The National Security Agency's Laboratory for Physical Sciences and Maryland:

A Partnership that Pushes the Limits of Technology

By: Dr. Gary Hughes

Student Sharon Yu, Prof. Julius Goldhar, and Asistant Research Scientist Daniel Mahgerefteh work on an optical fibre experiment at LPS.

Located off Metzerott Road, just beyond parking lot number 4, is the Laboratory for Physical Sciences (LPS). Here, physicists and engineers from the Technology and Systems Directorate of the National Security Agency (NSA) pursue collaborative research with professors, research scientists, and students from the University of Maryland's College Park and Baltimore County campuses. Approximately 20 faculty members and 15 graduate students from the Department of Electrical Engineering (EE) work at LPS. Faculty and students from Maryland's Physics and Materials Engineering Departments also work at the laboratory, as do a number of visiting professors and their students.

Strange Bedfellows

Although the current building is but five years old, the university's association with LPS actually dates back to the 1950's-shortly after President Harry Truman established the NSA. One of the first acts of the fledgling agency was to assemble a Scientific Advisory Board comprised of high-level scientists and engineers. This panel of experts was asked to identify the critical technical challenges faced by the agency. One challenge identified by the group was the need for NSA to guard against unforeseen technological advances in the fields of communications and computers. Subsequently, the Scientific Advisory Board advised the NSA to enact an open partnership with a prestigious university and conduct collaborative, unclassified research across the fields of physical science that influence technological advances in communications and computers. As a result, the Laboratory for Physical Sciences was established, right next to the University of Maryland in College Park.

Mission

NSA has encouraged LPS to aggressively pursue innovative and imaginative ideas that push communication and computer technologies beyond the limits otherwise possible-through more conservative, evolutionary developmental approaches. Dr. Thomas Beahn, the LPS's fourth and current director, readily accepts that such high-payoff research often involves high risk. He also recognizes that promising research often requires extended development before it can be applied successfully. Hence, Dr. Beahn has LPS supporting a healthy mix of studies, from demonstrations of technical feasibility to the development of prototype systems. Dr. Beahn repeatedly stresses that all work performed at the LPS needs to be evaluated against two criteria. The first of these is the project's inherent research value. This is measured, as are most university studies, by the number and quality of publications generated, degrees granted, and patents approved. The second measure is the value of the work to NSA. After supporting 40 full years of uninterrupted collaboration, the NSA has steadfastly been pleased with the results.

Projects

Last year's Annual Technical Review of ongoing research, performed by Maryland faculty and students at LPS, took place on November 18 and 19. Of the 25 projects reviewed, 15 involved significant contributions from EE. These projects included every area of current emphasis at LPS: 1) Integrated Optics, 2) Novel Communications Links, 3) In-Fiber Signal Processing, 4) Optical Devices and Techniques, 5) Materials and Microscopy, and 6) Microelectronics (Devices, Analysis & Packaging). Regardless of how these collaborative projects are grouped, each focuses on pushing beyond the technological boundaries holding back progress in some aspect of the communications or computing fields. Several examples follow.

LPS is pushing the technological boundaries presently throttling communication advances on a number of fronts. These include addressing the entire range of technical challenges to optically inter-connect desktop computers at 2.5 Gb/s. A representative sample of work underway is that of Prof. Mario Dagenais (EE), Prof. Y.J. Chen, from UMBC, and Dr. Dennis Stone from NSA. Together, they are building a wavelength partitioner for routing wavelength division multiplex (WDM) optical signals.

Directing packets between ever higher-speed WDM trunks may one day exceed the processing speed of electronics. This limit, however, may potentially be overcome by processing the signals through all-optical techniques. NSA's Dr. Lance Joneckis somewhat playfully describes the current thinking behind this approach with the following association: optics are to transport as electronics are to processing. Hence, there are a number of studies in all-optical packet header recognition and wavelength conversion at LPS.

Experiments aimed at increasing the speed of optical communication through the atmosphere are also being conducted at LPS. Communication rates through the atmosphere are limited by turbulence that distorts the wavefront of a transmitted signal. LPS is working on several methods of compensating for this distortion. One of these uses an optically addressed spatial light modulator to sense and compensate for the distortion. It is worth noting that the pixellated indium gallium arsenide photosensor used in this modulator was grown on one of the four molecular beam epitaxy machines in LPS's class 10 clean room. Having these machines, and the expertise to use them, enables LPS to grow a large percentage of the many optical and electronic specialty devices it needs to support its leading-edge research.

Increasing data storage density is another area of active research at LPS. Prof. Mel Gomez (EE) projects that conventional thin film recording will top out at 20 Gb/square inch. Although that's a 20-fold improvement over today's storage densities, he's beginning work in cooperation with Stanford University to assess the viability of a new paradigm that uses patterned media to obtain ultra high storage densities.

Lasting Value

NSA is pleased with the high quality and potential applicability of all of the research being performed at LPS. The fact that this has been the case for 40 years offers strong testimony that both NSA and the university have benefited from this partnership. The results remain impressive, even when the focus is narrowed to just the collaborative research performed at the LPS by members of EE and NSA's scientists and engineers. This collaboration has already resulted in 20 technical papers co-authored this year by the two groups.

For more information on LPS, please contact the Chairman of the Department of Electrical Engineering, Prof. Nariman Farvardin, at farvar@eng.umd.edu, or Dr. Gary Hughes, the Deputy Director of the Laboratory for Physical Sciences, at garyd@lps.umd.edu.