#	Pagina
attuale pagina	/open-h2020/projects/194488/results.html
-1	/open-h2020/projects/193454/results.html

Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - FlexNets (Quantifying Flexibility in Communication Networks)

Teaser

Summary

Communication networks have emerged to become the basic infrastructure for all areas of our society with application areas ranging from social media to industrial production and healthcare. New requirements include the need for dynamic changes of the required resources, for example, to react to social events or to shifts of demands. Existing networks and, in particular, the Internet cannot meet those requirements mainly due to their ossification and hence limitation in resource allocation, i.e., lack of flexibility to adapt the available resources to changes of demands on a small time-scale and in an efficient way.

In recent years, several concepts have emerged in networking research to provide more flexibility in networks through virtualization and control plane programmability. In particular, the split between data plane and a centralized control plane as defined by Software Defined Networking (SDN) is regarded as the basic concept to allow flexibility in networks. Virtualization also became important for the sharing of physical network resources through network virtualization and the softwarization of network functions previously implemented in middleboxes with the concept of Network Function Virtualization (NFV).

However, a deeper understanding of what flexibility means remains open. In this project, flexibility focuses on the dynamic changes of a network that is characterized by its resources (link rate and node capacities), connectivity (network graph) and its network functions with related resources (processing and storage). It is the objective of this research to analyse the fundamental design space for flexibility in SDN-based and virtualized networks with respect to cost such as resource usage, traffic overhead and delay. The outcome will be the definition of a measure for flexibility in communication networks and based thereon a set of quantitative arguments pro and contra certain design choices. An analytical model for the definition of such flexibility measure to quantitatively compare different network design choices and to assess the trade-off for flexibility vs. cost will be developed. To assess flexibility with respect to general graph properties a graph model will be designed. The detailed analysis is based on three use cases: dynamic resource allocation, QoS control in wireless and wired networks, and resilience.

In the state of the art, selected aspects of flexibility have been explored for certain network scenarios, a fundamental and comprehensive analysis is missing. A measure for flexibility and related design guidelines can help the communications industry to better understand the options provided by network softwarization for the support of all areas of our society that relies on communication networks.

Work performed

The work in the first 18 months of this five years project has focused mainly on the following tasks: (1) State of the art analysis and derivation of flexibility properties, (2) definition and discussion of a flexibility measure and cost modeling, (3) setup of an experimental facility for SDN and NFV and a measurement system, (4) detailed use case studies with a focus on dynamic resource allocation and optimization to evaluate and continuously adapt the flexibility measure. All planned work items and milestones have been achieved. Results have been published in a number of conference and journal papers.

State of the art analysis has revealed that the support of flexibility is often used as an argument for a new network design, however only little is explained about of how this flexibility is provided and, moreover, how flexibility can be quantified to compare with other competitive designs. Hence, the available qualitative arguments vary a lot in the literature, which is due to the fact that the use of flexibility as a measure is not well understood in the literature. We close this gap by providing a clear definition of flexibility in this project and we have started to show its practical application along with detailed case studies. For networks, flexibility refers to the ability to adapt the available network resources such as flows or topology to changes. Changes arise due to designers\' demand, such as shorter latency budgets or higher availability, or varying traffic characteristics such as sudden shifts of demands. Our proposed flexibility measure is based on the ratio of changes that could be fulfilled in a given timeframe over the total requested changes. When using flexibility as a measure, we have to be aware that it is unlikely to be a monotonous metric. Flexibility has to be defined on its intended background, for example, flexibility is regarded with respect to the placement of functions in a network. We refer to such background as a flexibility aspect, similar as QoS, which is not a single metric and refers to rate, loss, delay and jitter. Attempting to measure flexibility cannot neglect the costs that might be involved making a network design more flexible. Hence, flexibility shall always be evaluated in light of the costs involved in order to achieve flexibility. For example, providing several data centers in a network setup is likely to be more flexible with respect to changing demands for network functions as function can be potentially migrated to more places. However, operating more data centers also comes at a higher cost.

Use case studies performed to validate the proposed flexibility measure include virtual network embedding, network function placement, static and dynamic controller placement, QoS optimization and resilience. Each use case study focuses on a certain system setup that is optimized for certain design parameters. Flexibility is used as an evaluation parameter in simulation to compare different design choices such as number of data centers or number of SDN controllers.

In order to be able to practically evaluate design choices, an experimental facility consisting of 10 bare metal switches has been setup. The bare metal switches can be operated with different firmwares to run OpenFlow and also other network operating systems. To be able to measure network characteristics and network behavior, a carrier grade measurement system has been setup.

Final results

In the state of the art, selected aspects of flexibility have been explored for certain network scenarios, a fundamental and comprehensive analysis or even a measure for flexibility is missing. In the first 18 months of the project, a measure for flexibility in networks has been proposed and validated through use case studies. In those case studies different network designs have been challenged with change requests and their settings have been optimized to react to the changes in order to evaluate their flexibility with respect to the demanded changes. In this way the flexibility measure is continuously validated and revised. Publications and presentations of those case studies begin to raise the interest in fundamental flexibility analysis in the research community, which became visible through several invitations to keynote presentations.

With respect to the state of the art the use case studies themselves provide an important contribution based on their proposition and evaluation of potentially new network designs considering flexibility. The new flexibility measure has the potential to create a fundamental new thinking in network evaluation, in particular for the new concepts of network softwarization. For future network design with emerging applications, flexibility plays an important role to stay competitive for operators and for the society as a whole to be able to adopt new technologies that rely on communication networks quickly. Related design guidelines can help the communications industry to better understand the options provided by network softwarization for the support of all areas of our society that rely on communication networks.