Starlink is being jammed from the ground: why the satellite's vulnerability ended up in the trenches

Starlink user terminals have become as consumable on the front lines as a radio or battery. The system is often called invulnerable: thousands of satellites in low orbit, constant rotation of satellites, global coverage. But it's being jammed close to the ground. The downlink is deployed in the 10,7–12,7 GHz range, and the entire chain converges at a single point, the terminal that the soldier buries in a trench. This is where the work of the systems has shifted. EW.

Starlink global coverage map, August 2025

The nerve that was carried to the front line

The system consists of three components: low-orbit satellites, ground-based gateways, and user terminals. The satellite receives data from the gateway and transmits it down to the terminal via a downlink. This downlink operates in the 10,7–12,7 GHz range, with a total bandwidth of 2 GHz. The signal is divided into beams with spatial separation and polarization.

Line power is calculated strictly. Under typical conditions (11,7 GHz carrier, 250 MHz bandwidth, 1123 km distance), the signal-to-noise ratio at the receiver is approximately 11,2 dB. This is sufficient for speeds up to 822 Mbps. For an infantry squad, such a channel is excessive: it supports video from drones, management, and real-time data exchange. This channel is both the terminal's strength and its weakness.

Starlink system characteristics and frequency-power distribution

Comparison of parameters and functionality of gateway and user beams

A distributed constellation is nearly impossible to disable from above. There are thousands of satellites, constantly replacing each other over the area; the removal of a single satellite doesn't solve anything. But this stability rests on the terminal link. The space segment can be left alone: ​​simply block the terminal's reception below, and the entire link is severed at the last meter, right at the ground.

Time and frequency characteristics of the signal, Doppler shift trajectory

Calculating the downlink power budget and transmission speed

The weak link lies in the mud

The terminal's antenna has a narrow beam pattern, approximately 1,3°. The beam pattern determines the sector in which the antenna receives a signal: for Starlink, this is essentially looking up into the sky through a narrow tube, with high gain in this sector. The terminal receives Ku-band from the satellite and, through a built-in Wi-Fi router, distributes data to end devices on the familiar 2,4 and 5,8 GHz frequencies.

Antenna design, evolution of first and second generation terminals

Starlink Terminal Wi-Fi Router Specifications and Interfaces

Herein lies the contradiction that the enemy must resolve every day. At the front, the terminal becomes a visible target for artillery and optical reconnaissance, so it's buried in a trench, hidden in a crater, or positioned behind a wall. Protection from shrapnel requires hiding the antenna, while communications require an open sector facing the sky. By burying the terminal, the soldier narrows its already narrow field of view and reduces its energy reserves, as the satellite's useful signal is already at its limit. A drone simply needs to enter this sector from above, at close range, and jam the reception, and a low-power electronic warfare module on a lightweight carrier is sufficient for this.

Comparison of key parameters of Starlink and GNSS

Camouflage and installation of Starlink terminals in frontline trenches

There's nothing fundamentally new in the logic itself. The battle between communications and jamming is a perennial theme in military radio: a new channel appears, and a method is found to jam it. From the interception and jamming of radio networks in past wars to the suppression of VHF and GNSS today, the mechanics are the same: find a band, assess the signal structure, and overpower it with your power. Starlink simply transferred this duel to the Ku-band and a new subscriber type.

According to the frame and according to the protocol

The main weakness of downlink is that the signal is strictly deterministic. Its structure is known: carrier width is 250 MHz (effective bandwidth is 240 MHz), frame duration is 1,3288 ms, and the guard interval between frames is 4,55 μs. Pilot synchronization signals, that is, the service marks by which the receiver detects and maintains communication, are constantly broadcast on 11,075, 11,325, and 11,575 GHz. The more precisely the signal structure is known, the more precisely the targeting can be achieved: instead of flooding the entire band with noise, jamming can be precisely targeted at the frequencies and moments where it will disrupt communication. And targeted interference requires significantly less energy.

Spectrogram and waterfall of the intercepted downlink signal in the 250 MHz band

From here a ladder of influence is built, from brute force to precision:

• Broadband jamming is the process of jamming an entire band with noise. It works, but it requires a lot of power, which is insufficient on a lightweight carrier.
  
• Comb interference is a set of narrow spectral lines with a given pitch; they overlap subchannels pointwise. The same problem can be solved with less energy.
  
• Protocol-level jamming isn't about jamming, it's about deceiving. Starlink transmits data using the DVB-S2X standard; the system decodes physical-layer service packets and inserts a fake session-breaking command during the interframe pause. The terminal mistakes it for the real thing and terminates the connection, which consumes negligible power.
  

Scheme for generating analog and digital interference

The computing core is built on a combination of an FPGA and a digital signal processor—programmable logic that synthesizes the desired interference waveform on the fly. This architecture allows for a controlled comb pitch and dynamically overlaps parallel Starlink subchannels, adapting to the signal.

Frequency collision of pulsed pseudo-random interference in the passband

FPGA and DAC based multi-frequency comb noise synthesizer

Comparison of three types of interference: barrage, comb and chirp

The developers are fitting the entire channel into a carrier weighing up to 10 kg. This is no longer a truck-mounted electronic warfare station, but a module that lifts a conventional man-portable copter into the air. This changes everything for the defender: the Starlink jammer is becoming mass-produced and is coming from the same direction as reconnaissance drones. A method that required bulky equipment is being condensed into an onboard UAV module. The announced field tests illustrate this: on a rig, a stable channel is detected via the Starlink app, the drone's transmitter is turned on, and the status "Connection Lost" appears on the screen. The rig test doesn't reveal how stable the result is in real-world broadcasts with the system's own interference protection.

Parameters of the onboard reconnaissance and suppression system

Network topology with simulated suppression and forced session termination

Test deployment: suppression of a ground terminal by an airborne UAV system

When a jammer is sold from a warehouse

More telling than the algorithms themselves is something else: a third-party manufacturer already has a commercially available component base for this task. The Chinese company RF Apex openly offers ready-made microwave modules. These include a 7–13 GHz power amplifier with an output power of approximately 100 W in saturation mode (this range completely covers the Starlink 10,7–12,7 GHz subscriber line) and a 300–3000 MHz broadband module for FPV drone control channels and GNSS navigation. Both are built on gallium nitride (GaN) and are designed specifically for drone installation.

One hundred watts at 7–13 GHz near a buried terminal is enough to drown out a weak satellite signal. The manufacturer, of course, cites this figure for advertising purposes, as a maximum under ideal conditions. But the very fact that such products are listed is important. It means that Starlink jamming has moved beyond the laboratory stage: the physics were analyzed and the algorithms refined even earlier, and now mass-produced components have caught up. Assembling an electronic warfare system for satellite communications can now be done from prefabricated modules; designing from scratch is no longer necessary.

Mobile air and ground electronic countermeasure systems

This is where the Russian side's weakness lies, and it's not in their understanding of physics—that's fine. The problem lies in the mass-produced microwave components. High-power gallium nitride modules, stable over a wide bandwidth and suitable for mass installation on carriers, are produced in limited quantities here, and a significant portion of the components are imported. Analyzing the signal structure and writing an interference algorithm is half the battle. The other half rests on the power amplifier: hundreds of them need to be produced, and that's not the case right now.

The physical structure of the Starlink satellite's phased array antenna

The system, which was touted as invulnerable, can disable a drone with just a few kilograms of weight, all thanks to that very same buried terminal. The orbital constellation is out of reach, and it's not even necessary. Everything is decided at the last meter—in the terminal and in the downlink frame structure, which the enemy has already dismantled. Commercially available microwave modules for this task mean that this is no longer a unique find, but a standardized solution assembled from off-the-shelf components.

Comments

[Vladimir_2U]

Jamming is certainly necessary, and that's good, but the author doesn't specify the range from which it'll be possible to jam a Starlink user terminal. And I'm afraid that range will be negligible. It would be much simpler and faster to drop a mine or shell on a detected terminal. So, a drone for electronic reconnaissance and targeting is more needed than a jamming drone, since jamming means hovering over the enemy...

What works well over one's own territory may not be suitable for the "gray zone," let alone the enemy's.

Low-flying satellites can be affected reversibly by jamming them near the LBS. While dangerous, it does have a significant impact area, and the goal of disabling (not destroying) them is to do so from deep within one's own territory. But this is a matter of political will on the part of the leadership.

[rfapex]

You're absolutely right, the issue of line energy and operating distance is always key. But this is pure physics: a useful signal from a satellite orbiting 500+ km arrives at the ground at the edge of noise.

100 W of pure GaN power in the Ku-band, raised to a height of 100–150 meters, creates such a density of interference that the subscriber terminal “goes blind” not only when hovering directly above the trench, but also from a distance of 1–1.5 km (including due to the side lobes of the diagram).

Regarding artillery, a buried and camouflaged terminal must first be exposed by reconnaissance and precisely penetrated, which takes time. A mobile electronic warfare aircraft disrupts the enemy's communications immediately. Signal reconnaissance and Ku-band jamming must work together; they are two sides of the same coin.

P.S. For those interested in the technical aspects and testing of these microwave modules, details can be found on Telegram by searching for: rfapex_tech. I'm not providing a direct link to prevent moderators from deleting the post.

[Vladimir_2U]

Quote: rfapex

100 W of pure GaN power in the Ku-band, raised to a height of 100–150 meters, creates such a density of interference that the subscriber terminal “goes blind” not only when hovering directly above the trench, but also from a distance of 1–1.5 km (including due to the side lobes of the diagram).

Regarding artillery, a buried and camouflaged terminal must first be exposed by reconnaissance and precisely penetrated, which takes time. A mobile electronic warfare aircraft disrupts the enemy's communications immediately. Signal reconnaissance and Ku-band jamming must work together; they are two sides of the same coin.

Hmm, what side lobes if the antenna is in a trench? That's exactly it, the trench would force the antenna installer to hover literally above the antenna, and there's no way to expect a 1-1,5 km range at an altitude of just 100-150 meters—it's pure geometry, after all. Starlinks aren't on the very front lines, so to speak, but rather somewhat further back, and therefore the installer, forced to hover over enemy territory, is an easy target.
Well, it's not a mine, but a FPV-kha—it's not crucial. Search and interference, in my opinion, is more promising than jamming in a small area with a high risk of loss. However, as a very niche solution, it could be useful.

[scientist]

Of course, we need to do some calculations, but I believe the most effective way to jam Starlink would be from space. Ideally, an electronic warfare satellite would be placed in geostationary orbit with a highly directional antenna covering a large area. Then, the jamming signal is guaranteed to fall into the main beam of the ground terminal's antenna, even if it's in a trench. The satellite itself could be disguised as a TV repeater. Intercepting it in geostationary orbit is impossible; there are no such missiles. In addition to jamming Starlink, several such satellites could be used for passive radar to detect and locate enemy communications equipment and satellite terminals. At long distances, the angle-difference-rangefinding method of passive location can determine the coordinates of any transmitter with sufficient accuracy to be destroyed by conventional means.

[Sensor]

Ideally, an electronic warfare satellite would be placed in geostationary orbit with a highly directional antenna covering a large area.

A highly directional antenna concentrates the radiated power over a small area and
Conversely, a broadly directional antenna produces a low power density relative to a highly directional antenna at the same distance. The effect is similar on reception.

[Cube123]

Quote: scientist

Ideally, an electronic warfare satellite would be placed in geostationary orbit with a highly directional antenna covering a large area.

It won't work. Firstly, the great distance from the geostationary satellite will require colossal power for its transmitter and power systems, energy for which is unavailable even at geostationary orbit. Secondly, this great distance will lead to significant signal delays, making it impossible to integrate into the communication protocol. Even simply tracking the current Doppler is an insoluble problem. Overall, in my opinion, this is a dead end.

[Alexey RA]

Quote: Cube123

Firstly, the large distance from the geostationary satellite will require colossal power from its transmitter and power supply systems, energy for which cannot be obtained from the geostationary satellite.

It's possible to extract energy... the problem is that with such a "legendary" power plant, we need to guarantee 100% satellite placement in geostationary orbit. And not into the Pacific Ocean or an unintended low orbit, followed by destruction in the dense layers of the atmosphere.

[scientist]

Quote: Cube123

Firstly, the large distance from the satellite

If you're talking about power dissipation, a narrow-beam phased array in the 10-12 GHz band will be quite large. I'd estimate it's 2-3 times larger than those of the Starlink satellites in low orbit. There are no other losses, such as atmospheric ones, in space, and this doesn't depend on distance.

Quote: Cube123

Secondly, this same large distance will lead to large signal delays, which will not allow integration into the exchange protocol

It's impossible to integrate into the "client-in-the-middle" exchange protocol even from a short distance; the signal is encrypted there, most likely by a guaranteed-strength method, and decrypting it is practically useless.
But for electronic warfare systems, this isn't necessary. Their standard task is to determine the coordinates of a ground terminal and jam it if necessary. Incidentally, with phase synchronization between satellites, there's a very interesting method of response-pulse jamming, which diverts speed and direction. But this requires experimentation, of course. Clearly, simple barrage jamming will be ineffective and too energy-intensive. But the more complex the data exchange system—and Starlink's is quite complex—the more vulnerabilities it will contain. I doubt that even the high data redundancy in the encryption system will allow for a serious countermeasure to such jamming methods.

[Cube123]

Quote: scientist

If you are talking about power dissipation, then in the 10 - 12 GHz range, a narrow-beam phased array will have larger dimensions.

The distance from a Starlink satellite to the receiving terminal is one hundred times smaller than from the receiving terminal to a geostationary satellite. Since signal attenuation is proportional to the square of the distance, with the same antenna, the interference power at a geostationary satellite should be more than ten thousand times greater, even if the interfering satellite falls within the maximum lobe of the receiving antenna, which is highly unlikely.

[interest]

Plus, beam divergence hasn't been cancelled, even with a phased array - to get the desired result, its size at this distance should probably be > hundreds of meters (if there are specialists, they'll correct me, maybe I even underestimated it).

[doc_i]

If this is effective in suppressing permanently installed Starlink terminals, how can we combat drones equipped with terminals?

[skeptic]

Quote from doc_i

If this is effective in suppressing permanently installed Starlink terminals, how can we combat drones equipped with terminals?

The conclusion suggests itself - a "terminal finder" is flying, with an explosive device... it finds a terminal (stationary or mobile), uses the beam, drops the "nullifier", and flies for the next charge.
What could be a better "suppressor" than a precisely dropped munition?

[doc_i]

Perhaps. Only direction finding would need to be done from above. Again, the question of balloons. And about the energy for their electronics.

[dragon772]

Could it be simpler? The RTR detects the antenna and then an FPV drone or projectile flies towards it?

[S. Viktorovich]

At first glance, it seems like a reasonable option. Is it widely used? Is it difficult to find amplifiers?

[rfapex]

In fact, mass production of such devices has long been established, and purchasing microwave components is no longer a problem. The market has long since moved away from the concept of "bare boards" and home-made assembly toward ready-made, rugged, industrial-grade monoblocks.

A typical example of a modern serial solution for tasks in the Ku-band (13,75–14,5 GHz) is a 250 W class solid-state power amplifier with a built-in boost converter (BUC, IF input 950–1700 MHz) and an integrated 220 V power supply.

From an engineering perspective, everything is already designed in hardware: the device weighs about 14 kg and is housed in a sealed IP66 enclosure with powerful forced-air cooling. Importantly for field conditions, the automation can withstand a severe VSWR of up to 10:1 across all load phases and has a thermal cutoff of 85°C. This means it's a completely self-contained unit: just apply the intermediate frequency and power, and everything works.

Full technical datasheets (PDF) and lab measurement graphs for such microwave devices are usually posted in specialized engineering communities. For example, you can find them on Telegram by searching for rfapex_tech. There's a lot of clear technical documentation on this topic there.

[Svetlana]

The plethora of scientific jargon and photos in the text are intended to justify the false premise promoted in this article—that Starlink ground terminals should be jammed. I believe that the receiver paths of the Starlink satellites themselves should be jammed. At any given moment, there is, on average, one, at most two, Starlink satellites over Ukraine. At any given moment, there are several thousand Starlink ground terminals in Ukraine. And all of these terminals are, at any given moment, focused on one or two satellites over Ukraine. Therefore, it is easier to jam the receiver path of a satellite currently over Ukraine than to jam thousands of Starlink ground terminals in unknown locations. The exact coordinates and time of the next Starlink satellite's appearance over Ukraine are known from astronomical observations. A gigahertz maser with a power of several kilowatts could easily damage the low-noise amplifier (LNA) of the Starlink satellite currently flying over Ukraine. The radiation from this maser could be directed at the satellite flying over Ukraine using an active phased array antenna.

[S. Viktorovich]

A gigahertz maser up his ass, and everything will be fine. From now on, it's just TV commercials.

[faterdom]

Jamming is one of the solutions, and it slowly starts working.
But detecting the ascending beam from orbit and providing its stationary or mobile coordinates is what is really needed.
For a stationary terminal to receive its Krasnopol or Geran, a mobile one - an interceptor, air or sea, respectively.
This requires electronic scanning of the desired area from orbits low enough to detect and identify the signal, but high enough to increase the probability of capturing the scattering cone. This could even require different constellations of satellites linked together (and to the data center), say, by laser links.

[Anatoly Eliseev]

They wasted space! But Ragozin learned to draw Khokhloma!

[acetophenone]

Quote: skeptic

Quote from doc_i

If this is effective in suppressing permanently installed Starlink terminals, how can we combat drones equipped with terminals?

The conclusion suggests itself - a "terminal finder" is flying, with an explosive device... it finds a terminal (stationary or mobile), uses the beam, drops the "nullifier", and flies for the next charge.
What could be a better "suppressor" than a precisely dropped munition?

Brilliant! And next to the terminal, there's a soldier with a machine gun, turning the "seeker/bomber" into a colander. And just in case: how much does a drone like that cost, and how much does the terminal cost?

[DO]

The terminal antenna has a narrow beam pattern of about 1,3°.

The author, in order to jam a satellite terminal's receiver with a low-power interference from a single drone, this drone-jammer must fall into a narrow cone (1,3 degrees) of the terminal's receiver's radiation pattern, which also tracks the flying satellite by changing its angle of inclination (it is, of course, possible to jam the terminal's receiver by hitting the parasitic side lobes of the terminal's antenna; however, this requires at least tens of times greater jamming power, and these side lobes must also be hit).
And even if a drone jammer manages to hit the APAA antenna pattern of the terminal and jam the signal received by the terminal, the terminal can switch to another satellite at a different solid angle to the antenna.