Catálogo de publicaciones - libros

This paper focuses on the design and implementation of a low-complexity dialogue mechanism between the management and transport planes of an all-optical network. This dialogue aims at exchanging relevant information from monitoring the performance and degradation of optical signals with minimal disturbance to the optical services and minimum knowledge of the transport history of data, with a view to ensure service quality. Complexity of the dialogue is measured in terms of response delays in the event of failures.

The GMPLS standardization is paving the way for new configurable Traffic Engineering (TE) policies and new survivability schemes for transport networks. In this context, a centralized Path Computation System (PCS) has been implemented, suited for transport networks with a GMPLS control plane. After a brief description of the requirements for a PCS in a GMPLS network, some design issues for the proposed implementation are drawn, with particular emphasis on the centralized approach and on the strategies for achieving the connection survivability. Some results of an intensive testing campaign are shown for the validation of the design choices.

The operators are today facing the migration of their transport infrastructures toward intelligent optical layer by introducing in the core optical switches based on GMPLS/ASON protocol. The paper describes the advantages can be achieved by using the next generation Optical Platforms by comparing the ring and mesh network.

Network restorations in the distributed and centralised control plane are also compared by showing some lab test results. Finally are outlined the current limitations and the issues still open to be solved by the industry and the standard bodies.

A new rerouting scheme and spare capacity planning for optical link failures is demonstrated employing Bluetooth monitoring. Optical networks require adequate fault monitoring in order to accurately identify and locate network failures. Detailed physical layer information as well as link surveillance is carried to OLT for in-situ monitoring of QoS.

Although optical transmission technology for high-speed broadband networks is being studied actively, the conventional packet-based switching technology cannot assure high-quality communication for each connection. We therefore propose a new computing environment, called λ computing environment, that provides virtual channels utilizing optical wavelength paths connecting computing nodes. These channels provide the reliable high-speed paths needed for recently deployed distributed applications including SAN and Grid computing. In this paper, we propose and evaluate a method for accessing the virtual ring network for realizing a shared memory in a distributed fashion on WDM-based photonic networks.

With the automation in optical networks based on the introduction of a control plane (GMPLS, ASON) it will be possible to adapt the transport network dynamically to the traffic requirements of packet based client layers. This allows reducing leased line costs and improving the throughput. In some cases the network throughput limitation comes from the limited switching capacity of transit routers. This is especially the case when Multiservice-network elements are used, which usually are implemented as an extension of SDH network elements with packet switching capabilities. Hereby the packet switching matrix usually only covers a fraction of the whole interface capacity of the node, since the main switching is expected to be carried out in the TDM (SDH) domain. An intelligent Multilayer-Routing-Strategy can relief the packet layer by circumventing packet switching with transport shortcuts. In this paper we present such a new Multilayer-Routing-Strategy, that allows to route demand with a distributed routing mechanism. This is complemented by a centralized optimization instance, that optimizes the overall network status. Simulative investigations allow a first quantification of the advantage of this approach.

The GMPLS standardization is paving the way for the implementation of new configurable traffic engineering (TE) policies for transport networks. This paper takes aim at evaluating the effects of using bandwidth-dependent TE metrics in a centralized Path Computation System (PCS), suited for handling the routing requests in an operational transport network with a GMPLS control plane. The results of an intensive testing campaign show an evident improvement in the utilization of network resources when such TE metrics are enabled, whatever survivability requirement is imposed on the LSP (e.g. classical 1+1 protection, pre-planned or On-the-Fly restoration, etc.). Moreover, a simple policy function is suggested as a good trade-off between the achievable performance and the computing load on CPU.

In wavelength-division multiplexing (WDM) optical networks, the bandwidth request of a traffic stream can be much lower than the capacity of a lightpath. Efficiently grooming low-speed connections onto high-capacity lightpaths will improve the network throughput and reduce the network cost. In this paper, we propose and evaluate a new concept of traffic aggregation in mesh networks that aims to eliminate both the bandwidth underutilization and scalability problems existing in all-optical wavelength routed networks. Our objective is to improve the network throughput while preserving the benefits of all-optical wavelength routed networks.

Within GMPLS framework, the signaling protocol Resource reSerVation Protocol with Traffic Engineering extensions (RSVP-TE) is extended to support the requirements of an Automated Switched Optical Network architecture. This paper presents the extensions of the end-to-end connection services in an overlay network built on two control planes. RSVP-TE protocol extensions are first described between an IP/MPLS router and a SDH/GMPLS core optical cross-connect, defining GMPLS-UNI. Dimensioning of three scenarios proving the benefits of GMPLS-UNI is discussed.

Multiprotocol Label Switching (MPLS) was originally designed to provide higher packet forwarding rates in network equipment. Nevertheless, it was soon realized that it could also provide other advanced features, such as Traffic Engineering and Virtual Private Networks capabilities. The key feature of MPLS is a strict separation between control and forwarding operations, which reflects on the software and hardware architecture of the routers. The paper presents the results of an experimental study aimed at evaluating the performance of a Label Switching Router (LSR). In particular, the behaviour of the LSR control and forwarding planes has been analyzed in different working conditions as it concerns both the processing and computational effort due to the control plane and the data traffic to be forwarded.