News and Notes for Our Industrial Partners
June, 2008

 

RESEARCH

ICTAS (INSTITUTE OF CRITICAL TECHNOLOGY AND APPLIED SCIENCE) SUPPORTS COGNITIVE RADIO NETWORK TESTBED (VT CORNET)

Wireless@VT has embarked on the ambitious goal of building a unique heterogeneous wireless communication network testbed. With ICTAS support, we are building a part of the network consisting of cognitive radio nodes. In time, this network will be augmented and extended to include other nodes as well. This will be the first network of its kind at any university in the country. The testbed will enable researchers from VT and outside to implement and test their algorithms, protocols, applications and hardware technologies within a realistic environment. This will position Wireless@VT to attract large government and industrial funding in the area of adaptive and cognitive networks for many applications.

The testbed will be deployed in the new ICTAS-I building. Upon completion, the testbed will consist of 48 cognitive radio nodes. Each node will consist of a processor board, a Universal Software Radio Peripheral (USRP), and a RF frontend. The RF frontend will be based on the new Motorola RF transceiver chip that has a bandwidth of 100MHz to 4GHz. These nodes will be installed in the ceilings of the 4-story building, 12 per floor. A control room will be used to monitor the radio signals and control the nodes. The control room should be large enough to accommodate about 10-12 graduate students. Although the nodes will be wireless, they will also be interconnected via Ethernet. This is for the sole purpose of remotely updating the software to each of the nodes.

Many different applications and experiments will be deployed on this testbed. These include dynamic spectrum access and interoperability for public safety.

Currently we have a dynamic spectrum access (DSA) application running on the 5 node testbed that is built using the CIREN system, the MPRG team’s SDR Forum Radio Challenge entry. As each radio node comes online, they rendezvous on a channel in the FRS band, the same band as common 2-way radios. If they detect an interferer, in our case if someone talks on a 2-way radio in the same channel, they will vacate the channel and rendezvous on another channel that is unused. Currently, the nodes can send text and voice data to each other.

We are also improving the CIREN system and adding mesh network functionality and video streaming applications. A critical part of the success of this project is our ability to raise funds to support the testbed. We estimate we will need approximately $1.5 million dollars to achieve all of our objectives. Below is an overview of the Testbed v1.0 Setup.

IHCS: PROACTIVE CROSS-LAYER ADVERSARY LOCALIZATION FOR HOSTILE OR HARSH WIRELESS ENVIRONMENTS

The following is the abstract of a recently received NSF award. The Principal Investigator for this award is Dr. Yaling Yang. The Co-PIs are Drs. Jung-Min Park and Dr. Michael Buehrer. This 3 year project worth approximately $300,000 will begin this summer. This project will build on Dr. Buehrer’s previously funded NSF project on position location, and Dr. Park’s NSF CAREER award.

The objective of this research is to develop a localization system capable of localizing an adversary that is actively trying to disguise its location in a wireless network by distorting its signal features. The approach is a proactive, cross-layer localization design that incorporates attack traceback, cross-layer traffic manipulation, and physical layer position estimation. The attack traceback aspect focuses on narrowing down an adversary’s location to the coverage area of a couple of access points. The traffic manipulation aspect will develop trapping techniques to force or lure the adversary to exhibit their true location-related signal features. Leveraging these true location-related signal features, the physical layer position estimation aspect will develop proactive and robust localization techniques to accurately position the adversary.

The proposed project will help to establish accountability in wireless networks and will result in the development of key attack countermeasures. It is the first to address many technical challenges in localization and traceback. This project can also enhance the security of systems where location information is used to restrict access to critical resources. Furthermore, the proposed research results can be used to improve the accuracy of localization systems in harsh communication environments that severely distort the characteristics of emitted signals from legitimate users.

The proposed research will foster the integration of research and education by fortifying the existing curriculum with the project’s research results. The outreach component of the project will disseminate research results and pedagogical materials via education and industry outreach programs.

QUALCOMM PROJECT FUNDED

Dr. Buehrer is beginning a new project, funded by Qualcomm, titled “A Fundamental Analysis of Spectrum Sharing.” Spectrum sharing has become a topic of great interest in the past few years due to (a) the apparent lack of spectrum for new technologies and wireless applications, (b) the documented underutilization of currently allocated spectrum and (c) the emergence of cognitive radio research. However, while many agree that the sharing of spectrum may result ultimately in improved capacity, it is unclear how this sharing should be accomplished. There are a variety of possible approaches ranging from techniques that put little additional burden on the standard or application (e.g., non-coordinated frequency use¹) to more advanced techniques involving sophisticated cooperation or negotiation, which could place substantial burden on the application standard. This project proposes to develop a framework to examine the fundamental limits (i.e., performance bounds) of four basic approaches for spectrum sharing. Specifically, examining the performance limits of:

1. No cooperation or interference mitigation whatsoever - non-coordinated frequency use
2. Interference excision approaches - non-coordinated frequency use with interference mitigation at the receiver
3. Rule-based access - Purely passive, sensing-based approaches based on preset policies which require no cooperation
4. Active cooperation - Approaches which require interoperability between devices

The goals of the work are to determine the fundamental limits of cognition in the context of spectrum sharing. Specifically, what is the benefit of sensing/avoidance type approaches which require rules (and potentially cognition) but no interoperability as compared to traditional interference mitigation? Secondly, what is the (presumably additional) benefit of active cooperation which requires interoperability between standards? Is this interoperability fundamentally beneficial? If so, under what conditions is it beneficial?

The benefits of these various approaches to spectrum sharing will be examined in terms of fundamental limits. Although there has been some research into the capacity of spectrum sharing, original work will be provided in two respects. First, common framework will be created which will allow a comparison of the limits of the different approaches such that the benefits of cognitive approaches over less sophisticated approaches can be explicitly studied. This is of particular interest to those involved in standardization efforts who must make decisions on the ultimate benefit of incorporating sophisticated cooperation techniques into various standards.

Secondly, the focus will not be on Shannon capacity from an information theoretic point of view. The capacity of the interference channel (the model for cooperation between distributed nodes) is a difficult theoretical problem and doesn’t necessarily provide direct insight into the benefits of different approaches. Instead the focus is on approaches that examine individual link capacities or the sum of these link capacities.

¹By non-coordinated frequency use we mean that there is no attempt by the system to avoid or mitigate potential interference.