenFlow and software-defined network (SDNs) has resulted in a boom in SDN hardware and controller offerings, with varying degrees of maturity, popularity and support. However, few studies have been conducted to investigate the interaction between SDN hardware and software, as well as its impact on controller design and implementation. In this paper, we chronicle our experience with deploying two commodity SDN controllers and a new system, IronStack, of our own design in a production enterprise network at Cornell University, and describe the lessons learnt. We also report on several practical limitations of SDN and controller technology, and detail important future challenges for SDN adopters and developers.

The Internet has reached a breaking point, enabling ubiquitous accessibility to all digital demands. Though people have accelerated their use of overall digital demands due to the convenience of the Internet, where nearly everything is connected and accessible, it is becoming increasingly difficult to manage and modify networks. It is always a challenge to meet the emerging needs of network industry. SDN framework was proposed to overcome traditional network architecture (TNA) restrictions in order to cope with contemporary network management in terms of flexibility, agility, responding to faults, and dynamic changes. SDN assists network operators by uncoupling the control plane from its underlying ternary content-addressable memory (TCAM) hardware often called as forwarding plane by allowing the networks to programme as per the anticipated network functionality. Now, switching from traditional to software-defined networking styles, it is worth to check the robustness of the open flow (south bound interface)-based controller platform and the capability of its fault tolerance attribute to approach the successful SDN implementation. The performance of SDN open flow controllers plays a vital role in SDN implementation, which is meant for the separation between control and data plane to leverage the life of network industry. This paper comprehensively surveys the contrast between traditional network architecture (TNA) and software-defined network architecture. This research work discusses about the best feasible SDN implementation, as well as the selection of a useful controller platform for overcoming traditional network constraints and the performance assessment criteria for SDN open flow controllers. Further, the survey is continued with possible literature or work done by various researchers on performance analysis and evaluation of open flow-based SDN controllers as well as in a group of two or more controllers with respect to multifarious network condition, network criteria/parameter/metric, scaling of network load resources with most important varying network topology design. Ultimately, it strives to point ongoing research omission with respect to controller’s performance bottleneck benchmark suggesting optimized solution when it is well experimented.KeywordsTraditional network architecture (TNA)Software-defined networking (SDN)SDN open flow controllersOpen flow protocolPerformance evaluation/analysis

Software‐defined networking (SDN) is a relatively new approach in network management that proposes to separate the network control (Control plane) and the forwarding process (Data plane) to optimize the network infrastructure and improve network performance, controllability, manageability and flexibility. However, like every technology, SDN has brought its own new challenges in terms of security and scalability which are very important aspects that should be considered to design and build a resilient architecture in order to meet carrier grade network requirements. In this paper, we propose a secure SDN architecture with IEEE 802.1X port‐based authentication where we also consider the controller's scalability issue by devolving the access control capability to the data plane. In this way, we reduce the high demand and the workload on the SDN controller. Our proposed model presents a novel SDN network architecture and logical network segmentation which provides an optimal and secure network access with low latency. We have implemented and tested our architecture to show its performance (authentication delays). Copyright © 2016 John Wiley & Sons, Ltd.

Wireless Sensor Networks (WSNs) have a wide range of applications scenarios in computer vision, from pedestrian detection to robotic visual navigation. In response to the growing visual data services in WSNs, we propose a proactive caching strategy based on Stacked Sparse Autoencoder (SSAE) to predict content popularity (PCDS2AW). Firstly, based on Software Defined Network (SDN) and Network Function Virtualization (NFV) technologies, a distributed deep learning network SSAE is constructed in the sink nodes and control nodes of the WSN network. Then, the SSAE network structure parameters and network model parameters are optimized through training. The proactive cache strategy implementation procedure is divided into four steps. (1) The SDN controller is responsible for dynamically collecting user request data package information in the WSNs network. (2) The SSAEs predicts the packet popularity based on the SDN controller obtaining user request data. (3) The SDN controller generates a corresponding proactive cache strategy according to the popularity prediction result. (4) Implement the proactive caching strategy at the WSNs cache node. In the simulation, we compare the influence of spatiotemporal data on the SSAE network structure. Compared with the classic caching strategy Hash + LRU, Betw + LRU, and classic prediction algorithms SVM and BPNN, the proposed PCDS2AW proactive caching strategy can significantly improve WSN performance.

Software-defined networking (SDN) has become a popular paradigm for managing large-scale networks including cloud servers and data centers because of its advantages of centralized management and programmability. The issues of scalability and reliability that a single centralized controller suffers makes distributed controller architectures emerge. One key limitation of distributed controllers is the statically configured switch-controller mapping, easily causing uneven load distribution among controllers. Previous works have proposed load-balancing methods with switch migration to address this issue. However, the higher-load controller is always directly considered as the overloaded controller that need to shift its load to other controllers, even if it has no response time delay. The pursuit of absolute load-balancing effect can also result in frequent network delays and service interruptions. Additionally, if there are several overloaded controllers, just one controller with the maximum load can be addressed within a single load-balancing operation, reducing load-balancing efficiency. To address these problems, we propose SMCLBRT, a load-balancing strategy of multiple SDN controllers based on response time, considering the changing features of real-time response times versus controller loads. By selecting the appropriate response time threshold and dealing with multiple overloading controllers simultaneously, it can well solve load-balancing problem in SDN control plane with multiple overloaded controllers. Simulation experiments exhibit the effectiveness of our scheme.

Software defined networks (SDN) has gained a lot of attention from both academia and industry recently. SDN is a new networking paradigm that decouples the control plane from the data plane. In this paradigm, control functions are centralized in an SDN controller. The migration from traditional networks to SDN is highly challenging for network operators. Sudden migration may not be possible due to the cost and operational issues. Therefore, incremental deployment of SDN is considered as a viable solution, resulting in the formation of hybrid SDN networks. An important task in hybrid SDN networks is bandwidth allocation and routing considering the coexistence of both conventional and SDN-enabled routers. In this paper, we first formulate this problem as a multi-objective integer program with delay and capacity constraints. Then, a heuristic approach is proposed to combat the computational complexity of the model. Our proposed method is extensively analyzed through simulations, and the evaluation results confirm the effectiveness of our approach.

In this paper optical transport networks (OTN) and different control planes approaches are presented. For the Software Defined Networks (SDN) configurations and mode of operations such as implicit and explicit are discussed. The path latencies of two control plane approaches are examined and numerical results are presented on the path metric. The results are derived from a simple network simulation interconnected two data centers into different locations using SDN controller. In addition, a comparison between SDN and MPLS/PCE are presented.

This article describes how novel functionalities will take advantage of the cloud networking and will gradually replace the existing infrastructure of mobile networks with a virtualized one. Two technologies, namely software defined networking (SDN) and network function virtualization (NFV), offer their important benefits and a combination of them is an answer to the demands raised, such as central office re-architected as a data center (CORD). Open network operating system (ONOS) and POX are SDN controllers and offer an option to combine SDN and NFV addressing many ongoing problems in the field of mobile networks. In this paper, technologies and both controllers are compared and contrasted. Indicative cases of topologies are simulated and help evaluating both controllers. According to the experimental findings, ONOS is one of the most important controllers for practical, theoretical, research and educational purposes, while POX is a useful and simpler controller for other educative applications.