Design and deployment of a GMPLS control plane in IP optical networks.
Item
-
Title
-
Design and deployment of a GMPLS control plane in IP optical networks.
-
Identifier
-
AAI3283629
-
identifier
-
3283629
-
Creator
-
Li, Zhaoming.
-
Contributor
-
Adviser: Ibrahim Habib
-
Date
-
2007
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Engineering, Electronics and Electrical
-
Abstract
-
The primary goal of this thesis is to design, implement and deploy a Generalized Multiprotocol Label Switching (GMPLS) software engine in the experimental optical network called Circuit-switched High-speed End-to-End Transport Architecture (CHEETAH) network. This GMPLS software engine, called CHEETAH Virtual Label Switch Router (CVLSR), equips non-GMPLS devices such as Ethernet switches, routers, and other cross-connects with the GMPLS capabilities such that they can participate in the dynamic provisioning of end-to-end bandwidth-guaranteed connections. An important premise of dynamic provisioning is to allow the dynamic sharing of the large bandwidth pipes (e.g., 10 Gbps and above) of backbone networks by many users, thus cutting down the expensive costs of communications at these capacities. The CVLSR is also equipped with a novel dynamic network clustering capability to enable a broad range of applications such as remote visualization, grid computing and e-Learning. The CVLSR has been deployed and tested though experimental CHEETAH optical network across nation-wide high performance networks such as Hybrid Optical/Packet Infrastructure (HOPI) and Dynamic Resource Allocation via GMPLS Optical Networks (DRAGON) networks. We believe that this and similar efforts will result in a significant progress towards enabling the vision of dynamic provisioning of the end-to-end connections in the multi-region connection-oriented networks.;Next, a link bundling mechanism is presented to improve the efficient utilization of the optical circuits by sharing such circuits among multiple users or multiple applications. The concept of link bundling for dynamic allocation of the optical circuit has been demonstrated through both experimentations and simulations. We demonstrate that link bundling can handle the TCP/IP traffic without TCP connection abort when the buffer size of router interface is appropriately chosen. The packet loss introduced by link unbundling can be recovered by TCP retransmissions.;Finally, a linear programming based scheduling model is proposed to make optimized reservations for the book-ahead services subject to quality of service constraints such as minimizing connection blocking rate while maximizing network utilizations. Our algorithm is simulated based on a single routing area and it could be used for intra-area as well as inter-area and inter-domain scheduling.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
