Michael Duelli

Energy Efficient Virtual Network Embedding

Waste of energy due to over-provisioning and over-dimensioning of network infrastructures has recently stimulated the interest on energy consumption reduction by Internet Service Providers (ISPs). By means of resource consolidation, network virtualization based architectures will enable energy saving. In this letter, we extend the well-known virtual network embedding problem (VNE) to energy awareness and propose a mixed integer program (MIP) which provides optimal energy efficient embeddings. Simulation results show the energy gains of the proposed MIP over the existing cost-based VNE approach.

Resilience Analysis of Packet-Switched Communication Networks

In the presence of local network outages, restoration and protection switching mechanisms redirect the traffic over alternative paths to mitigate the effect of failures. However, some failure combinations still lead to loss of ingress-egress connectivity within a network or to severe congestion due to rerouted traffic. Congestion may also be caused by unexpected traffic shifts due to changed user behavior or due to changes of interdomain routing. This paper presents a framework for the analysis of ingress-egress unavailability and link congestion due to: 1) failures; 2) changes of user behavior; and 3) changed interdomain routing. It proposes algorithms to find the most probable combinations of 1)-3) according to some models, and to evaluate the connectivity and the relative link load of the network under these conditions. We have implemented this concept in a software tool and its visualization of the results leads to a comprehensive view of the networks resilience. It helps to anticipate potential ingress-egress disconnection and congestion before failures and overload occur or before planned modifications (new infrastructure, new routing, new customers) take effect. Thus, it detects weak points in a network, predicts the effectiveness of potential upgrades, and thereby supports careful bandwidth overprovisioning.

ALEVIN - A Framework to Develop, Compare, and Analyze Virtual Network Embedding Algorithms

Network virtualization is recognized as an enabling technology for the Future Internet. Applying virtualization of network resources leads to the problem of mapping virtual resources to physical resources, known as “Virtual Network Embedding” (VNE). Several algorithms attempting to solve this problem have been discussed in the literature, so far. However, comparison of VNE algorithms is hard, as each algorithm focuses on different criteria. To that end, we introduce a framework to compare different algorithms according to a set of metrics, which allow to evaluate the algorithms and compute their results on a given scenario for arbitrary parameters.

Provisioning and Operation of Virtual Networks

In todays Internet, requirements of services regarding the underlying transport network are very diverse. In the future, this diversity will increase and make it harder to accommodate all services in a single network. A possible approach to keep up with this diversity in future networks is the deployment of isolated, custom tailored networks on top of a single shared physical substrate. The COMCON (COntrol and Monitoring of COexisting Networks) project aims to define a reference architecture for setup, control, and monitoring of virtual networks on a provider- and operator-grade level. In this paper, we present the building blocks and interfaces of our architecture.

CAPEX-Aware Design of Survivable DWDM Mesh Networks

This paper reviews the basic architecture and component costs of opaque, transparent, and semitransparent DWDM networks and looks at the network design problem from a capital expenditure (CAPEX) point of view. Given are a fiber topology and a demand matrix with different bit rates. Required is the least-cost optical equipment for that topology together with the routing and potential muxponder-based aggregation of all demands such that they can be supported by the newly designed network. We look at the problem for networks without resilience requirements and for survivable networks using 1+1 protection against single fiber cuts. We model this problem for the three types of optical networks by integer linear programs (ILPs) in a canonical way.

Performance comparison of hardware virtualization platforms

Hosting virtual servers on a shared physical hardware by means of hardware virtualization is common use at data centers, web hosters, and research facilities. All platforms include isolation techniques that restrict resource consumption of the virtual guest machines. However, these isolation techniques have an impact on the performance of the guest systems. In this paper, we study how popular hardware virtualization approaches (OpenVZ, KVM, Xen v4, VirtualBox, VMware ESXi) affect the network throughput of a virtualized system. We compare their impact in a dedicated and a shared host scenario as well as to the bare host system. Our results provide an overview on the performance of popular hardware virtualization platforms on commodity hardware in terms of network throughput.

VNREAL: Virtual network resource embedding algorithms in the framework ALEVIN

Network virtualization is recognized as an enabling technology for the Future Internet that overcomes network ossification. However, it introduces a set of challenges. In any network virtualization environment, the problem of optimally mapping virtual resources to physical resources, known as virtual network embedding (VNE), is a critical challenge. Several algorithms attempting to solve this problem have been proposed in literature, so far. However, comparison of existing and new VNE algorithms is hard, as each algorithm focuses on different criteria. To that end, the VNREAL project introduces ALEVIN, a framework to compare different algorithms according to a set of metrics, easily incorporate new VNE algorithms, and evaluate these algorithms on a given scenario for arbitrary parameters.

Greedy design of resilient multi-layer networks

In this paper, we propose a deterministic greedy heuristic providing a construction layout for a cost-efficient multi-layer network that is able to carry a given set of traffic demands with and without protection on different layers. We apply the heuristic to different reference network topologies and protection requirements. Evaluations are conducted regarding equipment cost on different layers, blocking probability, path lengths, and number of demands affected by specific failures.

Minimizing Installation Costs of Survivable DWDM-Mesh Networks: A Heuristic Approach

The cost function for the capacity of optical links follows a step function. That means, the support of one more lightpath might require a costly upgrade of an optical cross connect (OXC), but then additional lightpaths can be supported at almost no further cost. This should be considered when lightpaths are routed through an optical network. In this paper we optimize the routing of the lightpaths to minimize the costs for the required optical equipment. We consider this problem for the failure-free case only and for survivable networks using dedicated path protection. We formulate the problems by integer linear programs (ILPs). In addition, we propose heuristics to solve the problem since solving ILPs is computationally expensive and not feasible for large problem instances. We show that our heuristics lead to good results within a fraction of time compared to ILP solvers.

Three Methods for Optimizing Single-Shortest Path Routing

Intra-domain routing in IP networks is based on the shortest path principle by assigning administrative weights (costs) to links. The resulting least-cost paths determine routes between pairs of routers. If several such equal-cost paths exist between a pair of routers, it may not be clear which of them is actually used to route traffic. This makes it difficult to predict the network traffic flow distribution. Therefore, the selected link costs should assure uniqueness of the shortest paths. On top of that, the link costs can be optimized with respect to some traffic objective. The resulting optimization problem, referred to as SSPP, turns out to be NP-hard. SSPP can be formulated as a mixed-integer programming problem and, as such, solved with branch-and- bound (B&B). In this paper, we consider three methods for SSPP. Two of them are exact methods based on B&B, namely branch- and-cut and constraint programming. Since the exact solutions of SSPP may require excessive computation time and may not always be effective when applied to practical networks, we also study a fast heuristic method. Finally, in a numerical study, we compare the effectiveness of the three approaches.