An optimized neutron activation analysis framework for arms-control treaty verification

Read full text
Authors: Xiao-Suo He, Hong-Qiang You, Li-Teng Xing
Cited by: 0
References: 14

Abstract

Advancing the discrimination capability and robustness of verification methods within a zero-knowledge framework remains a central challenge in contemporary nuclear arms-control. Building upon our previously proposed hash-algorithm-based neutron activation analysis system , we present an integrated optimization of both the verification apparatus and the associated algorithm. On the hardware side, we develop a neutron moderator that delivers a higher-quality and operationally practical thermal-neutron beam. By independently characterizing the spatial distributions of neutrons and prompt gamma rays, we determine the system’s optimal geometric configuration, thereby improving measurement fidelity at the source. On the algorithmic side, we propose a high-resolution difference-hash algorithm that incorporates a key-based encryption mechanism, substantially strengthening encrypted identification. Using previously established sample sets, we conduct simulated treaty-verification to evaluate system performance. The optimized apparatus and the new algorithm yield a marked gain in discrimination accuracy: under 1% noise, the system achieves an area under the receiver operating characteristic curve above 0.977—significantly higher than the 0.925 achieved by our earlier design. Moreover, the system preserves strong resistance to statistical analysis, brute-force reconstruction, and key-based security threats. Overall, the optimized system effectively prevents unauthorized access or tampering while substantially enhancing identification accuracy, establishing a robust pathway that couples physical protection with algorithmic encryption.

Fields of Study & Subjects

Nuclear Physics and Applications Radiation Detection and Scintillator Technologies Nuclear Issues and Defense

Publication details

Journal / source
Nuclear Engineering and Technology
Volume
58
Issue
6
Pages
104226–104226
Publication date
3 March 2026
Language
EN