The BRADLEY DEPARTMENT of ELECTRICAL and COMPUTER ENGINEERING

Research Areas

Our research in wireless and secure systems through the Wireless@Virginia Tech research center pushes the technology to greater use in society. We continue to make wireless communications and signal processing more secure, robust, and reliable, while enabling new applications, such as connected cars, clinical remote monitoring, and assisted living. Our research delves into spectrum sharing, signal processing, cognitive radio, geolocation, national security, and more.

Current Research

Medical communications

The hospital room of the future will create one of the most challenging wireless environments--with dozens of wireless devices of different standards operating over many different bands. We are tackling this problem on many fronts, including building the Cognitive Medical Wireless Testbed System (COMWITS). COMWITS will provide researchers the ability to test innovative cognitive wireless communications, sensor networks, and biomedical applications under real-world conditions.

Spectrum Sharing

Spectrum is a critical catalyst for economic growth, and commercial mobile broadband needs to support more users and faster data rates. The U.S. government has proposed sharing a number of federal frequency bands opportunistically with commercial systems such as 4G LTE. Spectrum sharing presents challenging issues, including interference, rogue operators, privacy, security, heterogeneous networks, propagation, and enforcement, among others.

Privacy

Security and privacy are critical challenges, particularly when defense systems coexist with non-government systems in the same spectrum bands. In one project, we are developing mechanisms and techniques for an obfuscated geolocation database that can enable the coexistence of primary and secondary users while preserving the operational privacy of the incumbent users.

We have developed an analytical framework for defining and evaluating protection zones, which are spatial separation regions defined around primary users. The legacy notion of interference protection is overly rigid and often results in poor spectrum utilization efficiency. We have proposed a novel framework for prescribing interference protection that includes the concept of Multi-tiered Incumbent Protection Zones (MIPZ). MIPZ can be used to quantitatively analyze a protection zone and to determine the tradeoffs between interference protection and spectrum utilization efficiency.

Legacy Military Systems

How can legacy military systems such as radars, satellite links, or tactical ad hoc networks share bandwidth in time, frequency, geography, and eigenspace? Our work on radars with fully active arrays, known as MIMO radars, and their ability to coexist spatially with cellular systems that use multi-antenna arrays such as 4G LTE, shows that with cognitive resource management in the cellular system, quality of service can be maintained. By characterizing the mutual channel, the radar is able to slightly alter radar pulses prior to transmission such that they fall into the nullspace of the communication systems' channels to significantly reduce interference.

Military Communication

As the military gives up frequency band resources to commercial broadband, it must do more with less. In hostile environments, enemy jammers may seek to deny access to radio spectrum. Our research focuses on cognitive approaches to radio resource management for military communications and radar systems to operate efficiently in these contested and congested environments.

In one project, we are developing a hyper-widespread spectrum (10 GHz wide) transmitter/receiver pair for transmitting a 100 kbps data rate signal that is more robust to jamming than currently available technology. Another effort takes the opposite perspective, investigating machine learning, such as q-learning and multi-armed bandit theory, applied to intelligent jamming.

Rogue Users

Transmissions from rogue secondary users who share spectrum with incumbents present a problem that straddles wireless technology, industrial economics, international standards, and regulatory policy. We have developed and implemented an ex-post spectrum enforcement technique that utilizes small shifts in carrier frequency to authenticate transmitters in spectrum-sharing scenarios. This scheme can be used to uniquely identify rogue transmitters under very harsh conditions (e.g., very low SNR, high multi-path fading, etc.).

Speaker identification

Identifying and verifying speakers from speech collected in a noisy environment gets more challenging as speech-to-noise power decreases. An ECE team reduced the error rate based on the hypothesis that discrimination power resides in the transients into and out of the vowel regions in speech. Since vowels contain relatively more energy, the transient regions can be identified. The team extracted features from both the vowel and transient regions and showed that speaker identity verification in a cocktail noise environment at 0 dB (same power in speech as in noise) led to an improvement in equal error rate.

Testbeds & Frameworks

We are building testbeds and frameworks to combat the threat of low-performing radios and networks developed and deployed with only ad-hoc performance testing and minimal regulatory compliance.

Our cognitive radio network (CORNET) testbed enables research and education on Software-Defined Radio (SDR), cognitive radio, and dynamic spectrum access. It consists of 48 indoor SDR nodes, 14 fixed outdoor nodes, and 6 mobile/portable units (O-CORNET). We are currently integrating LTE-capable nodes (LTE-CORNET). Except for the mobile units, all nodes can be remotely accessed. CORNET is unique in that it offers a wide range of experimental research and educational tools, including an FCC experimental license agreement for several frequency bands.

We are also developing a Cognitive Radio Test System (CRTS) for testing cognitive radios and engines in realistic operational scenarios. At a higher level, the open-source framework will enable evaluation, comparison, and refinement of methodologies.