Network Hypervisor Research Articles

Self-adjusting resilient control plane for virtual software-defined optical networks

Optical networks must promptly respond to failures and efficiently handle dynamic traffic in order to fulfill their role as a critical infrastructure. Leveraging network softwarization and virtualization, virtual software-defined networks offer sufficient flexibility towards this goal by sharing the physical infrastructure among multiple tenants whose traffic must traverse the network hypervisor. In a resilient optical control plane each switch must be assigned to a primary and backup hypervisor instance through short control paths, which challenge will be addressed in this paper. First, we propose an intelligent greedy hypervisor placement heuristic which maximizes acceptance ratio for current, and preparedness for future requests. Secondly, we introduce a graph neural network model that can be seamlessly integrated with either our integer linear program or heuristic method to yield high-quality placements in significantly less time compared to our prior solutions. This enhancement renders our approach applicable to larger networks, significantly expanding its practical utility. Finally, we propose a self-adjusting hypervisor migration strategy, which continuously adapts the placement to the dynamically changing virtual network requests, thus, ensuring service continuity by avoiding frequent control plane reconfigurations. Through simulations we show that our hypervisor placement and migration strategies provide a balanced control load while they can handle a wide variety of changes.

AgentVisor: Agent-Based Network Hypervisor for Autonomic Network Virtualization

A network hypervisor is a novel solution for network virtualization that allows the dynamic creation of virtual networks to facilitate independent business solutions to numerous tenants over a shared network infrastructure. This article proposes an agent-based network hypervisor (AgentVisor) architecture that virtualizes controller agents, separated from tenants' controllers to support high levels of adaptation, automation and autonomy of virtual software-defined networking networks (vSDNs). The controller agents enable closed control loop automation of systematic data collection, data-driven decision-making and autonomic control actions for monitoring, configuration and fault handling of vSDNs without intervention of tenants' controllers in a cloud/edge environment. The controller agent also offers an extended northbound interface with tenants' controllers to support policy and intelligence management based on artificial intelligence and machine learning algorithms. Through controller agent virtualization, the proposed AgentVisor architecture enables a multiagent system of autonomic vSDNs that facilitate independent business solutions for different markets over a shared infrastructure for fifth generation (5G) vision and beyond.

Resilient Control Plane Design for Virtualized 6G Core Networks

With the advent of 6G and its mission-critical and tactile Internet applications running in a virtualized environment on the same physical infrastructure, even the shortest service disruptions have severe consequences for thousands of users. Therefore, the network hypervisors, which enable such virtualization, should tolerate failures or be able to adapt to sudden traffic fluctuations instantaneously, i.e., should be well-prepared for such unpredictable environmental changes. In this paper, we propose a latency-aware dual hypervisor placement and control path design method, which protects against single-link and hypervisor failures and is ready for unknown future changes. We prove that finding the minimum number of hypervisors is not only NP-hard, but also hard to approximate. We propose optimal and heuristic algorithms to solve the problem. We conduct thorough simulations to demonstrate the efficiency of our method on real-world optical topologies, and show that with an appropriately selected representative set of possible future requests, we are not only able to approach the maximum possible acceptance ratio but also able to mitigate the need of frequent hypervisor migrations for most realistic latency constraints.

Resilient Control Plane Design for Virtual Software Defined Networks

Control plane survivability in virtual software-defined networks (vSDN) – where multiple tenants share the same physical infrastructure – is even more critical than in normal SDN networks. A reliable communication channel from the switches through the network hypervisor to the virtual controller is inevitable in order to avoid state inconsistencies, tenant isolation and security issues on the virtual switches. Although reliable controller placement and control plane design was thoroughly investigated in SDNs, there was a lack of attention for resilient hypervisor placement and control path design for vSDNs. Therefore, in this paper we make a two-fold contribution towards a survivable vSDN control plane. First, we propose an approximation algorithm for (hypervisor) placement which – in contrast with traditional approaches which minimize the average latency to the hypervisors as an objective function – focuses on finding the appropriate number of hypervisor instances to satisfy the control path length constraints declared in the service level agreements, leaving enough options open for self-driving network designs and intelligent algorithms. Second, we propose a general dynamic program that calculates minimum length paths traversing specific type of nodes in a given order, and apply it to find control paths from the virtual controller of the slice to the virtual switches traversing the corresponding hypervisor location. We conduct thorough simulations on real-world topologies to demonstrate the effectiveness of our approaches in no failure and single link failure scenarios.

Enabling SDN Hypervisor Provisioning Through Accurate CPU Utilization Prediction

Providing predictable performance to tenants is mission critical for network hypervisors. As a hypervisor acts as an intermediary between tenants controllers and the physical infrastructure, its resources (e.g., CPU, RAM) should be provisioned and allocated carefully. Initially, we demonstrate that state-of-the-art CPU prediction approaches are not suitable for provisioning network hypervisor CPU resources, since they predict only the mean CPU utilization. However, provisioning the resources with a mean value can significantly degrade the forwarding performance of a network hypervisor. In this article, we present a novel approach which provisions network hypervisor CPU resources efficiently, while avoiding performance degradation. We take three steps to achieve our goal: (i) conducting a profound measurement campaign to determine what is the minimum amount of CPU resources that needs to be allocated to a network hypervisor in order to have no performance degradation; (ii) revealing the key properties of virtual networks that affect the CPU utilization; (iii) designing a precise CPU prediction model. Using randomly generated virtual networks and arbitrary physical topologies, we show that our prediction model exhibits an average relative error of around 4%. Further, our evaluations indicate that provisioning the CPU resources of a network hypervisor based on the proposed prediction model does not degrade the hypervisor forwarding performance. Utilizing our approach, network operators can minimize their resources consumption while still providing predictable and undegraded forwarding performance to tenants.

Libera for Programmable Network Virtualization

Current network virtualization allows tenants to have their own virtual networks. However, new demands to virtual networks at a finer granularity have arisen as tenants want the ability to provision and control switches and links in their virtual networks. This study proposes a new concept called the p-NIaaS model. The p-NIaaS model enables tenants to program their own packet processing logic and monitor network status from any virtual network infrastructure, which is not possible with current network virtualization. This article presents the Libera network hypervisor that implements the p-NIaaS model. Libera overcomes the shortcomings of existing network hypervisors such as scalability, VM migration support, and flexibility. The evaluation shows that the Libera hypervisor is highly scalable and effectively supports VM migration. We also present the overheads of Libera. Libera incurs up to 11 percent overhead in comparison with a non-virtualized network, which we believe is promising in the first prototype of the p-NIaaS model.

5G network slicing using SDN and NFV: A survey of taxonomy, architectures and future challenges

The increasing consumption of multimedia services and the demand of high-quality services from customers has triggered a fundamental change in how we administer networks in terms of abstraction, separation, and mapping of forwarding, control and management aspects of services. The industry and the academia are embracing 5G as the future network capable to support next generation vertical applications with different service requirements. To realize this vision in 5G network, the physical network has to be sliced into multiple isolated logical networks of varying sizes and structures which are dedicated to different types of services based on their requirements with different characteristics and requirements (e.g., a slice for massive IoT devices, smartphones or autonomous cars, etc.). Softwarization using Software-Defined Networking (SDN) and Network Function Virtualization (NFV)in 5G networks are expected to fill the void of programmable control and management of network resources.In this paper, we provide a comprehensive review and updated solutions related to 5G network slicing using SDN and NFV. Firstly, we present 5G service quality and business requirements followed by a description of 5G network softwarization and slicing paradigms including essential concepts, history and different use cases. Secondly, we provide a tutorial of 5G network slicing technology enablers including SDN, NFV, MEC, cloud/Fog computing, network hypervisors, virtual machines & containers. Thidly, we comprehensively survey different industrial initiatives and projects that are pushing forward the adoption of SDN and NFV in accelerating 5G network slicing. A comparison of various 5G architectural approaches in terms of practical implementations, technology adoptions and deployment strategies is presented. Moreover, we provide a discussion on various open source orchestrators and proof of concepts representing industrial contribution. The work also investigates the standardization efforts in 5G networks regarding network slicing and softwarization. Additionally, the article presents the management and orchestration of network slices in a single domain followed by a comprehensive survey of management and orchestration approaches in 5G network slicing across multiple domains while supporting multiple tenants. Furthermore, we highlight the future challenges and research directions regarding network softwarization and slicing using SDN and NFV in 5G networks.

On Table Resource Virtualization and Network Slicing in Programmable Data Plane

Recently, the advances on programmable data plane (PDP) promote the studies on the network virtualization in a PDP-based substrate network (SNT). In this paper, we address the table resource virtualization and network slicing in PDP-based SNTs. We first leverage the idea of “Big-Switch” to design an effective table resource virtualization scheme in which the flow tables of a virtual switch (V-SW) can be installed in multiple adjacent substrate switches (S-SWs) according to their table sizes, while the feasible table size(s) on each S-SW are determined based on the global information of the SNT. By doing so, we can regulate the flow tables in the S-SWs in a more organized way to minimize memory fragmentation. Next, we address the network slicing based on the virtualization scheme, and come up with a three-layer VNE problem. To the best of our knowledge, such a VNE problem has not been studied before and the existing algorithms designed for traditional two-layer VNE problems can hardly solve it. We formulate an integer linear programming (ILP) model to solve the VNE problem exactly, and also design a time-efficient heuristic that can provide near-optimal solutions. Finally, we implement the heuristic in TPVX, which is a network hypervisor based on protocol-oblivious forwarding (POF), and also improve its performance by introducing source routing. The new TPVX is experimentally demonstrated in a real network testbed, and the results verify that our proposal maintains the additional latency caused by the three-layer VNE well and would not degrade the network services in virtual networks (VNTs).

Optimising virtual networks over time by using Windows Multiplicative DEA model

Recently, the prediction of the most efficient configuration of a vast set of devices used for mounting an optimised cloud computing services and virtual networks environments have attracted growing attention. This paper proposes a paradigm shift in modelling transmission control protocol (TCP) behaviour over time in virtual networks by using data envelopment analysis (DEA) models. Firstly, it proves that self-similarity with long-range dependency is presented differently in every network device. This study implements a novel fractal dimension concept on virtual networks for prediction, where this key index informs if the transport layer forwards services with smooth or jagged behaviour over time. Another substantial contribution is proving that virtual network devices have a distinct fractal memory, TCP bandwidth performance, and fractal dimension over time, presenting themselves as important factor for forecasting of spatiotemporal data. Thus, a continuous stepwise fractal performance evaluation framework methodology is developed as an expert system for virtual network assessment and performs a fractal analysis as a knowledge representation. In addition, due to the limitations of classical DEA models, the windows multiplicative data envelopment analysis (WMDEA) model is used to dynamically assess the fractal time series from virtual network hypervisors. For knowledge acquisition, 50 different virtual network hypervisors were appraised as decision-making units (DMU). Finally, this expert system also acts as a math hypervisor capable of determining the correct fractal pattern to follow when delivering TCP services in an optimised virtual network.

The (In)Security of Virtualization in Software Defined Networks

Software Defined Networking (SDN) is a new networking paradigm with the promise to increase simplicity and efficiency in network management through the separation of control functions from the forwarding functions. In SDN, the control functions are softwarized and logically placed in a centralized entity, i.e. the SDN controller. Network virtualization is one of the key features enabled and facilitated by the SDN, and it allows multiple virtual networks and the SDN controllers to share the same physical network infrastructure. This paper discusses the security of virtualization in the SDN, and it highlights and demonstrates critical vulnerabilities of key network hypervisors used in the SDN. In particular, the paper demonstrates how the isolation of different virtual networks can be broken, and enabling different types of attacks. Finally, the paper discusses the potential impact of these vulnerabilities and points to mitigation approaches.

LiteVisor: A Network Hypervisor to Support Flow Aggregation and Seamless Network Reconfiguration for VM Migration in Virtualized Software-Defined Networks

Network virtualization based on software-defined networking (SDN) has become a necessary technology to provide various services in cloud datacenters. Although many network hypervisors have been proposed to support SDN-based network virtualization, their forwarding techniques excessively consume the limited ternary contents addressable memory of OpenFlow-enabled switches. Moreover, they do not consider network reconfiguration after virtual machine migration. In this paper, we propose LiteVisor, that resolves the two problems mentioned above. It develops the Locator, Identifier, and Tenant sEparating (LITE) scheme. The LITE-based forwarding enables flow aggregation that reduces switch memory consumption. In addition, LiteVisor provides seamless network reconfiguration that does not require tenant controllers to be aware of virtual machine migration based on LITE. We evaluate LiteVisor in terms of the number of flow table entries and network reconfiguration time and compare it with OpenVirteX that is an open-source network hypervisor. The results show that the number of flow rules decreases by up to eight times compared with OpenVirteX in a fat-tree topology. We also demonstrate the seamless network reconfiguration of LiteVisor in the experiments and present the results of the reconfiguration time by the topology size and packet sending interval of hosts.

Network Hypervisors pros and cons A survey

Network Hypervisors pros and cons A survey

ONVisor: Towards a scalable and flexible SDN‐based network virtualization platform on ONOS

SummaryNetwork virtualization (NV) technologies have attracted a lot of attention as an essential solution for future networking infrastructure. The NV enables multiple tenants to share the same physical infrastructure and to create independent virtual networks (VNs) by decoupling the physical network in terms of topology, address, and control functions. One feasible way to realize full NV involves considering solutions based on the software‐defined networking (SDN) paradigm using its programmability. The SDN contributes many benefits to both network operations and management including programmability, agility, elasticity, and flexibility. There are several SDN‐based NV solutions; however, they suffered from a lack of scalability, high availability. Also, they have high latency between control and data plane because of proxy‐based architecture. In this thesis, we introduce a new NV platform, named Open Network Hypervisor (ONVisor). The design objectives include, among the features, (1) multitenancy, (2) scalability, (3) flexibility, (4) isolated VNs, and (5) VN federation. ONVisor was designed and implemented by extending Open Network Operating System, an open‐source SDN controller. The main features of ONVisor are (1) isolated control and data plane per VN, (2) support of distributed operations, (3) extensible translators, (4) on‐platform VN application development and execution, and (5) support of heterogenous SDN data‐plane implementations. Several experiments are conducted on various test scenarios in different test environments in terms of control and data plane performance compared to nonvirtualized SDN network. The results show that ONVisor can provide VNs a little bit lower control plane performance and similar data plane performance.

Experimental demonstration of building and operating QoS-aware survivable vSD-EONs with transparent resiliency.

Software-defined elastic optical networks (SD-EONs) provide operators more flexibility to customize their optical infrastructures dynamically. By leveraging infrastructure-as-a-service (IaaS), virtual SD-EONs (vSD-EONs) can be realized to further enhance the adaptivity of SD-EONs and shorten the time-to-market of new services. In this paper, we design and demonstrate the building and operating of quality-of-service (QoS) aware survivable vSD-EONs that are equipped with transparent data plane (DP) resiliency. Specifically, when slicing a vSD-EON, our network hypervisor (NHV) chooses to use "1:1" virtual link (VL) protection or on-demand VL remapping as the DP restoration scheme, according to the service-level agreement (SLA) between the vSD-EON's operator and the infrastructure provider (InP). Then, during an actual substrate link (SL) failure, the NHV realizes automatic DP restoration that is transparent to the controllers of vSD-EONs. We build a network testbed to demonstrate the creation of QoS-aware survivable vSD-EONs, the activation of lightpaths in the vSD-EONs to support upper-layer applications, and the automatic and simultaneous QoS-aware DP restorations during an SL failure. The experimental results indicate that our vSD-EON slicing system can build QoS-aware survivable vSD-EONs on-demand, operate them to set up lightpaths for carrying real application traffic, and facilitate differentiated DP restorations during SL failures to recover the vSD-EONs' services according to their SLAs.

Control Plane Latency With SDN Network Hypervisors: The Cost of Virtualization

Software defined networking (SDN) network hypervisors provide the functionalities needed for virtualizing software-defined networks. Hypervisors sit logically between the multiple virtual SDN networks (vSDNs), which reside on the underlying physical SDN network infrastructure, and the corresponding tenant (vSDN) controllers. Different SDN network hypervisor architectures have mainly been explored through proof-of-concept implementations. We fundamentally advance SDN network hypervisor research by conducting a model-based analysis of SDN hypervisor architectures. Specifically, we introduce mixed integer programming formulations for four different SDN network hypervisor architectures. Our model formulations can also optimize the placement of multi-controller switches in virtualized OpenFlow-enabled SDN networks. We employ our models to quantitatively examine the optimal placement of the hypervisor instances. We compare the control plane latencies of the different SDN hypervisor architectures and quantify the cost of virtualization, i.e., the latency overhead due to virtualizing SDN networks via hypervisors. For generalization, we quantify how the hypervisor architectures behave for different network topologies. Our model formulations and the insights drawn from our evaluations inform network operators about the trade-offs of the different hypervisor architectures and help choosing an architecture according to operator demands.

Multidomain Network Hypervisor for Abstraction and Control of OpenFlow-Enabled Multitenant Multitechnology Transport Networks [Invited

A virtualization architecture is presented for deploying multitenant virtual networks on top of multitechnology optical networks. A multidomain network hypervisor (MNH) and a multidomain SDN orchestrator (MSO) are introduced with this purpose. TheMNHprovides an abstract view of each virtual network and gives control of it to each independent customer SDN controller. The MNH is able to provide virtual networks across heterogeneous control domains (i.e., generalized multiprotocol label switching and OpenFlow) and transport technologies (i.e., optical packet switching and elastic optical networks). The MSO is responsible for providing the necessary end-toend connectivity. We have designed, implemented, and experimentally evaluated the MNH and MSO in an international testbed across Spain, UK, Germany, and Japan.

A service-aware virtualized software-defined infrastructure

The Internet infrastructure is gradually improving its flexibility and adaptability due to the incorporation of new promising technologies, such as the software-defined networks and the network function virtualization. The main goal is to meet the diverse communication needs of the users, while the global system operation satisfies the business and societal goals of the infrastructure and service providers. This calls for solutions that consider both local and global network viewpoints and provide sophisticated system control in a stable and predictable way, while being service-aware. We propose a fully integrated solution along these lines: the VLSP, a service-aware software-defined infrastructure for networks and clouds. The VLSP consists of three main distributed systems: a facility performing uniformly logically-centralized management and control of the infrastructure, called the virtual infrastructure management; an information management infrastructure able to maintain an accurate view of the infrastructure environment at both the local and system levels, called the virtual infrastructure information service; and a lightweight virtualization hypervisor able to perform configuration changes in the infrastructure resources, called the lightweight network hypervisor. We discuss representative use-case scenarios, while we demonstrate how VLSP tunes performance trade-offs for particular service demands.

Network Hypervisors: Enhancing SDN Infrastructure

Software-Defined Networking (SDN) has been widely recognized as a promising way to deploy new services and protocols in future networks. The ability to “program” the network enables applications to create innovative new services inside the network itself. However, current SDN programmability comes with downsides that could hinder its adoption and deployment. First, in order to offer complete control, today’s SDN networks provide low-level API’s on which almost any type of service can be written. Because the starting point is a set of low-level API calls, implementing high-level complex services needed by future network applications becomes a challenging task. Second, the set of emerging SDN technologies that are beginning to appear have little in common with one another, making it difficult to set up a flow that traverses multiple SDN technologies/providers.In this paper we propose a new way to set up SDN networks spanning multiple SDN providers. The key to our approach is a Network Hypervisor service. The Network Hypervisor offers high-level abstractions and APIs that greatly simplify the task of creating complex SDN network services. Moreover, the Network Hypervisor is capable of internetworking various SDN providers together under a single interface/abstraction so that applications can establish end-to-end flows without the need to see, or deal with, the differences between SDN providers.