Welcome back, aspiring cyberwarriors!

We are really glad to see you back for the second part of this series. In the first article, we explored some of the cheapest and most accessible ways to build your own hacking drone. We looked at practical deployment problems, discussed how difficult stable control can be, and even built small helper scripts to make your life easier. That was your first step into this subject where drones become independent cyber platforms instead of just flying gadgets. 

We came to the conclusion that the best way to manage our drone would be via 4G. Currently, in 2026, Russia is adapting a new strategy in which it is switching to 4G to control drones. An example of this is the family of Shahed drones. These drones are generally built as long-range, loitering attack platforms that use pre-programmed navigation systems, and initially they relied only on satellite guidance to reach their targets rather than on a constant 4G data link. However, in some reported variants, cellular connectivity was used to support telemetry and control-related functionality.

In recent years, Russia has been observed modifying these drones to carry different types of payloads and weapons, including missiles and MANPADS (Man-Portable Air-Defense System) mounted onto the airframe. The same principle applies here as with other drones. Once you are no longer restricted to a short-range Wi-Fi control link and move to longer-range communication options, your main limitation becomes power. In other words, the energy source ultimately defines how long the aircraft can stay in the air.

Today, we will go further. In this part, we are going to remove the smartphone from the back of the drone to reduce weight. The free space will instead be used for chipsets and antennas.

4G > UART > Drone

In the previous part, you may have asked yourself why an attacker would try to remotely connect to a drone through its obvious control interfaces, such as Wi-Fi. Why not simply connect directly to the flight controller and bypass the standard communication layers altogether? In the world of consumer-ready drones, you will quickly meet the same obstacle over and over again. These drones usually run closed proprietary control protocols. Before you can talk to them directly, you first need to reverse engineer how everything works, which is neither simple nor fast.

However, there is another world of open-source drone-control platforms. These include projects such as Betaflight, iNav, and Ardupilot. The simplest of these, Betaflight, supports direct control-motor command transmission over UART. If you have ever worked with microcontrollers, UART will feel familiar. The beauty here is that once a drone listens over UART, it can be controlled by almost any small Linux single-board computer. All you need to do is connect a 4G module and configure a VPN, and suddenly you have a controllable airborne hacking robot that is reachable from anywhere with mobile coverage. Working with open systems really is a pleasure because nothing is truly hidden.

So, what does the hacker need? The first requirement is a tiny and lightweight single-board computer, paired with a compact 4G modem. A very convenient combination is the NanoPi Neo Air together with the Sim7600G module. Both are extremely small and almost the same size, which makes mounting easier.

The NanoPi communicates with the 4G modem over UART. It actually has three UART interfaces. One UART can be used exclusively for Internet connectivity, and another one can be used for controlling the drone flight controller. The pin layout looks complicated at first, but once you understand which UART maps to which pins, the wiring becomes straightforward.

After some careful soldering, the finished 4G control module will look like this:

Even very simple flight controllers usually support at least two UART ports. One of these is normally already connected to the drone’s traditional radio receiver, while the second one remains available. This second UART can be connected to the NanoPi. The wiring process is exactly the same as adding a normal RC receiver.

The advantage of this approach is flexibility. You can seamlessly switch between control modes through software settings rather than physically rewiring connectors. You attach the NanoPi and Sim7600G, connect the cable, configure the protocol, and the drone now supports 4G-based remote control.

Depending on your drone’s layout, the board can be mounted under the frame, inside the body, or even inside 3D-printed brackets. Once the hardware is complete, it is time to move into software. The NanoPi is convenient because, when powered, it exposes a USB-based console. You do not even need a monitor. Just run a terminal such as:

nanoPi >  minicom -D /dev/ttyACM0 -b 9600

Then disable services that you do not need:

nanoPi >  systemctl disable wpa_supplicant.service

nanoPi >  systemctl disable NetworkManager.service

Enable the correct UART interfaces with:

nanoPi >  armbian-config

From the System menu you go to Hardware and enable UART1 and UART2, then reboot.

Next, install your toolkit:

nanoPi >  apt install minicom openvpn python3-pip cvlc

Minicom is useful for quickly checking UART traffic. For example, check modem communication like this:

minicom -D /dev/ttyS1 -b 115200
AT

If all is well, then you need to config files for the modem. The first one goes to /etc/ppp/peers/telecom. Replace “telecom” with the name of the cellular provider you are going to use to establish 4G connection.

And the second one goes to /etc/chatscripts/gprs

To activate 4G connectivity, you can run:

nanoPi >  pon telecom

Once you confirm connectivity using ping, you should enable automatic startup using the interfaces file. Open /etc/network/interfaces and add these lines:

auto telecom
iface telecom inet ppp
provider telecom

Now comes the logical connectivity layer. To ensure you can always reach the drone securely, connect it to a central VPN server:

nanoPi > cp your_vds.ovpn /etc/openvpn/client/vds.conf

nanoPi > systemctl enable openvpn-client@vds

This allows your drone to “phone home” every time it powers on.

Next, you must control the drone motors. Flight controllers speak many logical control languages, but with UART the easiest option is the MSP protocol. We install a Python library for working with it:

NanoPi > cd /opt/; git clone https://github.com/alduxvm/pyMultiWii

NanoPi > pip3 install pyserial

The protocol is quite simple, and the library itself only requires knowing the port number. The NanoPi is connected to the drone’s flight controller via UART2, which corresponds to the ttyS2 port. Once you have the port, you can start sending values for the main channels: roll, propeller RPM/throttle, and so on, as well as auxiliary channels:

Find the script on our GitHub and place the it in ~/src/ named as control.py

The NanoPi uses UART2 for drone communication, which maps to ttyS2. You send MSP commands containing throttle, pitch, roll, yaw, and auxiliary values. An important detail is that the flight controller expects constant updates. Even if the drone is idle on the ground, neutral values must continue to be transmitted. If this stops, the controller assumes communication loss. The flight controller must also be told that MSP data is coming through UART2. In Betaflight Configurator you assign UART2 to MSP mode.

We are switching the active UART for the receiver (the NanoPi is connected to UART2 on the flight controller, while the stock receiver is connected to UART1). Next we go to Connection and select MSP as the control protocol.

If configured properly, you now have a drone that you can control over unlimited distance as long as mobile coverage exists and your battery holds out. For video streaming, connect a DVP camera to the NanoPi and stream using VLC like this:

cvlc v4l2:///dev/video0:chroma=h264:width=800:height= \
--sout '#transcode{vcodec=h264,acodec=mp3,samplerate=44100}:std{access=http,mux=ffmpeg{mux=flv},dst=0.0.0.0:8080}' -vvv

The live feed becomes available at:

http://drone:8080/

Here “drone” is the VPN IP address of the NanoPi.

To make piloting practical, you still need a control interface. One method is to use a real transmitter such as EdgeTX acting as a HID device. Another approach is to create a small JavaScript web app that reads keyboard or touchscreen input and sends commands via WebSockets. If you prefer Ardupilot, there are even ready-made control stacks.

By now, your drone is more than a toy. It is a remotely accessible cyber platform operating anywhere there is mobile coverage.

Protection Against Jammers

Previously we discussed how buildings and range limitations affect RF-based drone control. With mobile-controlled drones, cellular towers actually become allies instead of obstacles. However, drones can face anti-drone jammers. Most jammers block the 2.4 GHz band, because many consumer drones use this range. Higher end jammers also attack 800-900 MHz and 2.4 GHz used by RC systems like TBS, ELRS, and FRSKY. The most common method though is GPS jamming and spoofing. Spoofing lets an attacker broadcast fake satellite signals so the drone believes false coordinates. Since drone communication links are normally encrypted, GPS becomes the weak point. That means a cautious attacker may prefer to disable GPS completely. Luckily, on many open systems such as Betaflight drones or FPV cinewhoops, GPS is optional. Indoor drones usually do not use GPS anyway.

As for mobile-controlled drones, jamming becomes significantly more difficult. To cut the drone off completely, the defender must jam all relevant 4G, 3G, and 2G bands across multiple frequencies. If 4G is jammed, the modem falls back to 3G. If 3G goes down, it falls back to 2G. This layering makes mobile-controlled drones surprisingly resilient. Of course, extremely powerful directional RF weapons exist that wipe out all local radio communication when aimed precisely. But these tools are expensive and require high accuracy.

Summary

We transformed the drone into a fully independent device capable of long-range remote operation via mobile networks. The smartphone was replaced with a NanoPi Neo Air and a Sim7600G 4G modem, routed UART communication directly into the flight controller, and configured MSP-based command delivery. We also explored VPN connectivity, video streaming, and modern control interfaces ranging from RC transmitters to browser-based tools. Open-source flight controllers give us incredible flexibility.

In Part 3, we will build the attacking part and carry out our first wireless attack.

If you like the work we’re doing here and want to take your skills even further, we also offer a full SDR for Hackers Career Path. It’s a structured training program designed to guide you from the fundamentals of Software-Defined Radio all the way to advanced, real-world applications in cybersecurity and signals intelligence.