http://www.elpasotimes.com/apps/pbcs.dl ... 20325/1001

Researcher says math can protect border

Louie Gilot
El Paso Times
Monday, December 12, 2005

Recently, U.S. politicians have come up with costly ideas to improve border security. Some want to hire more Border Patrol agents. Others push for a guest-worker program. And one, Rep. Duncan Hunter, R-Calif., wants to build a 2,000-mile-long wall.

Stefan Schmidt, a researcher at New Mexico State University's Physical Science Laboratory, came up with a pretty cheap alternative -- mathematics.

"We have ways of modeling to see how we can make the border more secure without paying too much money," Schmidt said.

Border Patrol spokesman Doug Mosier said his agency is constantly gathering intelligence to try to predict where smugglers are likely to cross into the United States and where to allocate resources. But the Border Patrol crime-fighting strategy doesn't yet involve mathematical formulas, Mosier said.

The idea of applying mathematics to security is not so far-fetched.

Mathematics played a role in military planning during World War II. One example is John Nash's work in game theory -- which studies incentives in human behavior -- featured in the 2001 Oscar-winning film "A Beautiful Mind." The opening line in that movie is "Mathematicians won the war."

More recently, Congress funded the Homeland Security Institute to advise the Department of Homeland Security and keep tabs on worthy new research. Representatives of the institute attended a meeting of mathematicians, including Schmidt, at Rutgers University last month to discuss how mathematics could be applied to the fight against terrorism.

Schmidt's idea is to use the Soviet Union's reflexive theory to tackle border security. The theory states the behavior of the adversary can be controlled through the selective release of information and disinformation.

On the border, it means that the Border Patrol could somehow convince drug smugglers that their preferred route is well-guarded and lead them to look for another passage. The Border Patrol could also lead the smugglers to believe that one crossing point is unguarded and set a trap for them. Under these scenarios, the Border Patrol would be able to direct smugglers to points that are cheaper and easier for the agency to guard.

Where these points are is calculated by formulas that take into account, "What you think. What you think they think. What they think. What they think you think," Schmidt said.

Strategies would involve disseminating false information along the border, a proposition that could become controversial.

But "propaganda is part of the game," Schmidt said. So are decoys, such as fake surveillance cameras. But mostly, Schmidt's formulas could be used to decide where to assign agents.

Schmidt says his prototype works well but he still needs a few more years to fine tune his model and scale it out, that is, apply it to a larger data pool.

Schmidt's research partner, Jonathan Farley, a Science Fellow at Stanford University's Center for International Security and Cooperation, said the research could be applied to things as varied as dealing with roadside bombs in Iraq --by leading insurgents to place the bombs in areas that would do less damage -- or where to place campaign advertising.

Border security is a logical application, he said.

"You can't have everybody along the border all the time," Farley said.

Schmidt and Farley are co-founders of Phoenix Mathematical Systems Modeling Inc., a company that develops mathematical approaches to homeland security.

Louie Gilot may be reached at lgilot@elpasotimes.com; 546-6131.