NEXT-GENERATION NETWORKING GROUP

In more recent years, several networking developments have emerged to dramatically alter the existing telecommunication infrastructure. The most important development, of course, is the rise of optical networking, first with wavelength division multiplexing (WDM) and more recently with optical multiplexers, and optical cross-connects. Optical networking heralds a new economic era, when bandwidth will be abundant and inexpensive. It means real-time establishment of optical paths, leading to new network architectures, new services, and ultimately new business models. Finally, optical networking forces us to go back and reconsider some of our most basic assumptions about how to build networks.

OVERVIEW OF RESEARCH AREAS

Fiber Optic Communication Systems, Networking & Architecture; IP/ATM/SONET-based DWDM and TDM Multiple-Access Broad-band Networks; High performance IP/MPLS routers, Next Generation Networking paradigm, Traffic Engineering/Provisioning &  Protection/ Restoring in a data-centric DWDM-based optical networks; Optical Amplifiers and components; CATV distribution over fiber-based local access ATM networks. 

Current Specific Ongoing Research Activities:

• How to scale metro Ethernet networks into a global multi-services infrastructure. Can we implement a truly seamless Ethernet infrastructure stretching from enterprise LAN to Metro to Global? This approach is novel and offers advantages over existing Layer-2 and MPLS solutions in that it divorces the Ethernet from legacy transport mechanisms like SONET/SDH and other layer-2 protocols. 

• Conducting research activities on issues related to the pros and cons of a router-centric architecture deployed on a thin optical layer versus a hybrid router/OXC-centric architecture deployed on a rich and intelligent optical transport layer. These studies included a detailed economical and technical analysis on the merits of a pure packet-switched network versus a hybrid, which packet switches only at the access point and circuit switches through the network.

• Currently conducting extensive work on modeling the Traffic Engineering problem in a Hybrid IP-Centric DWDM-Based Optical data Networking for both unicast and multicast traffic.

• Developing innovative fully distributed global information-based integrated routing and signaling framework for real-time provisioning of diverse traffic granularity (on a per-call basis including both full-lambda and sub-lambda traffic flows) entirely on the optical layer’s terms. The proposed integrated routing and signaling protocols go beyond those being developed within GMPLS by the IETF and OIF. Provisioning of diverse traffic granularity entirely on the optical layer’s terms, as this work will show, introduces numerous new challenges including additional control plane complexities that need to be addressed. To implement the proposed vision of an agile optical networking layer capable of supporting integrated routing and signaling algorithms to dynamically provision diverse traffic granularity, the following two salient features must be implemented: 1) Most of the networking functionalities and intelligence (including switching, protection, traffic engineering, and selective restoration) must be migrated down to the optical layer , and 2) The optical layer must also own and manage both the physical connectivity (optical resources) and logical connectivity (IP resources). We argue, as this work will show, that moving the networking functionality and intelligence down to the optical layer (favoring the intelligence of optical switches over routers), is more compelling in terms of simplicity, scalability, overall cost savings, and the feasibility for near-term deployment.

•  Implementation of a Secure Network based on Integrated QoS and Policy Base Network Management

Our society has come to rely heavily on the Internet. We take Internet based services for granted, though we are often reminded of the fragile nature of the Internet when a quasi-technical “script kiddy” takes down significant portions of the network with simple programs available on hacker web sites.  How is it that billions of dollars have been invested in Internet based services, yet they are vulnerable to such security breeches by unsophisticated hackers? Additionally, what level of disruption could well-funded networking professionals unleash if they were properly motivated to do so.

The well-known truth is that the public Internet has no defenses against many classes of attacks. On the public Internet all traffic is treated equally. Anonymous access to all destinations with all available bandwidth leaves a network open attacks that compromise sensitive information as well as Denial of Service (DoS) attacks that defraud users and operators of valuable resources.

Networks providing QoS services must explicitly support security aspects including Authentication, Access Control, Accounting and Message Privacy/Integrity. It’s our vision that the implementation of a true QoS-based network implies simultaneously providing conventional QoS mechanisms and comprehensive security infrastructure. Furthermore their integration will stretch the limits of both technologies. 

We are currently conducting work to integrate QoS based mechanisms and Security infrastructure. We will construct a Policy Based QoS enabled network with the core based on our Multi Protocol Label Switching (MPLS) enabled test-bed to demonstrate integration methodology and techniques. We will address concerns surrounding provisioning complexity and scalability with a novel distributed approach.