Catálogo de publicaciones - libros

Optical Burst Switching (OBS) has been proposed in the late 1990s as a novel photonic network architecture directed towards efficient transport of IP traffic. OBS aims at cost-efficient and dynamic provisioning of sub-wavelength granularity by optimally combining electronics and optics. In order to reduce the number of switching decisions in OBS core nodes, traffic is aggregated and assembled to bursts by the Burst Assembly Unit in an OBS ingress edge node. This Burst Assembly Unit is responsible for buffering incoming packets in queues and sending them as bursts as soon as a minimum burst length is reached and/or a timer expires. Typically, dozens of different queues must be able to handle high volumes of traffic.

This paper presents the design and implementation of a Burst Assembly Unit for a Network Processor. In an evaluation of the realized implementation we point out the ability to handle traffic at line speed while having fine grained timers for all queues.

This paper analyzes the properties of packet interarrival time (PIT) distribution functions of network segments including bottlenecks. In order to show the correlation between bottleneck behavior and packet interarrival time distribution, the alteration of probability distribution function (PDF) is observed through simulations including tighter and tighter bottleneck connections. The process of network bottleneck detection by passive monitoring requires effective metrics for distinguishing seriously congested links from normal or underutilized connections. The paper evaluates the third and fourth central moments (skewness and kurtosis, respectively) of PIT distribution as possible metrics for bottleneck detection. Simulation results as well as real measurement data analysis showed that PIT kurtosis can be a powerful measure of bottleneck behavior.

At high transmission speeds, complexity of implementation for fair queuing disciplines can impose a bottleneck to the overall system performance. Available scheduling algorithms set a fixed trade off between fairness and complexity, fairer systems involving more operations per packet and vice versa. In this paper first a new fair queuing scheme is proposed, with almost the same fairness and complexity propierties achieved so far by most used algorithms. Later on a tunable parameter is introduced, which allows the modification of the above mentioned trade off between fairness and complexity depending on working conditions and thus enlarging the field of application of the scheduler.

We propose a proper length function for an existing QoS routing algorithm (SAMCRA) that attempts to optimize network utilization while still offering QoS guarantees. This paper presents a comparison between several proposed algorithms via simulation studies. The simulations show that SAMCRA with a proper length performs similarly or even better than the best among the other algorithms and it has a fast running time.

Nowadays, many organizations need to be multihomed in order to achieve fault tolerant Internet access. Unfortunately, the hierarchical nature of IPv6 addressing architecture poses some threats on multihoming. The IETF is designing a solution based on the discussion of several approaches to solve the problem. Ingress filters are part of the problem, so ingress filtering compatibility mechanisms are needed. This paper discusses the host-centric multihoming approach and describes an implementation of an ingress filtering compatibility mechanism based on automatic tunnels and anycast addresses. The implementation has proven to work properly, being easily developed and deployed.

The interactions between the IGP and BGP routing protocols which are running inside an ISP’s network are sometimes hard to understand. The problem becomes particularly complex when there are dozens of routers/links and several thousands of destination prefixes. In this paper, we present a publicly available routing solver to evaluate routing what-if scenarios. The solver is able to model the complete network of an ISP and given the external routes learned by this ISP, to compute the paths towards all the destination prefixes. We demonstrate the use of our routing solver, C-BGP, by showing the results of an analysis of the link/router failure sensitivity in a transit network. Based on the analysis’ results, we can pinpoint links/routers whose failure has an important impact on the selection of BGP routes. The deployment of protection techniques that are used for optical links, SONET-SDH and MPLS should be considered for these links/routers.

Currently, a straightforward way to design BGP-based Traffic Engineering (TE) tools for stub Autonomous Systems (AS) is to rely on selfish routing mechanisms. Although TE tools can find an optimal solution, this optimum represents only a local optimum for outbound traffic. Indeed, this is one of the main limitations of the selfish routing approach. This approach makes the TE tools unaware of the effects of their route choices on transit AS throughout the chosen paths, due to uncoordinated routing decisions, and congestion can occur on distant intra- or inter-domain links. Thus, cooperation among AS is the key to avoid the performance degradation and routing instability caused by the selfish routing approach and it would be fundamental for the future QoS-aware Internet. With these objectives in mind this paper presents and discusses a framework for coordinated Inter-domain QoS Routing (QoSR) decisions among stub and downstream AS taking into account multiple traffic QoS constraints and routing preferences. The paper includes a description of the main mechanisms and algorithms that integrate the framework, and finally a discussion of the implementation issues.

The new IPv6 and the current IPv4 will coexist for many years. A wide range of techniques have been designed to make the coexistence possible and to provide an easy network transition. An organization involved in the IPv6 transition should address not only network issues but also final user applications. The management of the different services must also be revised. This is specially relevant in academic institutions where educational, production and research networks live together in the same environment. This document is intended to give the reader a more comprehensive and accurate picture about the IPv6 transition procedures that may be accomplished by the different University members to introduce IPv6 smoothly and successfully.

Recently, a first step towards a highly distributed IP-level topology discovery tool has been made with the introduction of the Doubletree algorithm. Doubletree is an efficient cooperative algorithm that allows the discovery of a large portion of nodes and links in the network while strongly reducing probing redundancy on nodes and destinations as well as the amount of probes sent. In this paper, we propose to reduce more strongly the load on destinations and, more essentially, the communication cost required for the cooperation by introducing a probing stopping rule based on CIDR address prefixes.

As the demand of ubiquitous Internet access and the current trend of all-IP communications keep growing, the necessity of a protocol that provides mobility management increases. The IETF has specified protocols to provide mobility support to individual nodes and networks. The Network Mobility (NEMO) Basic Support protocol is designed for providing mobility at IP level to complete networks, allowing a Mobile Network to change its point of attachment to the Internet, while maintaining ongoing sessions of the nodes of the network. All the mobility management is done by the mobile router whilst the nodes of the network are not even aware of the mobility.

The main aim of this article is evaluating the performance of the NEMO Basic Support protocol by using our implementation. We also discuss the design of an implementation of the NEMO Basic Support protocol.