NRT: Heterogeneous Wireless Access Network Test Bed (Rev.)

Stevens Institute Of Technology, Hoboken NJ

Investigators

Abstract

Over the next 10-15 years, the core and access networks will evolve in different directions. The core will evolve toward an all-optical system, while the access network will continue to experience a proliferation of technologies and network types, tuned to the economics of specific deployments and the requirements of users and applications. In particular, wireless technologies will play an increasing role not only for mobile applications but as a cost-competitive substitute for fixed wire access networks. The researchers propose to develop a test bed focused on the access network, containing both wired and wireless components. It will act as a shared-facility test bed for research on connectivity, quality of service, security, reliability/availability, and network and traffic management for heterogeneous access networks. The test bed will also serve as a proof-of-concept test bed for wireless networks that exploit advanced flexible software defined radio devices, steerable antennas and upper millimeter wave point-to-point links to improve network performance under resource constraints. There will be two main academic centers of activity (Stevens Institute of Technology, and University of Colorado), supported by activities at one other institution, (University of Utah). There will also be substantial participation from non-academic partners, including technology companies (Vanu, Inc., Lightpointe, and BBN). Use of the test bed by others will be solicited. Intellectual Merit: The access network is vitally important for service to the end user, yet frequently neglected in studies that focus on the high-speed core or the wireless link in isolation. A test bed in this area will enable collaborative research on ways to exploit and manage multiple networks simultaneously for overall optimization, maintain connectivity as mobile devices roam, exploit one network to support another, and improve security. In the past networking research has focused on layers 3 and higher, while wireless research has primarily looked at layers 1 and 2. A critical aspect of this test bed is that it will support cross-layer integrated experiments all the way from layer 1 to layer 5, leading to future networks that can exploit physical layer flexibility, application awareness, and advanced routing for overall better performance and service. The hardware and software deployed in the test bed will be designed for easy modification and repeatable experiments at all layers. The team building the facility brings deep experience in these modifiable systems: Dirk Grunwald (Colorado) with the Click modular router, John Chapin (Vanu, Inc.) with Vanu Software Radio, Jay Lepreau (Utah) with the Emulab software platform, Craig Partridge (BBN) with protocol software, and Igor Alexandrov (Stevens) with millimeter wave bands radio. The effort is led by Paul Kolodzy (Stevens), formerly of the FCC and DARPA, who brings a wide overview of access network issues. Broader Impact: The proposed test bed will provide a major research facility, bringing together a community from all over the country to focus on the interrelated problems facing access networks. In addition, novel educational approaches will be pursued where students will have access to multiple advisors with key complementary expertise, and key industrial partners. A series of high profile workshops will facilitate research agenda setting and community formation. Tangible products will include widely disseminated research papers and a series of white papers aimed at non-specialists. Finally a series of demonstrations of the test bed will be conducted for the research and lay communities. Progress on the research problems of access networks will reduce the cost and increase the service level provided to all users of the Internet, especially those in currently underserved communities. Additionally, devices and techniques that better handle heterogeneous wireless networks will improve the interoperability problems facing public safety and homeland security.