Does the aim-120 ignore ground clutter?

[Blinky.ben]

23 hours ago, Coxy_99 said:

IR missiles i understand they go for the heat, But why should active missiles ignore ground clutter if your on the ground, Or am i missing something?

Is the aircraft 10,000ft above the ground with the missile looking down? Or is the aircraft at 200ft? Cause then Doppler basically has nothing to do with it. And yes in this case a contact can be lost within the clutter and other things. But it depends on more things then I think DCS could programme. But the lower the contact to the ground with a look down aspect the more influence the ground will have.

[GGTharos]

2 hours ago, Skysurfer said:

Being stationary on the ground should not result in any PD radar lock. Now, if it's a pulse lock (like on the Tomcat) you can get locked and Captain-Jack-Sparrowed.  

 

No, you can't, and there's IRL experience to show for it.

[FoxAlfa]

Ground cutter is underrepresented in DCS, people give too much credit to Pulse Doppler... yes, it can filter some stuff and angles, but if you are low enough or looking at the ground the noise becomes too big to filter anything.

Two videos as proof, both Hornet and Tomcat lose lock on cold target when looking at the ground.
Both radars are much more powerful then 5 in dish and phone size processor unit of the AMRAAM. 

Edited February 20, 2021 by FoxAlfa

[Pikey]

Aircraft stationary on the ground should not be targeted, let alone hit with radar guided air to air missiles. That's all there is to that.

Doppler gate is what 40kts at close range for DCS air targets? This is a vestige of the fact that DCS sees a unit as "something" and not as a  moving object through 3d space.

However I can't speak for helicopter blades because that's a different thing and should be possible.

Still, its all moot because no trackfile.

[Blinky.ben]

I’ve learnt a lot from the community so might be a good time to give back. I thought I might try and clear something up about Doppler. Feel free to tell me to never post something this big again. This isn’t to point out how radars in DCS work but some people seem to compare DCS to real world and demand what is actually false.

 

First thing to make clear is we are talking about Automated Air radars and not ground based radars. Ground based radars are a lot more complex and don’t share a lot of disadvantages that Air based radars do.

 

However the Radars we are exposed to mostly in DCS are fairly simplistic AUTOMATED radars with one antenna. So they can’t as we like to call it “talk and listen” at the same time (PRF). So if the radar is currently sending out a pulse it can not receive any incoming pulses at that time, so a lot of pulses are returning back to the radar are not actually being received and are completely missed or only some of that pulse is received. But my point is here that the radar can miss up to more then 50-60% or less of the pulses but that comes down to timing and distances. But no matter what it’s never registering anywhere near 100% of the pulses. But it is sending out a gigantic amount of pulses per second which is why the odds games says this system works. Most Ground based Sam sites don’t share this issue and normally register most of the pulses cause they work in a network where they can send out pulses and listen at the same time.

 

When the pulse is sent out it will be modulated in such a way the radar can identify one pulse to the other, however some modulations give benefits over the others but I won’t be going into that. We are a long way before Doppler has even started to be processed. Cause we need to deal with the amplitudes and energy coming back. pulse/energy hits something it bounces back. Now a small amount of a pulse can bounce back at a different time before the rest of the same pulse bounces back, some pulses run out of energy and never make it back, this is where PRF and modulation is important because what part of the pulse did the radar get back or are two pulses coming back at the same time but one has a lot more energy then the other pulse but has the weaker pulse detected the jet your trying to find and has now been ignored?. Let’s say a jet is flying 1000ft above ground and your looking down at it. Your radar pulse is going to send almost the entire pulse back at you but with much less power in it due to attenuation. So within that pulse will be hundreds of energy spikes being the ground, rocks, trees, the jet, clouds....... So now the radar needs to start interrogating the pulse to find out what is within the pulses and what needs to be ignored. HOWEVER the radar has to protect itself first. So any pulses that are bouncing back with too much energy will automatically adjust the gain and reject the pulses with too much power in it. So instantly we have lost some pulses or parts of pulses that may or may not have the jet were looking for (KEEP IN MIND DOPPLER Considered yet).

 

Now we have the pulse making it’s way through the system and this is where certain modulations will play into their benefits. With what ever pulse we have left let’s imagine we are looking at the old TV that have no reception and the screen is just full of static. Now we need to find that white dot and track it which is Impossible. So we need to tune the reception more to remove the overwhelming amount of static. The radar is doing the same thing but these are called anti-clutter circuits which there are many types that all benefit in certain areas. Some are for detecting things in weather but some times the clouds are just too dense and the radar ignores those pulses to protect itself or the attenuation is too much. Some are good for ground clutter but again it will set a threshold to the point where the jet we are looking for will remain under that threshold and will always remain hidden cause the radar is ignoring it. It could be going through as many as 5 different clutter circuits. This is why Automated radars have a disadvantage to raw based radars cause the operator can set the threshold as they can still see the contact with the extra noise around. And again how much of the pulse are we getting back that have the jet in it? Once all this has been done then we are still left with a lot of spikes in energy/amplitude that the radar needs to determine is the target we are looking for, but we still don’t have enough information for that to happen (AGAIN DOPPLER/frequency still not considered) so now the radar sends out more and more pulse and does this whole process again and again until it builds a memory bank of the same spikes and NOW WE HAVE DOPPLER so the radar now says I have a lot of spikes above the threshold after removing a lot of other things but this spikes keeps coming back with a different frequency  so it must be moving. So now we get speed gates to remove other moving objects like birds or tress blowing in the wind and so forth. Now the radar looks at a particle part of the sky and decides to track it. But WAIT something got in the way let’s say a flock of birds cause depending on a few things a radar can’t determine between multiple contacts within a certain distance from each other so a flock of birds can mean one big contact, I’m not going into those distances cause I don’t know if this is open knowledge or not. However get this there is also something else that can get in the way CHAFF!!!!! which yes radars have features to help it not to track chaff but we don’t need it to track chaff we just need chaff to introduce many of the problems I have mentioned above NOW FORCING the radar to go through the whole process again cause the radar is now ignoring more and more pulses cause chaff is introducing multiple problems. and now the memory bank starts decreasing. Now it can predict where to look and if it finds the target quick enough it will build up that memory bank quicker but if it looses it for let’s say 3-5seconds or it moves outside the FOV then we have to start all over again. Now this doesn’t mean the radar will never find a contact in these situations but the truth is there is a good chance the contact will remain hidden if it is close enough to the ground below the horizon with the aspect your looking at it.

 

So you can see here DOPPLER is not actually this magical thing that has cured world hunger but simply a tool that’s way down the bottom of the tool box that the radar uses after much complex situations have been dealt with first. But again a pulse may never even get to that point.

 

This is why putting the object above the horizon is sssoooo very important to increase your chances of detection.

 

So with all this in mind we can’t have all this in DCS or maybe we can I know nothing about coding. But to make something like chaff to have more of an effect or look down penalties have more of a influence to account for other disadvantages not within DCS sounds more like a simulated environment I think.

Edited February 20, 2021 by Blinky.ben

[Coxy_99]

9 hours ago, Blinky.ben said:

Is the aircraft 10,000ft above the ground with the missile looking down? Or is the aircraft at 200ft? Cause then Doppler basically has nothing to do with it. And yes in this case a contact can be lost within the clutter and other things. But it depends on more things then I think DCS could programme. But the lower the contact to the ground with a look down aspect the more influence the ground will have.

 

Aircraft 30 to 40nm 35000 feet shoupd not be able to lob radar guided missiles on aircraft parked on the floor (not moving) not at 2000 feet or 10000 feet on the floor not going any where the missile looking dowm into that ground clutter should not track end of, Ive never seen this happen before in DCS ever ive been about a loooong time its happened twice on servers both being bluelag buddyspike. And funny enough track replays beimg bugged for over 2 years i cant get a track because a bug stops me.

[nighthawk2174]

4 hours ago, FoxAlfa said:

Ground cutter is underrepresented in DCS, people give too much credit to Pulse Doppler... yes, it can filter some stuff and angles, but if you are low enough or looking at the ground the noise becomes too big to filter anything.

Two videos as proof, both Hornet and Tomcat lose lock on cold target when looking at the ground.
Both radars are much more powerful then 5 in dish and phone size processor unit of the AMRAAM. 

 

Not sure what's going on with his radar there in the first bit of footage, however is plenty of footage of the same situation without any effects for the F18, and there's also plenty of F16 hudtapes with no effects either.  Also tbf in that F14 tape I was under the impression that when the dimond got centered like that it is a break away cue which is also supported by the fact the velocity and range carrots aren't jumping around.

 

[FoxAlfa]

57 minutes ago, nighthawk2174 said:

Not sure what's going on with his radar there in the first bit of footage, however is plenty of footage of the same situation without any effects for the F18, and there's also plenty of F16 hudtapes with no effects either.  Also tbf in that F14 tape I was under the impression that when the dimond got centered like that it is a break away cue which is also supported by the fact the velocity and range carrots aren't jumping around.

 

57 minutes ago, nighthawk2174 said:

Not sure what's going on with his radar there in the first bit of footage, however is plenty of footage of the same situation without any effects for the F18, and there's also plenty of F16 hudtapes with no effects either

 

 

 

 

Altitude and terrain are the biggest factors, something that would work well over flat terrain won't in hill or mountain area.

The biggest issue here is PRF, simply in one frame radar pickups target correct, next frame picks the ground, the next frame let say target and corrects.... it's a roulette... it can keep picking the target or keep picking the ground and lose the target. Tomcat clip is good for that, you can see target cue jump is it switches between looking the ground and then next resolving the target.

 

Your graph is also excellent for illustration for it, when looking at the ground noise increases dramatically, and signal strength can't go up unlimited due to emiter output power and other factors.

Also note that most of the footage we have is relatively high (5000+) at close target due to safety and even there we can see effects, and based on the manuals we know that radar performance degrades even more sub 5000 ft due to terrain noise. 

Edited February 20, 2021 by FoxAlfa

[Blinky.ben]

52 minutes ago, Coxy_99 said:

 

Aircraft 30 to 40nm 35000 feet shoupd not be able to lob radar guided missiles on aircraft parked on the floor (not moving) not at 2000 feet or 10000 feet on the floor not going any where the missile looking dowm into that ground clutter should not track end of, Ive never seen this happen before in DCS ever ive been about a loooong time its happened twice on servers both being bluelag buddyspike. And funny enough track replays beimg bugged for over 2 years i cant get a track because a bug stops me.

Totally agree

[nighthawk2174]

7 minutes ago, FoxAlfa said:

Altitude and terrain are the biggest factors, something that would work well over flat terrain won't in hill or mountain area.

The biggest issue here is PRF, simply in one frame radar pickups target correct, next frame picks the ground, the next frame let say target and corrects.... it's a roulette... it can keep picking the target or keep picking the ground and lose the target. Tomcat clip is good for that, you can see target cue jump is it switches between looking the ground and then next resolving the target.

 

Your graph is also excellent for illustration for it, when looking at the ground noise increases dramatically, and signal strength can't go up unlimited due to emiter output power and other factors.

Also note that most of the footage we have is relatively high (5000+) at close target due to safety and even there we can see effects, and based on the manuals we know that radar performance degrades even more sub 5000 ft due to terrain noise. 

 

Well I mean the PRF of these radars is going to be in the khz range but sure if the radar is in a conscan pattern here maybe it may not see the target for a split second at dogfight ranges.  But this wouldn't impact a monopulse seeker.  Also imho seems to me in the tomcat clip it may be more related to the very low hud refresh rate and break away cues.

[FoxAlfa]

2 minutes ago, nighthawk2174 said:

break away cues

nopu, the gun piper jumps.... it means it's getting the range wrong, simple as that.

[Shadow KT]

Posting two short random clips without any context shows exactly nothing. No information on electronic environment, radar status (all working as intended). Real life radars on jets are not as bullet proof as the ones in DCS and can have issues at times.... There is a chance that at that moment, the radar saw something that it might interpret as jamming and that's why it lost lock or got bad range for a moment or that it didn't just spaz out.

P.S Not saying that you should be getting hit by AMRAAMs, sitting on the ground from 40nm away. Kind of hilarious it does this, but it struggles to hit targets flying in a straight line

Edited February 20, 2021 by Shadow KT

[FoxAlfa]

17 minutes ago, Shadow KT said:

There is a chance that at that moment, the radar saw something that it might interpret as jamming and that's why it lost lock or got bad range for a moment or that it didn't just spaz out.

Yes, it did see, a huge radar reflector 2 nm in front called ground producing a lot of noise, that is the point. 

 

It is like trying to point a flashlight on something in front a wall of spotlights.

 

Edited February 20, 2021 by FoxAlfa

[Shadow KT]

18 minutes ago, FoxAlfa said:

Yes, it did see, a huge radar reflector 2 nm in front called ground producing a lot of noise, that is the point.

 

 

or it could have been chance + contact at extremely close range

Edited February 20, 2021 by Shadow KT

[FoxAlfa]

26 minutes ago, Shadow KT said:

or it could have been chance + contact at extremely close range

It has no issue when not looking straight down at same range

[Mad_Shell]

Even low level hovering or slow moving helicopters can be really hard to lock on for radars, despite the moving rotor. Here is an extract from a very recent publication (full publication:  https://data.epo.org/publication-server/rest/v1.0/publication-dates/20170208/patents/EP2033011NWB1/document.pdf )

 

" A hovering helicopter’s body Doppler, however, has merged with clutter and only the return of its rotor assembly extends outside of clutter (due to the rotation of the rotor assembly). Even if the helicopter were moving, different flight geometries could put the helicopter’s body Doppler within the clutter region. If a seeker tries to estimate the range and range-rate of the rotor return, it will find conflicting range-rate measurements since the rotor return constantly changes with time and scintillates (both in amplitude and angle). Thus, the seeker will disregard a majority of the helicopter’s rotor return, and the rotor return samples will not be used to classify the potential target as a helicopter. [0009] There remains a need for a robust technique for detecting and classifying hovering and slow-moving helicopters that is compatible with the existing base of Doppler radars "

[Coxy_99]

Locking me with radar firing 120s at me PARKED i need to make that clear firing actives on me, Im in a plane not a chopper im not at 2000ft or notching or taxiing, Im parked missile tracks. I should be off his radar.

[GGTharos]

3 hours ago, Mad_Shell said:

Even low level hovering or slow moving helicopters can be really hard to lock on for radars, despite the moving rotor. Here is an extract from a very recent publication (full publication:  https://data.epo.org/publication-server/rest/v1.0/publication-dates/20170208/patents/EP2033011NWB1/document.pdf )

 

While it may be a challenge we already have a combat example of this being done.

 

This article is about the missile seeker, so I could see it being difficult to engage helis before AMRAAM-B (which has successfuly engaged helis in a low altitude situation).

 

Thanks for the article, let's pass this on to ED because it has some math inside that could help better simulate the look-down effect.

Edited February 20, 2021 by GGTharos

[GGTharos]

7 hours ago, Blinky.ben said:

I’ve learnt a lot from the community so might be a good time to give back. I thought I might try and clear something up about Doppler. Feel free to tell me to never post something this big again. This isn’t to point out how radars in DCS work but some people seem to compare DCS to real world and demand what is actually false.

 

Nah you should definitely post educational stuff, and thanks for posting it - not everyone knows about all this stuff and hopefully those who would like to know have more questions than answers after you post.    I'll just say this:

 

7 hours ago, Blinky.ben said:

So with all this in mind we can’t have all this in DCS or maybe we can I know nothing about coding. But to make something like chaff to have more of an effect or look down penalties have more of a influence to account for other disadvantages not within DCS sounds more like a simulated environment I think.

 

No it isn't possible to simulate all this in practice, not for a game anyway.   I don't think there's enough processing power to deal with it, but there's room to add more complexity/nuance.

 

At some point, the probability of detection exceeds 0.9, and it's all about inside/outside notch gate. Here the stochastics become irrelevant, and that's all that's being simulated.  The reduction in range due to look-down is simulated as well.   There are no anti-clutter circuits, no actual measured doppler in the game and it just doesn't matter - to simulate them we'd have to know how they work in the sense that you'd need to implement different clutter ... types, eg. sea, urban, sand, rock, etc, then each clutter circuit and how well it works against all that.   Then you'd have to somehow estimate the amount of RF garbage that's present so that you could simulate the radar degradation.

Gain is something that would be simple enough to simulate, and probably the clutter types as well - but simple doesn't mean that it's low amounts of work ... terrain clutter would have to be marked as being this type or that type, and right now we don't know how automated this process could be made on ED's part (or if it can be realistically automated).   I imagine ED might not be interested in this.

 

Further you have to know how radar works and what you expect to see as a result ... so again deep knowledge, people screaming bug without having this knowledge, etc. etc.  You can see how bad it is already with the simple simulation that's in place.