Hornet Radar and Helos

[Cathnan]

In the last big patch the Hornet radar gained the ability to lock helos in ACM modes. It's been long standing now, that it can't detect Helos going below 100kts in RWS or TWS. No matter the settings. I always thought this was a bug, but it's been so long that I wonder, could this be how it's meant to be? Can anyone clarify if this is a bug? Or is this how it's supposed to be and they simply get filtered out as clutter.

[Cepheid]

I can't tell if the current state reflects real life. At least locking in ACM works now and opens up radar gun sight. But only when there is some doppler effect from the helo. For that the helo has to move somewhat fast in relation to ground clutter.

I tried approaching various halos in a test scenario from below so that I have Helo against the sky. ACM radar lock was only possible at ranges <5nm rendering AMRAAMs pretty much useless. I tried shooting Mad Dog AMRAAM at helo but that too is not reliable, sometimes it hits, sometimes it does not. If the Helo stops horizontal movement and just hovers, pointing its nose at you....no chance for any radar lock and very dangerous for you.

I also tried some crazy stuff like diving down on the helo for a gun kill. Then AI sometimes pulls up the nose trying to shoot you with it's gun.

Even 9X struggle with the low heat signature.

[amalahama]

The best way to shoot down a Helo is using a GBU-10

[Dragon1-1]

Helos are very dangerous to fixed wing, provided fixed wing is willing to get down to their level. So don't. Of course, if the radar doesn't see the helo otherwise, there's not much else you can do. A helicopter doesn't have much of a heat signature (particularly since MANPADS are typically IR guided, and they're a big threat to a helo, so a lot goes into countering heaters), and while the return from the rotor blades is very distinctive, it's also not very big. Fuselage often moves so slowly it hits the doppler notch by virtue of barely moving with regards to the ground, as far as the fighter is concerned. Plus, even when it's moving, it's not very big, either, particularly from front and rear aspect. 

I'm not sure how big the doppler gate is in the Hornet, but 100kts sounds like a likely value. 

[falcon_120]

This is not how it works in real life. Aircraft radars are able to detect a hovering helo thanks to the distinctive signature the moving blades provides.

Some months ago I watched a spanish interview in youtube of a spanish hornet pilot telling how they are able to detect wind turbines with their Apg65 and lock onto them.

Another RL sample is during desert storm when a F15E picked a mi8 on boarding troops and munitions with its radar and slave it to the tgp to kill it with a gbu 10. FwIW I read this long ago so this last one could be not true.

Enviado desde mi ELE-L29 mediante Tapatalk

[Dragon1-1]

10 hours ago, falcon_120 said:

This is not how it works in real life. Aircraft radars are able to detect a hovering helo thanks to the distinctive signature the moving blades provides.

Distinctive signature means little if there's not enough radar waves bouncing off the blades. It's impossible to hide or otherwise disguise the signature once it's in detection range, but remember that rotor blades are relatively thin strips of metal, definitely much less than a wind turbine (have you seen them? They're gigantic, no wonder they show up on AA radar). The actual area of the blades, and therefore overall RCS, is not going to be very large, and even if the software knows to specifically classify a helo-like return as a contact (which is probably why the turbines are lockable, the radar thinks they're giant helicopters), the question is how far out such a return can be seen.

I don't know the details on the F-15E incident, but if it was employing bombs, it might have been in air to ground radar mode, which does its filtering differently. It's possible a helo would show up as a weird spot that would lead the crew to slave the TGP to it just to see what it is. Notice that for this story to work, the Mudhen would've had to drop the bomb rather soon after finding the helo, so it couldn't have been very far away. Given that in US jets (including the Mudhen as per RAZBAM's videos, at least), AA radar is a B-scope, translating an AA contact into a position on the ground would be tricky, especially since the Mudhen doesn't quite have the level of MSI the Hornet does. So I'd bet on it being detected by eye in AG mode.

[falcon_120]

Yes, my point was, they can be detected even when hovering, how far away is a different topic I agree. But current implementation in DCS is like, speed below certain threshold makes the helicopter invisible in RWS/TWS even if you are 5nm away. You can lock it in ACM though what is not very consistent i think but at least is something.

Enviado desde mi ELE-L29 mediante Tapatalk

[robgraham]

Just something to remember, even a stationary helicopter isn't stationary to a radar.. Those blades while small do have a massive doplar return, they are never not moving, modern filters since the 70's actually look in part for the doplar effect of rotors for that reason.

Same goes for things like turbine blades, Props etc.. they all are actually very distinct radar return items.. despite being what you might think of as 'small' you can get the math fairly easy it's all public. 

[toilet2000]

I had a talk with Razbam's and their radar SME (which I assume is Galinette) told me through their CM that while the Doppler shift is great, the return in itself is pretty small. The radar does not only need a proper Doppler shift, but also a proper signal-to-noise ratio, which the very small heli blades do not provide according to the SME. So even though the signature is recognizable, the range at which it has a signal significant enough to be detected is quite low.

I myself don't have any experience in that regard, but I was curious and asked around. Just a thing to consider.

Edited August 5, 2022 by toilet2000

[Dragon1-1]

My point exactly. The doppler return from a strip of metal spinning in the air is distinctive, but it is not massive. It's proportionate to the strip's size, which in case of a helo blade is dinky compared to the size of the target the radar is normally looking for. So you're going to see it eventually, but not from far away.

[Cathnan]

This still leaves me wondering why I still don't see it in tws/rws despite my radar antenna being within gun range

Sent from my Pixel 2 XL using Tapatalk

[LetMePickThat]

On 7/31/2022 at 12:26 AM, Dragon1-1 said:

Distinctive signature means little if there's not enough radar waves bouncing off the blades. It's impossible to hide or otherwise disguise the signature once it's in detection range, but remember that rotor blades are relatively thin strips of metal, definitely much less than a wind turbine

Most NCTR systems use blade count to classify targets, when there's a signature ambiguity. It's doubtful that the main rotor on an helo wouldn't reflect enough energy to be detected when modern radar sets can distinguish a MiG-35 from a MiG-29 using the difference in blade signature between the RD-33 and the RD-33MK...

[BIGNEWY]

We had a fix in July for ACM mode and helo's, it is working as intended currently, slow helo's are difficult to lock. As far as I am aware anything under 100knts will be difficult to detect and lock. 

[silverdevil]

approved for public release

https://apps.dtic.mil/sti/pdfs/ADA351581.pdf

[Dragon1-1]

1 hour ago, LetMePickThat said:

Most NCTR systems use blade count to classify targets, when there's a signature ambiguity. It's doubtful that the main rotor on an helo wouldn't reflect enough energy to be detected when modern radar sets can distinguish a MiG-35 from a MiG-29 using the difference in blade signature between the RD-33 and the RD-33MK...

You're conflating two things: detection and identification. A MiG-29 has a normal-sized radar signature, the detection algorithm does not rely on return from turbine blades alone. After target is detect and SST lock acquired (NCTR always requires STT), the radar, now tracking only the MiG, can identify it by turbine blade returns, assuming range is close enough. It now knows it's looking at a plane, and it can focus all its energy on it, greatly increasing signal to noise ratio (which is why being unable to maintain lock past detection range in DCS is bogus, once locked, tracking range can increase greatly). A radar in search mode needs to detect the target during the instant in which it is under a small part of the radar beam, meaning signal to noise ratio for detection is much lower. Sure enough, a helo will, once detected, be immediately identifiable with NCTR, from any aspect (unlike a fixed wing aircraft, which has to be hot so the radar can see the turbine blades).

RD-33 and RD-33MK are basically two different engines, and as such, NCTR can probably see the difference in front fan design. It cannot tell between two engines if the fan is the same, because that's what it sees. So if the front fan has changed between RD-33 and RD-33MK, it can tell a MiG-29 from a MiG-35 just like it can tell Su-27 from either. If it hasn't, it'll report both as MiG-29. TBH, I don't know whether that's the case or not with this particular engine.

Edited August 8, 2022 by Dragon1-1

[LetMePickThat]

29 minutes ago, Dragon1-1 said:

You're conflating two things: detection and identification. A MiG-29 has a normal-sized radar signature, the detection algorithm does not rely on return from turbine blades alone. After target is detect and SST lock acquired (NCTR always requires STT), the radar, now tracking only the MiG, can identify it by turbine blade returns, assuming range is close enough. It now knows it's looking at a plane, and it can focus all its energy on it, greatly increasing signal to noise ratio (which is why being unable to maintain lock past detection range in DCS is bogus, once locked, tracking range can increase greatly). A radar in search mode needs to detect the target during the instant in which it is under a small part of the radar beam, meaning signal to noise ratio for detection is much lower. Sure enough, a helo will, once detected, be immediately identifiable with NCTR, from any aspect (unlike a fixed wing aircraft, which has to be hot so the radar can see the turbine blades).

RD-33 and RD-33MK are basically two different engines, and as such, NCTR can probably see the difference in front fan design. It cannot tell between two engines if the fan is the same, because that's what it sees. So if the front fan has changed between RD-33 and RD-33MK, it can tell a MiG-29 from a MiG-35 just like it can tell Su-27 from either. If it hasn't, it'll report both as MiG-29. TBH, I don't know whether that's the case or not with this particular engine.

 

I'm not conflating detection and identification. Even when knowing that a plane is here, NTCR would need to have a return from the blades to perform the blade count and subsequent identification. Assuming that the blades produced no return (or too small of a return), NTCR wouldn't be a thing. Sure, you'd be focusing more energy on the threat and thus have increased resolution and range, but the main rotor of your average helo has more surface and is more exposed than a compressor first stage, and should produce more returns for a given power output.

Also, NCTR doesn't always require STT, TWS and equivalent also work on modern AESA radars because of the ability to use multiples beams. Thales for instance publicly stated at Le Bourget that the RBE2 could perform multiple identifications at the same time. There are also various NCTR techniques that do not require the target to show compressor blades (i.e have an hot aspect), like narrow-beam interleaved search and track or other pseudo-imaging approaches.

[Dragon1-1]

42 minutes ago, LetMePickThat said:

I'm not conflating detection and identification. Even when knowing that a plane is here, NTCR would need to have a return from the blades to perform the blade count and subsequent identification. Assuming that the blades produced no return (or too small of a return), NTCR wouldn't be a thing. Sure, you'd be focusing more energy on the threat and thus have increased resolution and range, but the main rotor of your average helo has more surface and is more exposed than a compressor first stage, and should produce more returns for a given power output.

Remember that you need to account for physical surface area of the blades, not the entire rotor disk. Depending on a helo, this may actually be less than a nearly solid disk of a jet turbine (and may be made of composites and not of very reflective metal). Also, consider the orientation, too. In aspects at which NCTR is a thing, the radar is facing the turbine blades head-on. In contrast, in a co-altitude scenario, you're looking at the rotor disk edge-on. Not much surface area to be found there. If you're above, you're looking at the blades at an oblique angle which, as you might know from elementary optics, does not reflect the radar waves in the direction of the emitter. It does scatter them, so you do get some return, and the tail rotor might or might not be at a more convenient angle (if it has one, of course), but ultimately, radar returns from helicopter blades are weak and will get lost in the noise at ranges a fixed wing aircraft would be easily detectable.

Also, as a matter of fact, turbine blades do produce a rather small return. This means, among other things, that NCTR does not work at maximum detection range. This is even simulated in DCS, with fighters you don't get a print until you're at 20nm or so. I think you're underestimating the degree to which STT focuses the beam versus a regular scan and the S/N ratio advantage it brings. Basically, it has all the time in the world to count the blades, while a helicopter in search mode is illuminated for a fraction of second each scan, and, crucially, there's no correlation of marginal returns between scans. STT versus search mode is a massive difference.

Quote

Also, NCTR doesn't always require STT, TWS and equivalent also work on modern AESA radars because of the ability to use multiples beams.

AESA is a whole other story, and we're not talking about AESA because our Hornet doesn't have it. These things can do a lot of stuff that the older radars couldn't even dream of. Including STT-like continuous tracking of multiple targets at the same time, which of course enables NCTR.

Edited August 8, 2022 by Dragon1-1

[amalahama]

21 hours ago, BIGNEWY said:

We had a fix in July for ACM mode and helo's, it is working as intended currently, slow helo's are difficult to lock. As far as I am aware anything under 100knts will be difficult to detect and lock. 

An helicopter should be difficult to detect and lock, but not because of the speed, but for the relatively low radar signature. As it has been stated several times in this thread, rotor blades returns a very distinguish signature easy to pick up by Doppler radars, it should be independent of general Helo speed

[BIGNEWY]

3 minutes ago, amalahama said:

An helicopter should be difficult to detect and lock, but not because of the speed, but for the relatively low radar signature. As it has been stated several times in this thread, rotor blades returns a very distinguish signature easy to pick up by Doppler radars, it should be independent of general Helo speed

If you have public information showing this I will be happy to show it to the team. 

thanks

[escaner]

On 8/8/2022 at 2:57 PM, BIGNEWY said:

We had a fix in July for ACM mode and helo's, it is working as intended currently, slow helo's are difficult to lock. As far as I am aware anything under 100knts will be difficult to detect and lock. 

What you say is true when the helicopter (or any other target aircraft) is flying below your own Hornet due to the ground filter in the doppler radar. That filter is disabled when the target is above and the problem should disappear, yet in DCS is still impossible to detect helicopters flying above. And this is true for any doppler radar, not only the F-18's. I have not found this problem detecting helicopters in any other module.

[Yaga]

On 8/9/2022 at 5:37 AM, BIGNEWY said:

If you have public information showing this I will be happy to show it to the team. 

thanks

I get that "easy to pick up" is unquantifiable, but this is a pretty simple assertion and it's strange that you'd dismiss it, in its entirety, so quickly. It's also hard to source something proving that the radar "works" or "doesn't work". DCS isn't an EM simulator and no one needs it to be one. I understand the desire to create accurate organic outcomes and I do think conceptually this is a good ideal. However...

It has to work well. Simulated outcomes are great since you can't script an outcome for every situation conceivable. But if the system in place doesn't even generate reliable outcomes for situations that are common enough to be scripted, then you really haven't made any improvement at all. In fact, maybe the ecosystem has become so convoluted at this point with its myriad of coefs and formulated relationships that it can't even be adjusted easily anymore. 

When I use the radar in the F-18 or the F-16, I find it to be an absolute unmitigated nightmare of inconsistent and unreliable results, capabilities, and behaviors. In fact, the topic of detection ability only scratches the surface of the enormity of all the problems with processing logic, the general avionics presentation, workflow and just about every other area one must tediously slog through and overcome in order to use the radar as a radar. So when someone asserts something should be "easy to detect", it's hard to know where to even begin when the entire sum of environmental and systems modeling is working against them in nonsensical and illogical ways. 

Generally speaking, and almost certainly for our purposes, Fire control radars work. Modern weapons track. If they didn't, we'd develop ones that did. In fact, that's exactly what occurred over the last 60 or so years. 

If the deep simulation approach is to be continued, then you can't cut corners. Abstracting relationships at any level is ultimately abstraction, and any of that begs the question "after all this, aren't we still just scripting the outcome with so many extra steps?" 

Here's some citations for some research regarding helicopters and signal processing. The amount of existing material on the subject and related subjects is quite extensive, so there's plenty to read. I trust, given the scope of the project, that you have access to an academic DB that will have these journals: 
 

Quote

Setlur, Pawan et al. “Helicopter radar return analysis: Estimation and blade number selection.” Signal Process. 91 (2011): 1409-1424.

Zhang, Rui et al. “Multi-Aspect Micro-Doppler Signatures for Attitude-Independent L/N Quotient Estimation and its Application to Helicopter Classification.” Iet Radar Sonar and Navigation 11 (2017): 701-708.

Yoon, Sang-Ho et al. “Helicopter classification using time-frequency analysis.” Electronics Letters 36 (2000): 1871-1872.

Yoo, Kyungwoo et al. “Rotating Target Size and Rotation Vector Estimation in a Distributed Stepped Frequency Radar System.” IEEE Sensors Journal 20 (2020): 7189-7198.

Habermann, Danilo et al. “Discerning similar helicopter patterns by low PRF ground surveillance radar.” 2019 IEEE Radar Conference (RadarConf) (2019): 1-6.

Zhang, Rui et al. “Multi-Aspect Micro-Doppler Signatures for Attitude-Independent L/N Quotient Estimation and its Application to Helicopter Classification.” Iet Radar Sonar and Navigation 11 (2017): 701-708.

Point, Guillaume et al. “Parametric modelling of the radar signature of helicopters.” 2019 International Radar Conference (RADAR) (2019): 1-6.

Thank you, 

[BIGNEWY]

3 hours ago, Yaga said:

I get that "easy to pick up" is unquantifiable, but this is a pretty simple assertion and it's strange that you'd dismiss it

But I am not dismissing it, I am asking for evidence that I can take to the team. 

[Yaga]

4 hours ago, BIGNEWY said:

But I am not dismissing it, I am asking for evidence that I can take to the team. 

You're asking for evidence of what exactly? The team needs confirmation that rotorcraft have unique radar signature characteristics compared to fixed wing aircraft? We can definitely find such validation, but we could probably just show them a picture of each type of aircraft and have the same effect. 

If we do go and confirm that yes, there are unique attributes to Helicopter RCS in the real world, then what? It isn't relevant at all within DCS. That data doesn't exist to drive any simulated outcomes. The functions that would conceivably carry and use this data don't exist either. Just like the RCS characteristics attributable to pylons and stores in the real world, they don't factor in. This has.. some effect on accuracy across all the results it generates. 

Here's two works proposing methods for modeling helicopter radar signatures accurately, validated against real world measurements. Radar signature characteristics, unique to helicopters, are briefly covered in both:
https://www.researchgate.net/publication/350770098_Modelling_the_radar_signature_of_rotorcraft
https://www.researchgate.net/publication/336073108_Parametric_modelling_of_the_radar_signature_of_helicopters

This aren't usable methods here, I don't think. But it is the requested evidence showing that helicopters are not airplanes. I suppose that's something. 

Edited August 11, 2022 by Yaga

[AngelAtTheTomb]

Yaga, as an academic (and a nerd) I certainly appreciate the sources you're finding. Alas, I suspect BN is referring to something more precise - a document showing "F/A-18C Lot 20 FCR detects helicopters at X range". ED doesn't seem to have the appetite or the budget for a full radar simulation, so they'd need something they can build a radar script around, instead of "this is how radar works".

 

You're right that the radars are becoming a bit scattershot across all the modules. Some - M2000C, making everyone else look bad - are approaching a quasi-simulation. Some - F15C - are magic. The F/A-18's radar is a troublesome animal, and in my experience it performs well 90% of the time. I don't know enough about the real aircraft to know whether the other 10% is a simulator shortcoming or a real-world shortcoming.

 

I do find it hard to believe that helicopters would be essentially invisible to a modern FCR.

[BIGNEWY]

I am happy to bring anything to the teams attention, but yes as AngleAtTheTomb suggests more precise public information is best if possible.