Analytics // Security Monitoring
8/1/2014
01:00 PM
100%
0%

Some Security Tools Fail With Multipath TCP

Researchers at Black Hat USA will reveal how multipath TCP breaks assumptions about how TCP works -- and the implications for network security.

The need for hardy networking connections has led to the development of Multipath TCP (MPTCP), which allows a Transmission Control Protocol (TCP) connection to use multiple paths to maximize resource usage and increase redundancy. But the explosion in mobile and Internet of Things devices also requires network security tools keep pace with change.

Neohapsis researchers Catherine Pearce and Patrick Thomas say very few security and network management tools are up to that challenge when it comes to MPTCP streams. At a presentation at the upcoming Black Hat USA conference in Las Vegas, the two plan to discuss how MPTCP eviscerates assumptions about TCP that are made by both tools and network engineers alike.

TCP is a foundational protocol for communication on the Internet that has largely stood the test of time, says Thomas. "It's used everywhere; it's used by everything," he says. "It's been around since the dawn of the Internet."

Yet TCP is not set up for the future of the Internet people are seeing now -- a highly-connected world featuring different types of connectivity, says Thomas.

This had led to the development of MPTCP, which allows TCP to talk over multiple paths simultaneously, decoupling it from a specific IP address. Doing this, however, creates a new reality for security. For starters, it affects the ability of intrusion detection systems to inspect, correlate, and reassemble traffic. This can add a new wrinkle to fragmentation attacks, Thomas says.

"An intrusion detection system that is not multipath TCP-aware, sees five different connections coming from different IP addresses, has no conception that they are related, and on each of them it sees complete garbage data," he says.

"If any of your security decisions, tools, thought-processes, manual processes, if they rely on any of... these four things, then something in those is going to break," he says. "Those four things that we've got are: If you expect to see all app layer data within a TCP stream; if you expect to differentiate clients from servers based on the connection direction; if you expect to tamper with or close bad connections midstream; or if attempt to associate logical connections to IP addresses. If you make any security decisions based on any of those, then those security mechanisms are going to break in the face of MPTCP."

During their presentation, which is scheduled for Wednesday at 3:30 p.m., the researchers will show tools and strategies for understanding and mitigating the risk of MPTCP-capable devices on a network.

Multipath TCP changes the way the Internet works at its core, explains Pearce.

"If we're not ready for this the impact of this could be probably at least as great as IPv6," she says, adding that the goal of the talk at Black Hat is to raise awareness.

"We want security to keep up with technology," she says.

 

Brian Prince is a freelance writer for a number of IT security-focused publications. Prior to becoming a freelance reporter, he worked at eWEEK for five years covering not only security, but also a variety of other subjects in the tech industry. Before that, he worked as a ... View Full Bio

Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
EricD388
50%
50%
EricD388,
User Rank: Apprentice
8/1/2014 | 4:06:56 PM
Wireless Aggregation of Internet Bandwidth
This technology has been in the works for quite some time.  Delangis has three U.S. patents on this technology dating back to 2002.  Check out 7606156 and 8228801 for starters.
Register for Dark Reading Newsletters
White Papers
Cartoon
Current Issue
Dark Reading, January 2015
To find and fix exploits aimed directly at your business, stop waiting for alerts and become a proactive hunter.
Flash Poll
Video
Slideshows
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2014-3580
Published: 2014-12-18
The mod_dav_svn Apache HTTPD server module in Apache Subversion 1.x before 1.7.19 and 1.8.x before 1.8.11 allows remote attackers to cause a denial of service (NULL pointer dereference and server crash) via a REPORT request for a resource that does not exist.

CVE-2014-6076
Published: 2014-12-18
IBM Security Access Manager for Mobile 8.x before 8.0.1 and Security Access Manager for Web 7.x before 7.0.0 FP10 and 8.x before 8.0.1 allow remote attackers to conduct clickjacking attacks via a crafted web site.

CVE-2014-6077
Published: 2014-12-18
Cross-site request forgery (CSRF) vulnerability in IBM Security Access Manager for Mobile 8.x before 8.0.1 and Security Access Manager for Web 7.x before 7.0.0 FP10 and 8.x before 8.0.1 allows remote attackers to hijack the authentication of arbitrary users for requests that insert XSS sequences.

CVE-2014-6078
Published: 2014-12-18
IBM Security Access Manager for Mobile 8.x before 8.0.1 and Security Access Manager for Web 7.x before 7.0.0 FP10 and 8.x before 8.0.1 do not have a lockout period after invalid login attempts, which makes it easier for remote attackers to obtain admin access via a brute-force attack.

CVE-2014-6080
Published: 2014-12-18
SQL injection vulnerability in IBM Security Access Manager for Mobile 8.x before 8.0.1 and Security Access Manager for Web 7.x before 7.0.0 FP10 and 8.x before 8.0.1 allows remote authenticated users to execute arbitrary SQL commands via unspecified vectors.

Best of the Web
Dark Reading Radio
Archived Dark Reading Radio
Join us Wednesday, Dec. 17 at 1 p.m. Eastern Time to hear what employers are really looking for in a chief information security officer -- it may not be what you think.