Carela, V.; Barlet, P.; Mulla , O.; Solé-Pareta, J. Journal of network and systems management Vol. 23, num. 3, p. 401-419 DOI: 10.1007/s10922-013-9293-1 Data de publicació: 2015-07-01 Article en revista
Traffic classification is an important aspect in network operation and management, but challenging from a research perspective. During the last decade, several works have proposed different methods for traffic classification. Although most proposed methods achieve high accuracy, they present several practical limitations that hinder their actual deployment in production networks. For example, existing methods often require a costly training phase or expensive hardware, while their results have relatively low completeness. In this paper, we address these practical limitations by proposing an autonomic traffic classification system for large networks. Our system combines multiple classification techniques to leverage their advantages and minimize the limitations they present when used alone. Our system can operate with Sampled NetFlow data making it easier to deploy in production networks to assist network operation and management tasks. The main novelty of our system is that it can automatically retrain itself in order to sustain a high classification accuracy along time. We evaluate our solution using a 14-day trace from a large production network and show that our system can sustain an accuracy <96 %, even in presence of sampling, during long periods of time. The proposed system has been deployed in production in the Catalan Research and Education network and it is currently being used by network managers of more than 90 institutions connected to this network.
Pages, A.; Buttaboni, A.; Maier, G.; Siracusa, D.; Perello, J.; Spadaro, S. Journal of network and systems management Vol. 22, num. 3, p. 462-487 DOI: 10.1007/s10922-014-9304-x Data de publicació: 2014-07-01 Article en revista
In this paper, we propose a novel inter-domain connection provisioning mechanism for multi-domain translucent Wavelength Switched Optical Networks. The mechanism can operate in hierarchical and non-hierarchical path computation element-based architectures. It is based on a novel approach to derive the domain abstract topologies jointly with an inter-domain routing algorithm. The overall objective is to perform end-to-end route computations for incoming connections. We do so by taking into account the current load of the physical links and some energy parameters of the optical network devices with the purpose of distributing the load (as much as possible) along the physical links and keeping the energy consumption low. The performance of the whole mechanism is highlighted through extensive and illustrative simulation results that benchmark it against shortest path-based techniques.
Doyen, G.; Waldburger, M.; Sperotto, A.; Celeda, P.; Gorricho, J.; Schaaf, T.; Serrat, J. Journal of network and systems management Vol. 21, num. 4, p. 1-8 DOI: 10.1007/s10922-013-9298-9 Data de publicació: 2013-12-31 Article en revista
This article contains the report on AIMS 2013, which was held on June 25–28, 2013 at the Universitat Politècnica de Catalunya (UPC), Spain and was driven by the theme “Emerging Management Mechanisms for the Future Internet”. It covers the three main parts that formed the event program: the keynote and technical sessions of the main track, the PhD workshop and the hands-on tutorials. Finally, the report highlights the evolutions that would shape the future version of AIMS.
Presence information was the key enabler of the great success of instant messaging applications. This information was first limited to simple binary states such as online, offline, or busy. However, it is currently evolving towards a much more generic, flexible concept that includes all context that enables users or applications to adapt and control communications in a more efficient, personalized manner. Presence paves the way for the deployment of advanced communication services and is therefore a key component in Next-Generation Networks. The main barrier to the implantation of next-generation presence-based services is the great amount of traffic involved in the distribution of presence information across domains. In this paper, we give a deep insight into multiple strategies for reducing inter-domain presence traffic. We estimate the capacity demands of these strategies on the IMS servers. Moreover, we enhance two strategies for reducing presence load: Common Subscribe and Federated Common Subscribe. Not only do the proposed enhancements prevent the disclosure of some privacy, they also save presence traffic.