Krzysztof Walkowiak

Routing and Spectrum Assignment in Spectrum Sliced Elastic Optical Path Network

A spectrum-sliced elastic optical path network (SLICE) architecture has been recently proposed as an efficient solution for a flexible bandwidth allocation in optical networks. In SLICE, the problem of Routing and Spectrum Assignment (RSA) emerges. In this letter, we both formulate RSA as an Integer Linear Programming (ILP) problem and propose an effective heuristic to be used if the solution of ILP is not attainable.

On the advantages of elastic optical networks for provisioning of cloud computing traffic

This article provides motivation for the elastic optical network (EON) approach, an efficient and cost-effective solution for provisioning of cloud computing traffic. As opposed to wavelength switched optical networks (WSONs), the capabilities of which are limited by the use of rigid frequency grids, EON architectures allow for both scalable bandwidth provisioning and flexible resource allocation. The deployment cost, energy consumption, and bandwidth usage for both EON and classical WSON transport networks are compared in pan-European and U.S. backbone networks for 2012-2020 using Cisco traffic predictions. Results show that the EON concept significantly outperforms WSON in all examined criteria, and the gap between the two architectures increases in subsequent years. Moreover, potential advantages of anycast routing in transport networks with data center traffic are demonstrated.

Routing and spectrum allocation algorithms for elastic optical networks with dedicated path protection

Abstract Elastic optical network (EON) architectures have been recently proposed as a candidate solution for provisioning of both huge bandwidth and flexible connections in next generation optical networks. In this paper, we focus on survivable EON scenarios and, in particular, we address an offline problem of routing and spectrum allocation (RSA) with dedicated path protection (DPP) in EON. We formulate RSA/DPP as an Integer Linear Programming (ILP) problem. Since RSA is a difficult problem itself, we propose to apply a metaheuristic approach to provide near-optimal solutions to RSA/DPP. Namely, we develop a Tabu Search-based algorithm (TS), and a hybrid Adaptive Frequency Assignment-TS (AFA/TS) algorithm. We investigate the efficiency of the algorithms for a set of network and DPP scenarios and we show that the proposed algorithms outperform other reference algorithms. Eventually, we present some comparative results for different path protection scenarios.

Anycasting in connection-oriented computer networks: Models, algorithms and results

Anycasting in connection-oriented computer networks: Models, algorithms and results Our discussion in this article centers around various issues related to the use of anycasting in connection-oriented computer networks. Anycast is defined as a one-to-one-of-many transmission to deliver a packet to one of many hosts. Anycasting can be applied if the same content is replicated over many locations in the network. Examples of network techniques that apply anycasting are Content Delivery Networks (CDNs), Domain Name Service (DNS), Peer-to-Peer (P2P) systems. The role of anycasting is growing concurrently with the popularity of electronic music, movies, and other content required by Internet users. In this work we focus on the optimization of anycast flows in connection-oriented networks. We formulate a model of anycast connections and next propose a heuristic algorithm based on the Lagrangean relaxation aimed to optimize jointly routes for anycast and unicast connections. Results of numerical experiments are presented and evaluated. Finally, we analyze briefly problems related to anycasting in dynamic routing and multi-layer networks.

Joint anycast and unicast routing for elastic optical networks: Modeling and optimization

Anycast communication is gaining much interest due to the deployment and the growing importance of new network services, including Content Delivery Networks (CDNs) and grid/cloud computing, among others. Concurrently, elastic optical networks (EONs), being a very promising solution for future optical transport networks, are considered as effective and cost-efficient approach for supporting such bandwidth-demanding applications. In this article, we focus on anycast communication in EON. Specifically, we formulate a joint anycast and unicast Routing and Spectrum Allocation (RSA/JAU) off-line optimization problem and solve it by means of both Integer Linear Programming (ILP) and dedicated heuristic algorithms. We evaluate the performance of algorithms and assess the effectiveness of anycast communication under various network scenarios. The main conclusion is that anycast routing brings significant spectrum savings in EON.

Offline RSA algorithms for elastic optical networks with dedicated path protection consideration

In this article, we address the problem of static routing and spectrum allocation (RSA) in an elastic optical network (EON) with dedicated path protection (DPP) consideration. We formulate RSA-DPP as an Integer Linear Programming (ILP) problem. Since RSA is a difficult problem, we propose an Adaptive Frequency Assignment with Dedicated Path Protection (AFA-DPP) heuristic algorithm to provide near-optimal solutions to the problem. We investigate the efficiency of AFA-DPP for a set of network scenarios. Evaluation results show that the proposed algorithm outperforms other reference algorithms.

Optimization of Multicast Traffic in Elastic Optical Networks With Distance-Adaptive Transmission

In this letter, we propose three new approaches to optimizing multicast traffic in elastic optical networks that implement distance-adaptive transmission. Two of the approaches are based on integer linear programming modeling, whereas the last one is an effective heuristic. We also compare our approaches with reference methods proposed in the literature. The comparison is based on numerical experiments supported by a discussion on the effectiveness and applicability of the considered methods. The algorithms based on candidate tree modeling of multicasting outperform other methods for larger problem instances in terms of spectrum usage and applicability.

QoS dynamic routing in content delivery networks

Recently, much research in quality of service (QoS) routing has focused on the unicast communication technique. However, Content Delivery Network (CDN) approach becomes popular because CDN enables effective and inexpensive improvement of Internet service quality. Therefore, we analyze in this paper a network processing two kinds of demands: content demands to CDN servers and standard unicast demands. Since MPLS defines effective mechanism for traffic engineering, we assume that the CDN is located in MPLS network. To examine the QoS performance of CDN-enabled MPLS network we propose new constraint-based algorithms for CDN server selection and evaluate relative performance of these algorithms and the most effective existing QoS unicast routing algorithms in terms of the demand rejection ratio.

Off-line algorithms for Routing, Modulation Level, and Spectrum Assignment in elastic optical networks

Optical network architectures with elastic bandwidth provisioning are a very promising approach for next generation optical networks. Such elastic optical networks will enable efficient resource utilization and flexible sub-wavelength and super-channel connection provisioning to support heterogeneous and immense bandwidth demands. In this paper, we focus on the problem of Routing, Modulation Level, and Spectrum Allocation/Assignment (RMLSA) which emerges in such networks. We both formulate RMLSA as an Integer Linear Programming (ILP) problem and propose an effective heuristic method called Adaptive Frequency Assignment - Division and Collision Avoidance (AFA-DCA) to be used if the solution of ILP is not attainable.

Tabu search algorithm for routing, modulation and spectrum allocation in elastic optical network with anycast and unicast traffic

The paper presents a novel method based on the standard tabu search (TS) approach, dedicated to solve the routing, modulation and spectrum allocation (RMSA) problem in elastic optical networks (EONs). The considered formulation of the RMSA problem covers simultaneously unicast (one-to-one) and anycast (one-to-one-of-many) traffic demands. This is a very important issue taking into account the fact that anycasting gains more and more importance in contemporary Internet due the growing popularity of services like cloud computing, content delivery networks, and video streaming. In this paper, we formulate RMSA as an integer linear programming (ILP) problem and we study four different objective functions, which are related to, respectively, cost, power consumption, maximum and average spectrum usage. We evaluate the performance of our TS method based on the comparison with both optimal results yielded by the CPLEX solver and the results obtained by reference heuristic algorithms proposed in the literature. Moreover, we evaluate benefits of the use of anycasting in EONs. The performed simulation experiments demonstrate that the proposed algorithm outperforms other reference methods. What is more, we show that the anycast transmission can provide significant savings compared to the typical unicast transmission.