Cisco WLAN Flaws May Be Typical Of Many Proprietary Systems, Researcher Says

Researchers next week at Black Hat Europe will outline vulnerabilities in Cisco wireless LAN technology they say may be indicative of flaws that exist in other proprietary technologies as well.

In a session called "Hacking Cisco Enterprise WLANs," researchers Enno Rey and Daniel Mende -- both employed by German penetration-testing firm ERNW -- will offer a look at some of the flaws found in existing, proprietary Cisco wireless LAN products

"We did our proof-of-concepts on Cisco WLAN products because we happened to have access to them at our site," Rey says. "But we're not Cisco-bashing here. What we found could be found in a lot of proprietary technologies made by a lot of other vendors, as well."

In the presentation, Rey and Mende will demonstrate how proprietary technologies -- particularly older technologies that are no longer strategic to the vendor -- often fail to receive the vulnerability assessments and scrutiny of more current Web- and standards-based technologies. The ERNW researchers evaluated three generations of Cisco wireless LAN products, ranging from the first-generation Cisco Structured Wireless-Aware Networks (SWANs) first introduced a decade ago to the more current Cisco Unified Wireless Network (CUWN). In each case, they found flaws that were relatively easy to spot and wouldn't be difficult to exploit, Rey said.

"You wouldn't be able to find them with a standard Nessus scan, but it didn't take us long to find these," Rey remarks. ERNW has notified Cisco of the vulnerabilities and has agreed not to disclose technical details until the networking giant has a chance to evaluate and patch the flaws.

In the case of SWAN, the researchers took a hard look at Cisco's proprietary Wireless LAN Context Control Protocol (WLCCP), which enables wireless access points to communicate. They found a number of flaws in the authentication methods used by the APs that could allow an attacker to extract cryptographic material -- including the encryption keys used on the wireless network.

A central part of SWAN's authentication process relies on a Cisco protocol called LEAP, which has been proved to be insecure, Rey observes. Cisco has addressed this weakness by generating another encryption key on top of LEAP, but because the key is generated via the LEAP authentication process, it remains vulnerable, he says.

The more current generation of Cisco WLAN products, CUWN, is more secure than SWAN, but still contains exploitable flaws, Rey states. In particular, management interfaces used by the WLAN systems, often based on Web interfaces and/or security-weak SNMP protocols, could put the proprietary technology at risk.

"When we do our penetration testing, we often find weaknesses in the management interfaces," Rey says. "The Web interfaces often have well-known vulnerabilities, such as cross-site scripting. The weaknesses in SNMP are also well known. In this case, we found those weaknesses in Cisco WLAN products, but we often find those same weaknesses in other technologies, such as voice over IP or storage systems."

A key point of the presentation, Rey says, is to point out that proprietary systems often are not vetted and tested as scrupulously as more mainstream systems. "A large vendor acquires a small vendor, and the smaller vendor's products are integrated without the same attention to detail as they would be if the larger vendor had developed the products itself," he says. "The quality assessment and security assurance practices are not as thorough."

And the vendors are not the only ones at fault for the vulnerabilities that ERNW often finds in its penetration testing clients' environments, Rey says. "We see administrators exposing their management interfaces to the wider corporate network, making them accessible," he says. "Sometimes the complexity of the proprietary solutions is such that the IT people are happy it's running at all. They get it up and running, and they say, 'We'll harden it later.' But it doesn't happen."

Enterprise IT administrators need to understand that each time they bring in a new set of devices or technologies -- particularly proprietary ones -- they may introduce new security risks, even through simple management interfaces that are built into the newly-installed devices, Rey observes. "Anytime you put in a new technology, you need to look at the vulnerabilities you may be creating," he says. "Those new devices may have different management procedures that aren't integrated with your existing secure management procedures."

Have a comment on this story? Please click "Discuss" below. If you'd like to contact Dark Reading's editors directly, send us a message.