Uninhabited Aerial Vehicles and loitering munitions are increasingly being used in the suppression/destruction of enemy air defence battle.
Air launched anti-radiation missiles with a loitering capability were deployed by the Royal Air Force during the Cold War, but the utility and availability of lower cost loitering munitions is making them increasingly attractive for the contemporary S/DEAD fight. They are hard to detect by radar and, crucially, they do not expose human aircrews to the danger of flying within the engagement zones of a surface-to-air missile or anti-aircraft artillery systems.
UAVs can gather imagery intelligence on ground-based air defences, and they can collect electronic or communications intelligence on the radars and communications an integrated air defence system relies upon. Loitering munitions can stay aloft for prolonged periods outside these engagement zones – once a radar activates, they can detect the radar’s signals and use these to guide towards the antenna and then explode.
Tactically, UAVs and loitering munitions have advantages as S/DEAD weapons as they may fly relatively slowly. The Russian Army’s Orlan-10 UAV has a maximum speed of 81 knots (150 kilometres-per-hour), well below the 800 knots of a Lockheed Martin F-16 fighter. This matters because these UAV speeds may fall out of the ‘velocity gate’ used by ground-based air surveillance radars.
A velocity gate is a programmable threshold used by a radar, and helps to reduce spurious targets or ‘clutter’ for the radar operator. For example, a radar maybe programmed to detect and track all targets with a speed of between 150 knots and 1,934 knots, so it would detect targets like helicopters at the lower end of this speed threshold, or anti-radiation missiles at the higher end.
But other targets such as like birds which move at much slower speeds, or even swarms of insects would be ignored. This is to avoid the radar and its operator becoming inundated with targets to detect and track, many of which may be spurious.
The problem posed by a UAV is that its velocity may be below a radar’s velocity gate, causing it to be ignored.
Another problem concerns Radar Cross Sections (RCS) which denotes how large the target appears to the radar. As with the velocity gate, radars are programmed to look for specific targets with specific RCSs which is also to avoid problems caused by false alarms.
Every airborne target will have an RCS, but these vary greatly in size which can be caused by its actual size, the materials it is constructed from, or its overall shape. Aircraft like the Lockheed Martin F-35 use angles in their construction which help scatter incoming radar signals away from the aircraft and the transmitting radar antenna, while non-metallic composite materials used in the aircraft’s skin help to absorb incoming Radio Frequency (RF) energy.
RF is the ‘stuff’ which forms a radar signal. As metal is electrically conductive, it can help to bounce the incoming radar signals back to the transmitting radar, while composite materials do not contain metal which also helps to reduce the aircraft’s RCS.
A question of power
A McDonnell Douglas F-4 Phantom-II fighter has an RCS of about six square metres. RCS is routinely measured in decibels-per-square milliwatt (dBm) with the Phantom having a 7.8dBm RCS.
To understand how RCS works, let’s pit this F-4 against an antiquated FuMG-62D Würzburg fire control/ground-controlled interception radar which was used extensively by the Luftwaffe) during the World War II. We have chosen the FuMG-62D because it is one of the few military radars where all its performance parameters are in the public domain – for understandable reasons, radar manufacturers and militaries like to keep many of these statistics classified. Such information can allow enemies to divine what techniques they need to use to jam these radars.
The FuMG-62D could transmit radar signals with power levels of seven kilowatts/kW or 11 kW. We need to ascertain the Effective Radiated Power (ERP) of the radar which combines the strength of the signal the radar can push out and the gain of the antenna. The gain is how much of the signal the radar can point towards the target, which is a function of antenna design, among other factors. We determine this by adding the radar’s incoming power (3.3kW/65.1dB) to the 102.7dB antenna gain which gives us an ERP of 167.8dB.
Radar signals lose strength the further they go. Our radar signal is transmitted from the antenna and zooms through the ether at light speed, and hits the target. The Würzburg radar had a maximum range of 40km and transmitted on a frequency of 560MHz. Let’s suppose that our F-4 is right at the edge of the radar’s range. We need to know the strength of the radar signal when it hits our target.
By the time the signal reaches the target, its strength has diminished to 49dB. Nonetheless, this 49dB is not the strength of the signal that bounces back to the radar because the F-4 has a metal construction which is particularly good at reflecting RF. Therefore, the strength of the signal will increase to 72.7dB at the F-4 when it begins its journey back to the radar antenna. But by the time it returns to the antenna, the signal will have diminished significantly to -46dB.
Therefore, radars must be very sensitive to detect such weak signals. How much of this energy will re-enter the radar through the antenna so that it can be used by the radar operator to determine that a target is there and that target’s range? The gain of the antenna can help here because of the sharpness of radar beam it can develop, which adds to the signal strength. Thus, when we take our -46dB figure, but take account of the antenna gain (102.7db), we get a signal strength of 56.7dB.
Now let’s suppose our Würzburg radar’s target is a much smaller loitering munition. For this demonstration we give our target an equivalent RCS to the 0.05m2 (-13dBm) RCS of Boeing’s AGM-86 cruise missile as an exemplar, only because public information on the RCS of weapons like Israel Aerospace Industries’ Harpy loitering munition is non-existent.
All the other elements will stay the same including the target’s range of 40km and the radar’s 560MHz frequency. The same signal strength, 49dB, reaches the target. This time, we get a signal strength of 51.9 reflected to the radar. As before, this signal moves through the ether, eventually reaching the radar antenna where it has a strength of -66.8dB. How much of this signal strength goes into the radar? It will be 35.9dB, noticeably less than the F-4.
We can see that a smaller target like a loitering munition returns less signal strength to the radar than a larger target like the F-4. Some UAVs or loitering munitions may have such small RCSs that they are discounted by the radar.
If we take a flock of geese as a comparison, one immediately sees where the problem is. Open sources say that a flock of geese can have a radar cross section of between 0.1m2 and 1m2 translating into an RCS of between 38.9dBm and 48.9dBm. The loitering munition falls squarely within this bracket, so a radar operator tuning their set to avoid detecting flocks of birds could risk ignoring a loitering munition.
One way around this problem is to attempt detection using techniques like micro-doppler processing. Radars exploit a phenomenon known as the doppler shift which takes its name from the Austrian physicist Christian Doppler (1803 – 1853).
Prof. Doppler determined that the frequency of a wave depends on the speed of the wave’s source relative to the speed of the observer. Known as the doppler effect, although this may sound complex, we observe it more often than we think. A person standing still may hear a police car approaching them. As the car approaches the pitch of its siren seems to increase but, as the car zooms past and continues onwards, the pitch of the siren seems to fall, but in reality the pitch is unchanged.
All waves – be they sound waves or the radio waves radars rely upon – have peaks and troughs. The frequency of a wave is the measurement of the distance between each peak or trough. As the police car siren approaches the pedestrian, the peaks and troughs take progressively less time to reach their ears. The listener is getting more peaks and troughs per second reaching them as the car approaches. This increase in peaks and troughs translates into an increase in frequency for the pedestrian and is known as the doppler effect. Likewise, this process works in reverse when the police car drives away.
Radars rely on the doppler effect as they indicate when a target is moving relative to the radar’s position. Put simply, if the target is approaching the radar, the frequency of the radar echoes bouncing off the target increase. If the target is moving away, they will decrease.
On its own, this does not solve our problem of detecting a target which may have such a small RCS that it is discounted by the radar, but radars can harness the doppler effect in other ways. UAVs and loitering munitions need some form of propulsion.
UAVs will have spinning rotors, propellers, or small jet engines with spinning fan blades and, as these are moving, they produce additional doppler shifts alongside the aircraft’s momentum. By detecting and computing these shifts, it is possible to determine that the target is a UAV and not a bird. True, birds flap their wings, but the difference with a UAV is that the aircraft’s blades will be constantly in motion to ensure it keeps flying. In contrast, a bird may flap its wings and then glide for a while.
The advent of Cognitive Radar, which harnesses Artificial Intelligence (AI) techniques, may enhance these micro doppler processing techniques still further. It may be possible to alter or upgrade a radar’s software to tune it to detect the micro-doppler shift triggered by a UAV or loitering munition’s spinning blades.
In peacetime, a radar could be calibrated by flying a wide array of different UAVs and inert loitering munitions at various ranges from the antenna. The radar can be ‘trained’ to recognise the micro doppler shift of the spinning blades, compared to the flapping wings of a bird. The radar’s software will then flag the former to the operator while discounting the latter.
By using AI techniques, the radar could recognise micro doppler shifts from other uninhabited aircraft it detects on a day-to-day basis. These signatures may not be identical to the ones the radar has been trained with but would be similar. Over time, this would allow the radar to hone its UAV detection skills. In wartime this would provide the radar with a solid set of skills to detect, recognise and track hostile UAVs and loitering munitions. This will not only help protect blue force air defences against red force S/DEAD capabilities, it would provide UAV detection writ large.
Having radars with the requisite techniques to detect and track UAVs and loitering munitions are useful, but other techniques can aid the detection and tracking of these threats.
UAVs will almost certainly need a radio link connecting the air vehicle to an operator. Civilian UAVs tend to use radio links on frequencies of 2.4GHz and 5.8GHz, and similar frequencies may be used by the UAV to share imagery it is collecting with those who need this intelligence.
Electronic Warfare (EW) systems like Electronic Support Measures (ESMs) may be able to detect radio traffic on these frequencies and may be able to determine the location of the UAV and its pilot. Once this is done, electronic attack in the form of jamming can be directed against the UAV.
With the radio link broken, the UAV may simply land in situ. Alternatively, it may automatically fly back to the place from where it was launched. Determining where this is may provide coordinates for artillery which can engage the aircraft’s location on the ground. The pilot’s location can be treated in a similar fashion.
Most UAVs and loitering munitions will use Global Navigation Satellite System (GNSS) Position, Navigation and Timing (PNT) signals transmitted from constellations like the US Global Positioning System (GPS). These typically transmit signals from satellites to Earth on frequencies of 1.1GHz to 1.6GHz. By transmitting powerful jamming signals on similar frequencies, it may be possible to drown out the PNT signal which this may cause the aircraft to land in situ or return to its point-of-origin.
Some threats like ‘kamikaze’ UAVs may get around this threat by relying on an Inertial Navigation System (INS), a self-contained system using gyroscopes and accelerometers to help the aircraft navigate. This is the case for the Shahed-136 which the Islamic Republic of Iran has supplied to Russia for use against Ukraine.
This underscores why a layered approach to protecting ground-based air defences is needed to counter the use of UAVs and loitering munitions in the S/DEAD battle.
EW systems help detect the signals these aircraft rely upon and to jam them once discovered. They can also help deprive aircraft of PNT signals from GNSS constellations. When no such signals are present, it may become necessary to employ other methods to detect, identify and track these aircraft which is where techniques like micro-doppler processing come in.
UAVs and loitering munitions are now a fact of life in the S/DEAD fight. Fortunately, air defenders are not powerless against these threats.