Hackers say coming air traffic control system lets them hijack planes
FAA says it can spot hacking attempts, but won't allow independent 'stress tests'
January 11, 2013 — CSO — An ongoing multibillion-dollar overhaul of the nation's air traffic control (ATC) system is designed to make commercial aviation more efficient, more environmentally friendly and safer by 2025.
But some white-hat hackers are questioning the safety part. The Next Generation Air Transportation System (NextGen) will rely on Global Positioning Systems (GPS) instead of radar. And so far, several hackers have said they were able to demonstrate the capability to hijack aircraft by spoofing their GPS components.
The Federal Aviation Administration (FAA) has declared that it already has multiple measures to detect fake signals. But it has so far not allowed any independent testing of the system.
The hacking exploits are not new. National Public Radio's "All Tech Considered" reported last August that Brad Haines, a Canadian computer consultant known online as "RenderMan," noted that the radio signals aircraft will send out to mark their identity and location under NextGen, called automatic dependent surveillance-broadcast (ADS-B), were both unencrypted and unauthenticated.
By spoofing those signals, Haines said he could create fake "ghost planes."
"If I can inject 50 extra flights onto an air traffic controller's screen, they are not going to know what is going on," he told NPR. "If you could introduce enough chaos into the system - for even an hour - that hour will ripple though the entire world's air traffic control."
Haines presented his findings at the Defcon hacking conference in Las Vegas last summer [http://www.csoonline.com/article/713233/the-black-hat-bsideslv-and-defcon-post-mortem].
Then there is the group of researchers from the University of Texas that successfully hijacked a civilian drone at the White Sands Missile Range in New Mexico during a test organized by the Department of Homeland Security (DHS) last summer.
The system used to hijack the drone cost about $1,000. The NextGen program is expected to cost taxpayers $27 billion, plus another $10 billion spent by the commercial aviation industry.
In a third case, NPR reported that Andrei Costin, a Romanian graduate student in France, was able to build a software-defined radio hooked to a computer that created fake ADS-B signals in a lab. It cost him about $2,000. Costin made a presentation at last summer's Black Hat conference.
Paul Rosenzweig, founder of Red Branch Law & Consulting and a former deputy assistant secretary for policy at DHS, wrote in a post last week on Lawfare that this amounts to the FAA continuing to dig itself into a deeper hole. One problem, he wrote, is that the eventual goal is to eliminate radar, which is inefficient because it requires planes to fly on designated radar routes.
"But the hardware for radar broadcasting and reception can't (that I know of) be spoofed," he wrote. "Today, when planes fly using GPS they 'double check' their location with radar. [But] the entire plan behind NextGen is to eventually get rid of the radar system -- an expensive 20th century relic, I guess. But then we are completely dependent on GPS for control."
The FAA told NPR that besides confirming ADS-B signals with radar, the NextGen system will automatically check to make sure the correct receivers are picking up the correct signals. If a "ghost plane" is sending a signal to the wrong receiver, it would be spotted as fake. Third, it will use a technique called "multilateration" to determine exactly where every ADS-B signal is sent from.
Nick Foster, a partner of Brad Haines, praised the use of multilateration. "But I still wonder if it would be possible to fool the system on the edges," he told NPR. "I think the FAA should open it up and let us test it."
The risks of GPS hacks extend beyond aviation. Logan Scott, a GPS industry consultant, told Wired magazine last year [http://www.wired.com/dangerroom/2012/07/drone-hijacking/2/] that GPS is also used to control the power grid, to power banking operations including ATMs and to keep oil platforms in position. The world's cellular networks also rely on it.
And given that it is free, unauthenticated and unencrypted makes it vulnerable. "The core problem is that we've got a GPS infrastructure which is based on a security architecture out of the 1970s," Scott said.
Not everybody sees the GPS vulnerability as a major safety problem, however. Martin Fisher is now director of information security at Wellstar Health System, but worked previously in commercial aviation for 14 years. He said radar will still be around, even when the transition to NextGen is complete.
"Don't for a moment believe there won't be radar anymore," he said. "Commercial aircraft will still have anti-collision radar and proximity alarms."
Beyond that, he said, "do not make the assumption that the pilots flying your aircraft simply follow the instructions of ATC like automatons. These are very highly trained men and women with years of experience flying day, night, good weather, bad weather."
Paul Rosenzweig said he would still be much more comfortable if the FAA would allow the system to be "stress tested."
Whatever bugs are in the system, there may be more than 12 years available to fix them. The Washington Post reported in September that Calvin L. Scovel III, inspector general for the Department of Transportation, told a House subcommittee that the program was "four years behind schedule and $330 million over budget."
Read more about access control in CSOonline's Access Control section.
Other stories by Taylor Armerding