In the last post we reviewed a few hardware implants that may have been used by surveillance states to monitor or gain access directly to a computer at a hardware level and how to potentially detect them. Those types of attacks would aid in persistent compromise. This second post in the series will discuss a few similar hardware devices released in a recent leaks that are used for Radio Frequency (RF) data exfiltration and how you might go about detecting them.
The examples I will cover today are:
HOWLERMONKEY: A short to medium range Radio Transceiver
RAGEMASTER: A hardware implant in a VGA cable that sends video data over RF
LOUDAUTO: A hardware device that sends amplifies audio over RF
SURLYSPAWN: A hardware implant in a keyboard that emits keystrokes over RF
The devices used can vary, but they all employ a similar method of communication via an unknown radio/radar protocol.
Each of these technologies are physical implants, so to detect them you should first check for evidence of a device which has been wired in to an existing device such as a keyboard or other peripheral. Keep in mind, most manufacturing of electronics these days will use PCB traces to wire in all of their chips. If you open a device and see something with micro-gauge wiring putting itself into the device, you may want to ask the manufacturer if that’s normal (or just get another device.)
In some cases, the RF bug will be embedded in the cable itself. For these bugs, the only physical way to detect them would be to cut open the cable’s sleeve and inspect inside the cable. Unfortunately, this option will likely destroy or degrade the cable’s functionality, as you will not be able to put it back together.
An alternative method to detect these types of devices is purely theoretical at this time. If you had a device that could pick up the RF signals from the devices, you would be able to pick up any transmissions from these bugs. You may have seen this in spy movies from the 70’s and 80’s. The only problem is, the frequencies are not known and based on the information leaked the devices may only power on when data is being extracted.
This would be where recent advents in open source may help us. With software defined radio (SDR) gaining popularity, amateur RF listening devices and programs are available for as little as $20 to scan under 2GHz or more beefy ones like the hackRF are available for a few hundred dollars and can listen on frequencies up to 6GHz. This is a huge drop in price to what similar devices would have cost just 10 years ago. Now, why is this a big deal with detecting bugs? These devices can be programmed to listen to a large part of the RF spectrum and detect anomalies based on the software written for them. A suspicious operator could validate if there are RF transmitters in a device by monitoring the spectrum while the device is off (to get a baseline for ambient RF background noise) and then monitoring it again after the suspect device is turned on and transmitting data via radio frequency. The detection device would pick up the signal and alert the user.
A reminder, this series of posts is just a thought project. We will not know with much certainty what steps to take to detect these bugs until one of these bugs is physically found and has been reverse engineered. While I get pretty specific as to what to do with the RF monitoring, I really have not scratched the surface (far beyond the scope of a single blog post). If you're interested in looking further in to software defined radio (SDR), I would recommend checking out the GNURadio project.
In the next post I’ll cover some information on persistent compromises used against routers and device firmware and BIOS.