Very low level air defence against flying drones

I've pointed at the issue of aerial drones as a battlefield air defence challenge that cannot be met with traditional dedicated battlefield air defence hardware. You simply cannot defend against 5,000...50,000 € drones by firing expensive missiles at them, and the handful of self-propelled anti-air gun crews cannot really defend other troops against drones that approach at less than 10 m altitude.

Small arms have to be the (metaphorical) last line of defence, particularly against the smallest drones - but small arms have a very poor probability of hit against drones a few hundred metres away if the user can spot and identify a drone at such a distance at all. This isn't trivial at night, for example.

Radio jammers have been developed, deployed and used to disrupt the radio command link and/or satellite navigation of flying drones.* A properly designed reconnaissance drone that encounters this would use its inertial navigation capability** to return to a pre-designated landing point, usually this would be the take-off location. The drone user would likely regain control before the drone would arrive. Well-designed autonomous killer drones would be most unimpressed by radio jamming.

Air defences that depend on acoustic sensors for alerting would be incapable of detecting gliding drones or drones that fly like an owl; silently.

Daylight and UV cameras would not suffice at night time.

Radars would have difficulty telling birds from drones that emulate bird behaviour, and the emissions might be treacherous.

Laser radar (LIDAR/LADAR - essentially scanning the sky with a laser) is incapable of identification as well - and I doubt it would be practical for 3D search for tiny objects.

Net projectors tend to be very limited in range and are bulky.

Hard kill weapons may reasonably range from 5.56 mm to 30 mm, and at 30 mm an electronic-timed  cannister round as in the German Puma IFV may be used. The timing would require some information about range, though.

- - - - -

Well, here's a possible standard subsystem for use on military vehicles. It's not totally specialised, so the introduction of such hardware may be realistic:

Think of a remotely-controlled weapon station (RCWS). The installed weapon would ideally be some .338 revolver machinegun with a controllable rate of fire, but realistically it would be an ordinary 7.62x51 mm machinegun (then preferably MG3). The traverse would be 360°, elevation -15°/+90°.
a typical RCWS, this example has a 12.7 mm gun
The RCWS would have two rings of microphones which can be used for sniper detection, can feed info in the radio network for triangulation of artillery and mortar fires, detection of helicopters and the microphones could also detect the typical noises of drones. They wouldn't be 100% reliable, but most drones might be noticed at useful distances.
There would also be a ring of four wide field of view uncooled infrared sensors, and one coaxial uncooled infrared sensor with zoom. The ring would provide a permanent all-round stare to detect threats and to provide the vehicle crew with all-round day/night vision. They could detect muzzle flashes and patterns/movement of ground targets as well. These staring sensors should detect drones at all relevant distances; drones should not be able to identify a stationary vehicle before its RCWS sensors detect them. The sensor range may thus be greater than the weapon range (a correct warning is very useful in itself).
The coaxial sensor would be used for identification. Its effective range should allow identification of targets (all kinds) at the effective range of the gun against those targets. A coaxial laser rangefinder could provide range information on the target and greatly improve the weapon's accuracy. It could also be used as an interrogator device of an identification friend or foe system in which the lased target may respond with coded radio message if it's friendly. Such a laser could - if the wavelength fits - be used for simulation purposes in training systems like MILES and AGDUS.

An alternative setup would use a rotating scanning uncooled IR sensor that switches to coaxial mode once a target is found. This approach was used by dedicated air search sensors like AD/AD, FIRST and also a couple less well-known systems (including a Swedish and a French one, but I keep forgetting their designations).

The all-round staring sensor setup could also be used for an active protection system; or an active protection system's staring sensors could provide the air search for a RCWS that has only a coaxial sensor. 

Either way, having enough infrared air search is going to be expensive; ten thousands of Euros per vehicle including necessary spare parts. The whole package could easily cost 200,000 € even with competent project management and a large production run. It would cost more if active protection system launchers are included as well and this does not include a software-defined radio to fully exploit the potential by networking the stations and integrating them into the formation's air defence and arty/mortar detection.

Now think of thousands of such RCWS deployed with an army corps, installed on combat and non-combat vehicles. Low altitude drones would have a much harder time. Autonomous killer drones could still overwhelm defences if they attack as a saturating swarm, of course. The only realistic defence against a swarm of autonomous killer drones may be another swarm of autonomous killers drones.

Autonomous killer drones are  quite a challenge, but currently there's little indication that the armed bureaucracies could cope with the much less difficult well-designed recce drones. There are hardly any programs for developing or buying equipment with the recce drone challenge in mind (known to me). The directional radio jammers won't help against well-designed drones and they are apparently never integrated with proper sensors.


*: Many companies came up with simple directional (mostly Yagi-type) antennas installed on something rifle-like to jam the radio bands relevant to commercial drones and pretend that they are oh-so great. Such technology is worthy of a 1930's electrical engineering student's homework.
**: Accelerometers are cheap nowadays, that's why there's a  the quadcopter boom; quadcopters require accelerometers for stabilisation.


  1. Hi Sven:
    My email is down for a while. I'm Israel, so I can respect your rule of sticking to a nickname.
    Anyways. I think your idea of using one of these remote firing stations could be coupled with a fairly simple and somehow available weapon: an automatic shotgun.

    Instead of a normal machine gun, couple a Saiga-style automatic shotgun and increase the chances of bringing down the drone with heavy buckshot. One of these stations, with enough ammo, could send a wall of buckshot directly to the threat. The buckshot used could be standarized, so infantryman equiped with shotguns could carry the same type of ammo, or both kinds of ammo to be compatible.

    Thanks Sven,


    1. The effective range of shotguns may suffice to protect the own vehicle, but it does little else. A RCWS machinegun could be expected to hit at about 500...1,000 m and is very deadly at 10...500 m, depending on circumstances. Shotguns are not really effective past 100 m, not even with flechettes.