Amazon Alexa is a smart speaker that can be programmed to control home automation devices via voice commands. For example, Stuart Hinson wanted to be able to control his wirelessly controlled blinds simply by verbally asking Alexa to close or open them. Stuart's blinds could already be controlled via a 433 MHz remote control, so he decided to replicate the control signals on an ESP8266 with 433 MHz transmitter, and interface that with Alexa. The ESP8266 is a cheap and small WiFi capable microchip which many people are using to create IoT devices.
Fortunately replicating the signal was quite easily as all he had to do was record the signal from the remote control with his RTL-SDR, and use the Universal Radio Hacker software to determine the binary bit string and modulation details. Once he had these details, he was able to program the ESP8266 to replicate the signal and transmit it via the 433 MHz transmitter. The remaining steps were all related to setting up an HTTP interface that Alexa could interface with.
If you're interested, we've also previously posted about another Alexa + RTL-SDR mashup which allows Alexa to read out ADS-B information about aircraft flying in your vicinity.
SimpliSafe is an American DIY home security system company that claims over 2 million customers. Their system relies on 433/315 MHz ISM band wireless radio communications between its various sensors, control panels and remote controls. Back in 2016 we already posted about research from Dr. Andrew Zonenberg and Micheal Ossmann who showed that the SimpliSafe wireless communications are unencrypted, and can easily be intercepted, decoded, and spoofed. SimpliSafe responded to those concerns by downplaying them and mentioning that sophisticated hardware was required.
Adam began with some initial manual RF analysis with an RTL-SDR, and then later worked with rtl_433 dev Christian Zuckschwerd to add PiWM demodulation capability, which is the modulation used by SimpliSafe systems. Now Adam is able to easily decode the serial number, pin codes, and status codes transmitted by SimpliSafe sensors and key pads in real time with just an RTL-SDR.
This is very concerning as not only could a burglar easily learn the alarm disarm pincode, but they could also profile your behavior to find an optimal time to break in. For example if you arm your alarm before bed, and disarm in the morning your sleep schedule is being broadcast. It is also possible to determine if a particular door or window has been left open. With a tuned Yagi antenna Adam was able to receive signals from 200+ feet (60m) in free space, and 115 feet (35m) through walls.
In addition to the lack of encryption, Adam also discovered that the SimpliSafe system was susceptible to jamming attacks, and that the tamper detection system can be easily compromised. Adam has disclosed all concerns and findings to SimpliSafe who are aware of the problems. They assure him that next generation systems will not suffer from these flaws. But unfortunately for current generation owners, the hardware will need to be eventually replaced as there is no over the air update capability.
Johannes Smit wanted to be able to view the live data from his SWR WH2303 weather station and send it to a database. Whilst the weather data acquisition software that he paid for worked well, he thought that there must be a cheaper and more fun way to grab the data. But unfortunately the manufacturers would not respond to his request for the RF protocol specifications. So Johannes decided to reverse engineer the protocol using his RTL-SDR instead.
Next he fired up Universal Radio Hacker (URH) and captured a sample of the weather station signal. Using URH he was able to determine the modulation type (FSK) and the bit length parameter (150us). Johannes' next step was to open the weather station, find the RF chip, look up the RF chip information on the web and find the spec sheet. From the spec sheet and internet forum searches he was able to determine the properties of the packet including the sync word and preamble. With this data he was able to determine the packet structure.
Finally he captured a packet and recorded the exact data shown on the weather station at the time of the packet. With this he was able to search the binary data string for the data shown on the weather station, indicating the location of a particular piece of data within the string.
Johannes' tutorial shows just how powerful tools like Universal Radio Hacker can be, and his tutorial is an excellent start for those looking at reverse engineering any of their own local RF protocols.
Foo-Manroot first explains how easily capture and replay a signal with the HackRF. If the signal is simple without any security like rolling codes then a simple replay attack like this will allow the HackRF to control the device quite easily. In the next section he goes on to explain how to actually analyze and synthesize the packets yourself using Python and GNU Radio. Finally he also shows that a brute force attack can be applied once you know how to synthesize the signal. Brute forcing runs over every possible packet combination in a short time and this can be pretty fast for simple protocols like those used in wireless remote controls. His post also includes all the GNU Radio files required so it is easy for someone to replicate his work easily.
If you are interested in controlling simple OOK devices like a wireless powerplug with replay attacks then we have a tutorial for doing this with a simple RTL-SDR and Raspberry Pi running RpiTX which might be useful for those who don't have a HackRF.
In this talk Samy Kamkar shares the exciting details on researching closed systems & creating attack tools to (demonstrate) wirelessly unlocking and starting cars with low-cost tools, home made PCBs, RFID/RF/SDR & more. He describes how to investigate an unknown system, especially when dealing with chips with no public datasheets and undisclosed protocols. Learn how vehicles communicate with keyfobs (LF & UHF), and ultimately how a device would work that can automatically detect the makes/models of keyfobs nearby. Once the keyfobs have been detected, an attacker could choose a vehicle and the device can wirelessly unlock & start the ignition. Like Tinder, but for cars.
Over on his blog "ele y ciencia" has written up two very useful blog posts - one on how to decode AFSK signals from scratch and the other on how to reverse engineer any unknown digital signal. The blog is written entirely in Spanish, but Google translate does a decent enough job at getting the message across (in Chrome right click anywhere on the page and select Translate to English or use the Google translate webpage).
The first post is about decoding an AFSK protocol and explains that you need to record the signal with an RTL-SDR or other SDR, apply a low pass filter to obtain the signal envelope and then apply thresholding with the known baud rate to obtain the demodulated digital signal. The tutorial is high level and just explains the process, but doesn't show how to do it in any software. Later on in the post he goes on to show how he reverse engineered a train-land radiotelephone system and a TCM3105 modem chip which utilizes a FSK system.
In the second post he shows how to decode any unknown digital signal using just an RTL-SDR and Audacity. He starts off with finding and recording an unknown digital signal with an RTL-SDR and then reverse engineers it in a sort of manual fashion without using any tools like Universal Radio Hacker. The post goes through the full details and steps that he took, and in the end he gets data out of the signal discovering that it is data from a Fleet Management System used in his country for monitoring data such as speed and engine data from commercial vehicles like trucks and buses.
The two posts are very detailed and could be an excellent reference for those interested in reverse engineering some unknown digital signals in your area.
Over on GitHub we've just seen the release of a program called rtlmm made by user ps2 which decodes MiniMed RF packets with an RTL-SDR. We weren't entirely such what MiniMed was, but from Googling the name it appears that it is a product by a company called Medtronic who sell medical equipment such as portable automatic insulin pumps and glucose monitors for diabetic patients. These products have RF telemetry links that transmit to a meter which can receives data and forwards it to your phone via Bluetooth LE. Sniffing the telemetry from these sensors could allow you to build up your own data without the need of the meter.
Rtlmm was inspired by a similar program called rtlomni which is a program released a few months ago and made by F5OEO. rtlomni works with Omnipod diabetes insulin pumps and monitors which are similar products to MiniMeds offerings.
The PandwaRF (formerly known as GollumRF) is an RF analysis transceiver tool that can be very useful for investigating ISM band devices that communicate with digitally modulated RF signals. It can be used for applications such as performing replay attacks, brute force attacks, and other analysis. The RX/TX frequency range of the device is from 300 – 928 MHz, with a transmit power of up to +10 dBm.
The PandwaRF is based on the CC1111 chip which is the same chip used in devices like the Yard Stick One from Great Scott Gadgets (creators of the HackRF). Compared to the YS1 the PandwaRF is essentially the same, but designed to be much more portable, with a built in battery and an Android app that you connect to via Bluetooth. This makes it very useful for taking out in the field as no laptop is required to use it, just a phone or tablet. The PandwaRF can be used just like a YardstickOne when plugged into a PC however.
We should also clarify that CC1111 based devices like the PandwaRF and YS1 are not classed as SDRs. Rather they are RF transceiver chips that can demodulate, decode and transmit a fixed set of digital modulation schemes, such as OOK/ASK, 2-FSK, 4-FSK, GFSK, and MSK. While these devices are not able to receive or transmit any arbitrary signal like an SDR, they make reverse engineering, analysis, replay attacks, brute force attacks etc much simpler for common modulation schemes compared to using an SDR for the same purpose.
Early on in the year PandwaRF sent us a sample of their device for review. Unfortunately during that time their Android software was extremely buggy and we were simply unable to use the device properly. Others reported similar troubles on forums and blog comments. However fast forward to today and it now seems that the Android software is stable and functioning properly.
We first tested the PandwaRF on a simple task which was a replay attack. The goal was to record the signal of a cheap wireless RF alarm, and see if we could replay it back. The wireless alarm is controlled with a keyfob.
First we used the Spectrum Analyzer tool in the PandwaRF app to try and get the frequency of the keyfob. The Spectrum Analyzer tool allows you to see about 1.2 MHz of bandwidth. We assumed the signal would be around 433 MHz. After pressing the button a few times the peak showed up at about 433.9 MHz on the spectrum analyzer. The refresh rate of the spectrum analyzer is quite low, so if the signal is not continuous it’s possible to miss the signal, which is we why we had to try several presses before the signal showed. A standard SDR like an RTL-SDR might be better for this initial frequency searching. We confirmed the frequency to be at 433.893 MHz on an RTL-SDR blog V3.
Next we switched to the RX/TX tool. Here you can enter the frequency of interest and set the expected modulation. We know that this device is ASK/OOK modulated, so we chose this setting. You also need to set the data rate. If you don’t know this value then the app has a data rate measuring tool. So we just pressed on the Measure button, and then pressed a button on the remote until it converged to a data rate of 5,121.
Next you need to set the ‘desired payload’. This is how many bytes long the packet is and determines how long the capture is. As we were unsure we simply set it to 250 bytes to ensure that a longer capture was taken. The PandwaRF will keep on receiving until it receives the desired payload of 250 bytes or is stopped manually. Setting it longer allows us to capture a longer signal, and ensure that the replayed signal is received. For this alarm device it is okay if the same signal is played multiple times in a short time frame.
The final setting is the RX Frame length. This determines how many bytes will be captured before transferring the data to Android. So for example, if you set the desired payload to 100 Bytes, and the RX Frame length to 52 bytes, then in total you will capture 104 Bytes of data. The PandwaRF can only transfer in 14, 33, 52, 71 or 90 bytes, so select one that is closest to a multiple of your desired payload.
Finally we pressed on ‘Sniff’ and pressed the ‘bell’ button on the remote. The PandwaRF detected the signal and recorded the data. Now pressing Xmit replays the signal successfully causing the alarm bell to sound.
Brute Force Attack
The PandwaRF can also be used as a brute forcing tool. With cheap alarms the alarm code is relatively short, so can be brute forced in a matter of minutes. The PandwaRF already had a preset mode for our cheap Forecum door alarm, so we simply selected this mode and started the brute force. It gave an estimated brute force time of 28 minutes, which is the time it takes to run through every possible alarm code.
The PandwaRF app currently supports the Idk and PT2262 chipsets, as well as some models of DIO, Extel and Forecum house alarms. If the device that you want to brute force is not yet in their database, then you’ll probably need to do some analysis first on the PC with an SDR. Software like Universal Radio Hacker and DSpectrumGUI are good tools for this. Once you know the structure of the data, then you can program PandwaRF to perform the brute force attack.
Note that their newer ‘PandwaRF Rogue’ product is supposed to be significantly faster at brute forcing. For example the Android software gives us a estimated duration of 28 minutes with the standard PandwaRF, and only 3 minutes with the Rogue.
The Rogue is also able to brute force 32 bit codewords with zero delay in between transmissions. The standard PandwaRF has a minimum delay of 100 ms which can really slow things down. It also allows for function mask bit skipping, enable more brute force patterns and can split the brute force attempt into several steps. Also as we’ve seen from their videos the Rogue has more pre-set commercial devices built into its app.
So if brute forcing is your main use for the PandwaRF then it seems to make sense to get the Rogue. Unfortunately the Rogue is significantly more costly, coming in at 990 euros, vs 145 euros for the standard PandwaRF. Of course you could still use the standard PandwaRF on a PC with tools like rfcat to perform a faster brute force attack as well, just like you would with a YardstickOne.
The code runs on the Android device and not on the PandwaRF, so each RF command generates a bluetooth transfer which can be quite slow. They write this is why they have created a specific brute force implementation in the app, so that they can run their native brute force code on the PandwaRF itself, which is must faster than transferring the RF command for every brute force step.
Overall the PandwaRF is a very handy tool for doing replay and brute force attacks while in the field. It can also be converted back into a PC based CC1111 device, like a Yardstick One simply by plugging it into a computer with a USB cable so you’re not missing out on that functionality either.
Compared to the Yardstick One the cost is a bit more, with the Yardstick One costing $99 USD at most outlets, and the PandwaRF costing 145 Euros (~$173 USD). So it is probably only really worth it if you are doing field testing.
That said, now that the PandwaRF software seems stable it is an excellent tool for investigating wireless devices in a simpler way compared to with an SDR. An SDR is still much more powerful, but tools like this simplify the process significantly. The best set of tools for reverse engineering would be a SDR combined with a device like this.
In the future it looks like they plan to implement new features such as De Bruijn (OpenSesame) attack’s and rolling code attacks and we look forward to testing those out.