<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd" xmlns:dc="http://purl.org/dc/elements/1.1/" version="2.0">
  <channel>
    <title>DSpace Collection:</title>
    <link>http://hdl.handle.net/2117/3099</link>
    <description />
    <pubDate>Fri, 24 May 2013 02:26:14 GMT</pubDate>
    <dc:date>2013-05-24T02:26:14Z</dc:date>
    <itunes:owner>
      <itunes:email>webmaster.bupc@upc.edu</itunes:email>
      <itunes:name>Universitat Politècnica de Catalunya. Servei de Biblioteques i Documentació</itunes:name>
    </itunes:owner>
    <itunes:explicit>no</itunes:explicit>
    <itunes:keywords />
    <item>
      <title>Prediction-based routing as RWA in multilayer traffic engineering</title>
      <link>http://hdl.handle.net/2117/16382</link>
      <description>Title: Prediction-based routing as RWA in multilayer traffic engineering
Authors: Puype, Bart; Marín Tordera, Eva; Colle, D.; Sánchez López, Sergio; Pickavet, Mario; Masip Bruin, Xavier; Demeester, Piet
Abstract: Multilayer traffic engineering (MLTE) allows&#xD;
coping with ever-increasing and varying traffic demands in IP-over-Optical multilayer networks. It utilizes cross-layer&#xD;
TE (Traffic Engineering) techniques to provision optical lightpath capacity to the IP/MPLS (Internet Protocol/ Multi-Protocol Label Switching) logical topology on-demand.&#xD;
Such provisioning however causes optical connection arrival rates that pose strong performance requirements to Routing&#xD;
and Wavelength Assignment (RWA) strategies. Collecting up-to-date network information for the RWA with rapidly changing network states can be quite difficult. Exposing optical layer state information to the IP layer in the overlay model, or transforming this optical layer information in a workable&#xD;
representation in an integrated control plane is similarly problematic.&#xD;
Prediction-Based Routing (PBR) has been proposedas a RWA mechanism for optical transport networks; it bases routing not on possibly inaccurate or outdated network state, but instead on previous connections set-up. In this article, we propose to implement PBR as the RWA mechanism in the&#xD;
optical layer of a multilayer network, and use the predictive capabilities of PBR to expose dynamic optical network information&#xD;
into the multilayer traffic engineering algorithm with minimal control plane overhead. Some simulations show the benefits of using the PBR in the optical layer for MLTE purposes.</description>
      <pubDate>Tue, 31 Jul 2012 10:33:56 GMT</pubDate>
      <guid isPermaLink="false">http://hdl.handle.net/2117/16382</guid>
      <dc:date>2012-07-31T10:33:56Z</dc:date>
      <itunes:author>Puype, Bart; Marín Tordera, Eva; Colle, D.; Sánchez López, Sergio; Pickavet, Mario; Masip Bruin, Xavier; Demeester, Piet</itunes:author>
      <itunes:explicit>no</itunes:explicit>
      <itunes:keywords>Multilayer traffic engineering, RWA, Prediction-based routing, OTN, ASON</itunes:keywords>
      <itunes:summary>Multilayer traffic engineering (MLTE) allows&#xD;
coping with ever-increasing and varying traffic demands in IP-over-Optical multilayer networks. It utilizes cross-layer&#xD;
TE (Traffic Engineering) techniques to provision optical lightpath capacity to the IP/MPLS (Internet Protocol/ Multi-Protocol Label Switching) logical topology on-demand.&#xD;
Such provisioning however causes optical connection arrival rates that pose strong performance requirements to Routing&#xD;
and Wavelength Assignment (RWA) strategies. Collecting up-to-date network information for the RWA with rapidly changing network states can be quite difficult. Exposing optical layer state information to the IP layer in the overlay model, or transforming this optical layer information in a workable&#xD;
representation in an integrated control plane is similarly problematic.&#xD;
Prediction-Based Routing (PBR) has been proposedas a RWA mechanism for optical transport networks; it bases routing not on possibly inaccurate or outdated network state, but instead on previous connections set-up. In this article, we propose to implement PBR as the RWA mechanism in the&#xD;
optical layer of a multilayer network, and use the predictive capabilities of PBR to expose dynamic optical network information&#xD;
into the multilayer traffic engineering algorithm with minimal control plane overhead. Some simulations show the benefits of using the PBR in the optical layer for MLTE purposes.</itunes:summary>
    </item>
    <item>
      <title>OBGP+: An improved path-vector protocol for multi-domain optical networks</title>
      <link>http://hdl.handle.net/2117/9655</link>
      <description>Title: OBGP+: An improved path-vector protocol for multi-domain optical networks
Authors: Yannuzzi, Marcelo; Masip Bruin, Xavier; Sánchez López, Sergio; Marín Tordera, Eva
Abstract: One of the essential components for the dynamic provisioning of lightpaths across multiple domains is the Routing and Wavelength Assignment (RWA) strategy adopted. The consolidation that path-vector protocols have had in practice, has motivated the optical extension to BGP (OBGP). We claim, however, that a routing model mostly centered on the exchange of reachability information–like the one we have today with BGP or the one offered by OBGP–will not be sufficient for multi-domain optical networks. Routing domains must be able to exchange both reachability as well as aggregated Path-State Information (PSI). Understanding that this is a missing piece in the routing models provided by BGP and OBGP is easy nowadays, but contributing with solutions capable of highly improving the performance of a path-vector without impacting on key aspects of the protocol–fundamentally, its scalability, its convergence properties, and the number of routing messages exchanged between domains–is a challenging task.&#xD;
&#xD;
In this paper we propose OBGP +, which is a very simple extension of a path-vector protocol supporting the computation and advertisement of PSI between optical domains. The PSI that we propose to use is highly condensed in the form of a single integer value. In order to avoid the typical increase in the number of routing messages associated with the update of PSI, we propose to piggy-back the updates in non-dummy Keepalive messages exchanged between OBGP+ neighbors. Extensive simulations reveal that, despite its simplicity: (i) OBGP+ is able to drastically reduce the blocking experienced with a path-vector protocol like OBGP; (ii) OBGP+ needs much less number of routing messages than OBGP to achieve such performance; and (iii) the convergence and restoration features of OBGP+ are also better than those of OBGP, which is particularly important for connections that lack a protection path.</description>
      <pubDate>Wed, 13 Oct 2010 09:49:59 GMT</pubDate>
      <guid isPermaLink="false">http://hdl.handle.net/2117/9655</guid>
      <dc:date>2010-10-13T09:49:59Z</dc:date>
      <itunes:author>Yannuzzi, Marcelo; Masip Bruin, Xavier; Sánchez López, Sergio; Marín Tordera, Eva</itunes:author>
      <itunes:explicit>no</itunes:explicit>
      <itunes:keywords />
      <itunes:summary>One of the essential components for the dynamic provisioning of lightpaths across multiple domains is the Routing and Wavelength Assignment (RWA) strategy adopted. The consolidation that path-vector protocols have had in practice, has motivated the optical extension to BGP (OBGP). We claim, however, that a routing model mostly centered on the exchange of reachability information–like the one we have today with BGP or the one offered by OBGP–will not be sufficient for multi-domain optical networks. Routing domains must be able to exchange both reachability as well as aggregated Path-State Information (PSI). Understanding that this is a missing piece in the routing models provided by BGP and OBGP is easy nowadays, but contributing with solutions capable of highly improving the performance of a path-vector without impacting on key aspects of the protocol–fundamentally, its scalability, its convergence properties, and the number of routing messages exchanged between domains–is a challenging task.&#xD;
&#xD;
In this paper we propose OBGP +, which is a very simple extension of a path-vector protocol supporting the computation and advertisement of PSI between optical domains. The PSI that we propose to use is highly condensed in the form of a single integer value. In order to avoid the typical increase in the number of routing messages associated with the update of PSI, we propose to piggy-back the updates in non-dummy Keepalive messages exchanged between OBGP+ neighbors. Extensive simulations reveal that, despite its simplicity: (i) OBGP+ is able to drastically reduce the blocking experienced with a path-vector protocol like OBGP; (ii) OBGP+ needs much less number of routing messages than OBGP to achieve such performance; and (iii) the convergence and restoration features of OBGP+ are also better than those of OBGP, which is particularly important for connections that lack a protection path.</itunes:summary>
    </item>
    <item>
      <title>Toward a new route control model for multidomain optical networks</title>
      <link>http://hdl.handle.net/2117/9629</link>
      <description>Title: Toward a new route control model for multidomain optical networks
Authors: Yannuzzi, Marcelo; Masip Bruin, Xavier; Fabrego, G; Sánchez López, Sergio; Sprintson, A; Orda, A
Abstract: The design of the control plane model for&#xD;
multidomain optical networks poses complex&#xD;
challenges and introduces many open problems.&#xD;
Some initiatives have proposed Optical BGP&#xD;
(OBGP), which is an extension of BGP supporting the advertisement and signaling of optical information between routing domains. We argue, however, that future optical networks offer the opportunity to avoid inheriting the limitations of&#xD;
BGP, especially in terms of routing and traffic engineering control. In this article we present a route control model replacing BGP/OBGP.&#xD;
Extensive simulations confirm that our route&#xD;
control model is able to drastically reduce the blocking experienced with OBGP, and this can be accomplished without increasing the number or frequency of routing updates exchanged between domains.</description>
      <pubDate>Mon, 11 Oct 2010 16:18:17 GMT</pubDate>
      <guid isPermaLink="false">http://hdl.handle.net/2117/9629</guid>
      <dc:date>2010-10-11T16:18:17Z</dc:date>
      <itunes:author>Yannuzzi, Marcelo; Masip Bruin, Xavier; Fabrego, G; Sánchez López, Sergio; Sprintson, A; Orda, A</itunes:author>
      <itunes:explicit>no</itunes:explicit>
      <itunes:keywords />
      <itunes:summary>The design of the control plane model for&#xD;
multidomain optical networks poses complex&#xD;
challenges and introduces many open problems.&#xD;
Some initiatives have proposed Optical BGP&#xD;
(OBGP), which is an extension of BGP supporting the advertisement and signaling of optical information between routing domains. We argue, however, that future optical networks offer the opportunity to avoid inheriting the limitations of&#xD;
BGP, especially in terms of routing and traffic engineering control. In this article we present a route control model replacing BGP/OBGP.&#xD;
Extensive simulations confirm that our route&#xD;
control model is able to drastically reduce the blocking experienced with OBGP, and this can be accomplished without increasing the number or frequency of routing updates exchanged between domains.</itunes:summary>
    </item>
    <item>
      <title>Reducing the effects of routing inaccuracy by means of prediction and an innovative link-state cost</title>
      <link>http://hdl.handle.net/2117/6923</link>
      <description>Title: Reducing the effects of routing inaccuracy by means of prediction and an innovative link-state cost
Authors: Masip Bruin, Xavier; Marín Tordera, Eva; Yannuzzi, Marcelo; Serral Gracià, René; Sánchez López, Sergio
Abstract: The routing inaccuracy problem is one of the major&#xD;
issues impeding the evolution and deployment of Constraint-&#xD;
Based Routing (CBR) techniques. This paper proposes a promising&#xD;
CBR strategy that combines the strengths of prediction with&#xD;
an innovative link-state cost. The latter explicitly integrates a&#xD;
two-bit counter predictor, with a novel metric that stands for&#xD;
the degree of inaccuracy (seen by the source node) of the state&#xD;
information associated with the links along a path. In our routing&#xD;
model, Link-State Advertisements (LSAs) are only distributed&#xD;
upon topological changes in the network, i.e., the state and&#xD;
availability of network resources along a path are predicted&#xD;
from the source rather than updated through conventional LSAs.&#xD;
As a proof-of-concept, we apply our routing strategy in the&#xD;
context of circuit-switched networks. We show that our approach&#xD;
considerably reduces the impact of routing inaccuracy on the&#xD;
blocking probability, while eliminating the typical LSAs caused&#xD;
by the traffic dynamics in CBR protocols.</description>
      <pubDate>Wed, 14 Apr 2010 09:21:54 GMT</pubDate>
      <guid isPermaLink="false">http://hdl.handle.net/2117/6923</guid>
      <dc:date>2010-04-14T09:21:54Z</dc:date>
      <itunes:author>Masip Bruin, Xavier; Marín Tordera, Eva; Yannuzzi, Marcelo; Serral Gracià, René; Sánchez López, Sergio</itunes:author>
      <itunes:explicit>no</itunes:explicit>
      <itunes:keywords />
      <itunes:summary>The routing inaccuracy problem is one of the major&#xD;
issues impeding the evolution and deployment of Constraint-&#xD;
Based Routing (CBR) techniques. This paper proposes a promising&#xD;
CBR strategy that combines the strengths of prediction with&#xD;
an innovative link-state cost. The latter explicitly integrates a&#xD;
two-bit counter predictor, with a novel metric that stands for&#xD;
the degree of inaccuracy (seen by the source node) of the state&#xD;
information associated with the links along a path. In our routing&#xD;
model, Link-State Advertisements (LSAs) are only distributed&#xD;
upon topological changes in the network, i.e., the state and&#xD;
availability of network resources along a path are predicted&#xD;
from the source rather than updated through conventional LSAs.&#xD;
As a proof-of-concept, we apply our routing strategy in the&#xD;
context of circuit-switched networks. We show that our approach&#xD;
considerably reduces the impact of routing inaccuracy on the&#xD;
blocking probability, while eliminating the typical LSAs caused&#xD;
by the traffic dynamics in CBR protocols.</itunes:summary>
    </item>
    <item>
      <title>Improving the Performance of Route Control Middleboxes in a Competitive Environment</title>
      <link>http://hdl.handle.net/2117/6861</link>
      <description>Title: Improving the Performance of Route Control Middleboxes in a Competitive Environment
Authors: Yannuzzi, Marcelo; Masip Bruin, Xavier; Marín Tordera, Eva; Domingo Pascual, Jordi; Fonte, Alexandre; Monteiro, Edmundo
Abstract: Multihomed subscribers are increasingly adopting intelligent route control solutions&#xD;
to optimize the cost and end-to-end performance of the traffic routed among the&#xD;
different links connecting their networks to the Internet. Until recently, IRC practices&#xD;
were not considered adverse, but new studies show that in a competitive environment,&#xD;
they can lead to persistent traffic oscillations, causing significant performance&#xD;
degradation rather than improvements. To cope with this, randomized IRC techniques&#xD;
were proposed. However, the proliferation of IRC products raises concerns,&#xD;
given that randomization becomes less effective as the number of interfering IRC&#xD;
systems increases. In this article, we present a more scalable route control strategy&#xD;
that can better support the foreseeable spread of IRC solutions. We show that by&#xD;
blending randomization with adaptive filtering techniques, it is possible to drastically&#xD;
reduce the interference between competing route controllers, and this can be&#xD;
achieved without penalizing the end-to-end traffic performance. In addition to the&#xD;
potential improvements in terms of scalability and performance, the route control&#xD;
strategy outlined here has various practical advantages. For instance, it does not&#xD;
require any kind of protocol or coordination between the competing IRC middleboxes,&#xD;
and it can be adopted readily today because the only requirement is a software&#xD;
upgrade of the available route controllers.</description>
      <pubDate>Tue, 06 Apr 2010 15:19:08 GMT</pubDate>
      <guid isPermaLink="false">http://hdl.handle.net/2117/6861</guid>
      <dc:date>2010-04-06T15:19:08Z</dc:date>
      <itunes:author>Yannuzzi, Marcelo; Masip Bruin, Xavier; Marín Tordera, Eva; Domingo Pascual, Jordi; Fonte, Alexandre; Monteiro, Edmundo</itunes:author>
      <itunes:explicit>no</itunes:explicit>
      <itunes:keywords />
      <itunes:summary>Multihomed subscribers are increasingly adopting intelligent route control solutions&#xD;
to optimize the cost and end-to-end performance of the traffic routed among the&#xD;
different links connecting their networks to the Internet. Until recently, IRC practices&#xD;
were not considered adverse, but new studies show that in a competitive environment,&#xD;
they can lead to persistent traffic oscillations, causing significant performance&#xD;
degradation rather than improvements. To cope with this, randomized IRC techniques&#xD;
were proposed. However, the proliferation of IRC products raises concerns,&#xD;
given that randomization becomes less effective as the number of interfering IRC&#xD;
systems increases. In this article, we present a more scalable route control strategy&#xD;
that can better support the foreseeable spread of IRC solutions. We show that by&#xD;
blending randomization with adaptive filtering techniques, it is possible to drastically&#xD;
reduce the interference between competing route controllers, and this can be&#xD;
achieved without penalizing the end-to-end traffic performance. In addition to the&#xD;
potential improvements in terms of scalability and performance, the route control&#xD;
strategy outlined here has various practical advantages. For instance, it does not&#xD;
require any kind of protocol or coordination between the competing IRC middleboxes,&#xD;
and it can be adopted readily today because the only requirement is a software&#xD;
upgrade of the available route controllers.</itunes:summary>
    </item>
    <item>
      <title>Managing Interdomain Traffic in Latin America: A New Perspective based on LISP</title>
      <link>http://hdl.handle.net/2117/6767</link>
      <description>Title: Managing Interdomain Traffic in Latin America: A New Perspective based on LISP
Authors: Yannuzzi, Marcelo; Masip Bruin, Xavier; Grampin Castro, Eduardo; Gagliano Molla, Roque; Castro Casales, Alberto; Germán Duarte, Martín
Abstract: The characteristics of Latin American network&#xD;
infrastructures have global consequences,&#xD;
particularly in the area of interdomain traffic&#xD;
engineering. As an example, Latin America&#xD;
shows the largest de-aggregation factor of IP&#xD;
prefixes among all regional Internet registries,&#xD;
being proportionally the largest contributor to&#xD;
the growth and dynamics of the global BGP&#xD;
routing table. In this article we analyze the&#xD;
peculiarities of LA interdomain routing architecture,&#xD;
and provide up-to-date data about the&#xD;
combined effects of the multihoming and TE&#xD;
practices in the region. We observe that the&#xD;
Internet Research Task Force initiative on the&#xD;
separation of the address space into locators&#xD;
and identifiers can not only alleviate the growth&#xD;
and dynamics of the global routing table, but&#xD;
can also offer appealing TE opportunities for&#xD;
LA. We outline one of the solutions under discussion&#xD;
at the IRTF, the Locator/Identifier&#xD;
Separation Protocol, and examine its potential&#xD;
in terms of interdomain traffic management in&#xD;
the context of LA. The key advantage of LISP&#xD;
is its nondisruptive nature, but the existing proposals&#xD;
for its control plane have some problems&#xD;
that may hinder its possible deployment. In&#xD;
light of this, we introduce a promising control&#xD;
plane for LISP that can solve these issues, and&#xD;
at the same time has the potential to bridge the&#xD;
gap between intradomain and interdomain traffic&#xD;
management.</description>
      <pubDate>Mon, 22 Mar 2010 18:04:13 GMT</pubDate>
      <guid isPermaLink="false">http://hdl.handle.net/2117/6767</guid>
      <dc:date>2010-03-22T18:04:13Z</dc:date>
      <itunes:author>Yannuzzi, Marcelo; Masip Bruin, Xavier; Grampin Castro, Eduardo; Gagliano Molla, Roque; Castro Casales, Alberto; Germán Duarte, Martín</itunes:author>
      <itunes:explicit>no</itunes:explicit>
      <itunes:keywords />
      <itunes:summary>The characteristics of Latin American network&#xD;
infrastructures have global consequences,&#xD;
particularly in the area of interdomain traffic&#xD;
engineering. As an example, Latin America&#xD;
shows the largest de-aggregation factor of IP&#xD;
prefixes among all regional Internet registries,&#xD;
being proportionally the largest contributor to&#xD;
the growth and dynamics of the global BGP&#xD;
routing table. In this article we analyze the&#xD;
peculiarities of LA interdomain routing architecture,&#xD;
and provide up-to-date data about the&#xD;
combined effects of the multihoming and TE&#xD;
practices in the region. We observe that the&#xD;
Internet Research Task Force initiative on the&#xD;
separation of the address space into locators&#xD;
and identifiers can not only alleviate the growth&#xD;
and dynamics of the global routing table, but&#xD;
can also offer appealing TE opportunities for&#xD;
LA. We outline one of the solutions under discussion&#xD;
at the IRTF, the Locator/Identifier&#xD;
Separation Protocol, and examine its potential&#xD;
in terms of interdomain traffic management in&#xD;
the context of LA. The key advantage of LISP&#xD;
is its nondisruptive nature, but the existing proposals&#xD;
for its control plane have some problems&#xD;
that may hinder its possible deployment. In&#xD;
light of this, we introduce a promising control&#xD;
plane for LISP that can solve these issues, and&#xD;
at the same time has the potential to bridge the&#xD;
gap between intradomain and interdomain traffic&#xD;
management.</itunes:summary>
    </item>
  </channel>
</rss>

