National Security
Researchers at PNNL are developing innovative modeling and simulation technologies and techniques to help secure the homeland and protect the American people. The National Security application area includes modeling and simulation in behavior, networks, radiation detection and materials discovery, and radiation transport.
Behavioral Modeling
The ability to understand and predict human behavior may help prevent terrorist acts from occurring and enhance national security. PNNL is developing innovative behavioral modeling and simulation tools to characterize or predict various aspects of human individual and social actions and interactions. For example, we use behavioral modeling tools to examine interactions between members of terrorist groups for impending actions or behavioral cues from suicide bombers. We also use behavioral modeling for information analysis activities including cognitive models of human information processing and decision making motivate the design of decision aids, visualizations, and other performance support systems to mitigate human limitations. In cyber security, behavioral modeling approaches are used to support the recognition of malicious activity or intent gleaned from data monitored in computer networks. In the areas of energy, environmental, transportation, or computing infrastructures, modeling of group behaviors may give clues to possible impacts on critical infrastructures of concern to the national security.
Critical Infrastructure
Modeling is a valuable tool for the inspection of critical infrastructures such as nuclear reactors, railway components, natural gas pipelines, and bridges. Models are generated to determine the viability of using a technique, such as acoustics, for inspection. A key part of the process is to determine the degradation or compromise of a critical component and its acoustic signature. Modeling helps to guide the experimental procedure needed to perform an optimal measurement procedure. Laboratory work involving test parts can provide feedback to the models and determine whether a particular technique (such as acoustics) or a combination of techniques (such as acoustics combined with eddy-current analysis) will result in a viable method for infrastructure inspection.
Network Modeling
Computing networks are essential for the nation's economic stability and thus are a national security concern. Cyber security aims at protecting this infrastructure both proactively and reactively. Network modeling is concerned with modeling the interactions between computational devices viewed by topology, functional relationship, connection media, or scale. PNNL's network modeling includes the study of
- network communications, algorithm, and data transfer optimization
- security properties of (both simple and complex) network architectures
- failure/survivability modes of networks
- efficient protection of networked systems
- security, resource, and performance tradeoffs of network communications protocols in sensor networks, enterprise communications networks, and infrastructures
- efficient mathematical models, such as scale-free network models
- efficient protection of networked systems.
Simulation of computer networks provides insight into approaches to protecting ubiquitous networks from cyber attacks
Radiation Detector Material Discovery
In the simulation area, PNNL is developing the capability to simulate the physics of signal generation in radiation detection media. This capability will lead to the ability to predict the performance of radiation detection materials and thereby accelerate the discovery and development of next-generation materials. PNNL's simulation tools cover primarily the energy cascade leading from an incident gamma ray or neutron to the production of "thermalized" electronic (semiconductors) or optical (scintillators) signal.
Radiation Transport for Detection
Radiation detection, a PNNL strength, has been applied to non-proliferation activities, among others. Modeling of scenarios from source to detection is a significant capability for application to non-proliferation and homeland security concerns. Coupled to the physics-based modeling, accurate statistical characterization of benign and nuisance radiation sources is critical to evaluating the performance of existing and theoretical detection systems through this source-to-detection modeling process.
PNNL has extensive experience in radiation transport modeling, including the use of MCNP, GEANT and deterministic transport codes such as Attila. PNNL is taking the lead in combining Monte Carlo methods with deterministic transport codes to create rapid computations for large, complex geometries. Parallel versions are used, and work is progressing on applying emerging hardware solutions to these problems.