Current Graduate Student Opportunities

The NWC offers both paid and for credit student research positions. We strongly prefer candidates interested in making a multiyear commitment to the center. In general, students start with an unpaid semester-long reading course.

Graduate Research Assistantship Opportunity (posted 3/29/2012)

A graduate research assistantship in applied mathematics/signal processing is available in one of three possible areas:

Cognitive Communications Electronic Warfare (CogEW)
Cognitive Radio (CR) is an emerging radio technology where radios autonomously adapt their transmissions to achieve an objective such as avoiding interference with other RF devices and jammers or avoiding detection.  Work includes developing sensing and machine learning algorithms to characterize and learn the adaptive behaviors of cognitive radios in real-time.  A non-parametric balanced multi-scale Bayesian inference (BMBI) engine has been shown in simulations to learn the dynamic behavior of a cognitive radio network.  In addition, a blind source localization algorithm was developed to geolocate emitters by probabilistically combining angle-of-arrival and signal strength information.  Future work will focus on improving the behavioral learning algorithms and leveraging that information to autonomously create and optimize electronic attack strategies against cognitive radios in real-time. 

Emitter Recognition for Behavioral Analysis (ERBAN)
Analyzing patterns of behavior in wireless communications networks. The first research objective, emitter recognition, involves supervised pattern classification of the RF transmission of wireless communications devices to identify the model type. A major technical challenge involves dimensionality reduction of the RF features so that emitter recognition can be performed rapidly, and nonlinear system identification associated with analog components of these devices will be incorporated as well.  The next research topic involves analyzing wireless communications patterns for threat behavior using machine learning techniques to predict whether these communications patterns can be associated with threats.  This is analogous to the problem of detecting anomalous patterns of communications in a wired network.  Techniques such as network change detection, graph spectra, and parametric Bayesian machine learning will be applied.  

Investigating novel forms of compressed sensing for communications and nonlinear systems identification of analog transmitters and receivers.