Warhead Verification with Sensitive Design Data Protection

A major challenge in verifying that a nuclear warhead actually is a warhead lies in protecting critical design information. Confirming certain technical attributes will likely require overcoming the hurdle of protecting sensitive design data and national secrets.

An approach for conducting verification measurements while simultaneously protecting sensitive design information is proposed by Sandia National Laboratories physicists. The CONfirmation using a Fast-neutron Imaging Detector with Anti-image Null-positive Time Encoding (CONFIDANTE) is designed to protect sensitive information for the warhead owner, or host, while accommodating the monitor, who may be seeking to verify that sensitive information to confirm the inspected item is a warhead.Sandia National Laboratories researchers (left to right) Peter Marleau, Patricia Schuster, and Rebecca Krentz-Wee developed a new method for verifying warhead attributes. (Photo by Dino Vournas)

“CONFIDANTE is an implementation of a zero-knowledge proof (ZKP) as a way to demonstrate the validity of a claim while providing no further information beyond the claim itself,” explains physicist Peter Marleau. “Unlike other ZKP confirmation methods, which rely on a measuring instrument that has been pre-loaded with sensitive information, CONFIDANTE allows the monitoring party to conduct the measurement in real time without accessing sensitive design data.”

In mathematical cryptography, ZKP is accomplished by challenging a host to solve a problem that is only possible if the host possesses the information being authenticated. After repeated challenges, the host can prove it possesses that information without revealing any details about the information itself. A ZKP object-comparison system was developed three years ago by the U.S. Department of Energy’s Princeton Plasma Physics Laboratory and Princeton University to potentially support warhead confirmation while protecting sensitive design data.

Using this system, confirmation that an alleged warhead has the characteristics of a warhead is demonstrated through neutron transmission and emission counts measured by an array of radiation detectors. To protect sensitive design data during the measurement process, the Princeton method prepares the radiation detectors with a template rather than directly comparing in real time the images of a warhead being verified with a trusted warhead.

The template is the complement of the measurement expected from a real warhead. If the two match, they cancel each other out leaving only statistical noise, yielding no further information. The “templates” are effectively destroyed by the measurement, so the monitor does not have the opportunity to maintain the data to which a measurement is compared.

However, the template, the process of pre-loading it, and the detector itself are off-limits to the monitor, undermining measurement authenticity.

Sandia researchers sought to resolve this problem by use of time-encoded imaging (TEI), a new approach for indirect detection and localization of special nuclear materials, which relies on encoding directional information in the time-dependent modulation of fast neutron detection rates. Sandia developed TEI to overcome the precise calibration and high cost of typical detection, which uses arrays of detectors.

TEI uses a single detector within a cylindrical coded mask. As the mask rotates, radiation from the object is modulated by a pattern of apertures and mask elements on the cylinder. Using TEI, a single detector can do the work of multiple detectors in creating an entire 2D image of the object.

“We realized that if we designed the mask such that the pattern on one half of the cylinder is the inverse of the other half, an object on one side of the system will project the inverse image of an object on the opposite side of the system at all times if and only if the two objects are identical. The image and anti-image will effectively cancel each other out and the detector will show a constant unmodulated rate,” said Marleau. “And we can do it without ever recording potentially sensitive information.”

Because no information other than statistical noise is stored or recorded in the detector — unlike a template approach — the host party in theory can certify that no sensitive information is at risk. The monitor then can have full access to the data in real time, potentially even conducting the measurement themselves.

Proof-of-concept tests conducted at Lawrence Livermore National Laboratory using identical plutonium dioxide hemispheres demonstrated feasibility. Sandia researchers now plan to improve CONFIDANTE with a more compact gamma ray version of the imager.