French mobile femtocells vulnerable to rogue base station hack

Security researchers have detailed further flaws in the femtocell base station technology supplied by mobile carriers to consumers and small businesses as a means to improve 3G mobile connectivity in buildings by taking advantage of existing broadband connections.

Security shortcomings in Vodafone's femtocell signal booster technology – actually fixed at the start of 2010 – hit the news last month with the publication of research by The Hacker's Choice (THC). The group said vulnerabilities it uncovered in the process of reverse-engineering Vodafone-supplied base station kit created a means to turn a femtocell into an interception device.

The hack (until it was fixed) created a possible mechanism to either intercept or make calls at the expense of victims, providing they were first tricked into connecting to a compromised base station.

New research suggests the security of femtocells is a widespread problem and not restricted to Vodafone. Base station technology supplied by SFR, France's second mobile carrier, is also vulnerable, according to research presented at the Black Hat conference in Las Vegas last weekend.

Nico Golde and other security researchers discovered that the software bug recovery process applied by SFR to faulty base stations is flawed. SFR understandably doesn't want to send around an engineer every time a femtocell goes awry, so it pushes a software rebuild out to affected devices in the hope that this will resolve any glitch, an approach it takes when basic support functions ("have you tried turning it of and then on again?") fail.

Unfortunately the Ubiquisys-manufactured devices supplied by SFR are not configured to authenticate image servers. That means an attacker can push his own configuration and (backdoored) firmware. SFR recently made this process more complicated by failing to address the root cause of the problem.

"The images are signed, but the public key can be provided in the configuration file (which is not signed)," the researchers explain. "We were able to analyse the procedure because an unencrypted recovery image could be retrieved. This has been fixed, but we now have the tools to decrypt them.

"However, there are still other ways to get unencrypted images," they add.

The security shortcoming creates a possible mechanism to establish rogue femtocells. This, in turn, opens the door to a range of attacks capable of targeting end-users being logged into a femtocell. Base station owners as well as network operators are at risk.

Possible exploits run the range from turning compromised devices into IMSI-catchers to call interception, making calls or sending messages at the expense of victims, and taking control of other femtocells. Compromised devices might even be used as the first part of stepping-stone attacks against operators' network, Golde and his colleagues suggest.

Other carrier networks may also be open to attack, though this has yet to be tested. "This is definitely not only a problem limited to one operator," Golde told El Reg. Golde worked on the research together with fellow security researchers Kévin Redon and Ravi Borgaonkar.

The attacks are explained in greater depth in a series of slides that can be found here (76-page PDF/6MB). ®