Aim120 can be trashed with a barrel roll and chaff headon

[stefasaki]

10 hours ago, Notarobot said:

It's been reported numerous times and most multiplayer servers will ban you for it. What's your in game name by the way?

We’re not talking about the famous AOA-roll here related to syncing issues in MP, this also works in SP.

Edited November 3, 2024 by stefasaki

[GRY Money]

With the new update it seems even easier to do it

[Hobel]

vor 23 Stunden schrieb GRY Money:

With the new update it seems even easier to do it

My subjective impression based on testing and game experience is that it can't be fixed  only by guidence.

There has definitely been some improvement this year compared to before, the missile used to burn a lot of energy vs a roll at long range, now it has to get very close so it starts to have problems. 

But look at the results ingame basically, and I think that's an important point, the missile often only flies a few meters past between 10-20m, that's actually a very good result with these extreme maneuvers if you ask me.

Nevertheless, it doesn't change the end result, the role or in other words "abuse" unfortunately works very well.

 

 

Edited December 5, 2024 by Hobel

[pan1ker]

Кавказ_BVR_training_v23-20241204-190713.trk Tacview-20241204-190741-DCS-Client-Caucasus_BVR_training_v23.zip.acmi

 

Edited December 5, 2024 by pan1ker

[Xhonas]

This is definitely not fixed. Can you take a look at it again? 

[Mad_Shell]

On 11/2/2024 at 8:12 AM, Маэстро said:

I just meant that it may be quite hard to find more optimal solution, because our implementation is based on a number of works of scientists for whom this field is of primary interest.

And it's also based on some knowledge about several real systems.

@Маэстро  I found an article from 1994, dedicated specifically to this problem of high G barrel roll creating a big miss distance for missiles. 

Their solution is to introduce a phase lead in the yaw and pitch guidance channels (1.5 radian gives excellent results)

imado1994.pdf

[Маэстро]

On 12/7/2024 at 8:52 PM, Mad_Shell said:

@Маэстро  I found an article from 1994, dedicated specifically to this problem of high G barrel roll creating a big miss distance for missiles. 

Their solution is to introduce a phase lead in the yaw and pitch guidance channels (1.5 radian gives excellent results)

imado1994.pdf 621.83 kB · 19 downloads

I'm afraid it is not practical. If target will do any other maneuver than barell roll missile will not properly guide. Enableing and disabling this pase shift is not an option. How missile will identify type of maneuver to enable phase shift? How fast it be able to indetify that such maneuver had started or ended? What if target will not fly directly to the missile? etc.
 

On 12/5/2024 at 11:47 PM, Xhonas said:

This is definitely not fixed. Can you take a look at it again? 

It cannot be "fixed". Please read careafully my first post in this topic.
But since last patch make things even worse we will roll back filter prarmeters closer to previous values.

[Mad_Shell]

20 minutes ago, Маэстро said:

I'm afraid it is not practical. If target will do any other maneuver than barell roll missile will not properly guide. Enableing and disabling this pase shift is not an option. How missile will identify type of maneuver to enable phase shift? How fast it be able to indetify that such maneuver had started or ended? What if target will not fly directly to the missile? etc.
 

It cannot be "fixed". Please read careafully my first post in this topic.
But since last patch make things even worse we will roll back filter prarmeters closer to previous values.

Yeah, there is definitely the problematic of recognising the maneuver based on target acceleration vectors, and the delay to process that information. I can see how non trivial it could be. 

Speaking about delay, since you introduced a delay in the servocommands, what is the value of that delay? Reducing it would be the most effective way to deal with high g barrel rolls according to a few papers I found, but it should remain realistic of course

Edited December 9, 2024 by Mad_Shell

[Маэстро]

1 hour ago, Mad_Shell said:

Yeah, there is definitely the problematic of recognising the maneuver based on target acceleration vectors, and the delay to process that information. I can see how non trivial it could be. 

Speaking about delay, since you introduced a delay in the servocommands, what is the value of that delay? Reducing it would be the most effective way to deal with high g barrel rolls according to a few papers I found, but it should remain realistic of course

Sorry, I can not provide any value. AMRAAM Klaman filter is non-stationary, filtering level (for a specific target type/size) changes based on current measurement noise level, relaive velocity and range. Different filter state variables(los angle, angular rate, target acceleration) have distinct lags("delays"). It's obvious that reducing lags may help, but as I already said these lags can not be just reduced.

[AeriaGloria]

7 hours ago, Mad_Shell said:

Yeah, there is definitely the problematic of recognising the maneuver based on target acceleration vectors, and the delay to process that information. I can see how non trivial it could be. 

Speaking about delay, since you introduced a delay in the servocommands, what is the value of that delay? Reducing it would be the most effective way to deal with high g barrel rolls according to a few papers I found, but it should remain realistic of course

This part of the AIM-120 paper that ED released when the FM was remade is likely what you want

[Xhonas]

12 hours ago, Маэстро said:

It cannot be "fixed". Please read careafully my first post in this topic.
But since last patch make things even worse we will roll back filter prarmeters closer to previous values.

Ah come on, this is so frustrating. DCS is incredible but these small issues can really hinder the enjoyment of playing this game. Can't you adjust the proximity fuze then since you really aren't up to tweak the guidance? Ik you don't want to create a "wonder weapon", but come on, what we have right now and what we had in the previous update is not realistic at all. Your math may be right, but the practical results in the game are completely arcade-y. 

Edited December 9, 2024 by Xhonas

[Mad_Shell]

1 hour ago, Xhonas said:

Ah come on, this is so frustrating. DCS is incredible but these small issues can really hinder the enjoyment of playing this game. Can't you adjust the proximity fuze then since you really aren't up to tweak the guidance? Ik you don't want to create a "wonder weapon", but come on, what we have right now and what we had in the previous update is not realistic at all. Your math may be right, but the practical results in the game are completely arcade-y. 

Barrel rolls and periodic sinusoidal maneuvers are a real challenge for missiles IRL. You can find online tons of articles, including very recent ones. dedicated to this problem, because the miss distances can be of several tens of meters, which no realistic value of proximity fuse will solve. The missile has to dynamically adapt its proportional navigation laws to the target maneuvers, which is non trivial.

[Xhonas]

32 minutes ago, Mad_Shell said:

Barrel rolls and periodic sinusoidal maneuvers are a real challenge for missiles IRL. You can find online tons of articles, including very recent ones. dedicated to this problem, because the miss distances can be of several tens of meters, which no realistic value of proximity fuse will solve. The missile has to dynamically adapt its proportional navigation laws to the target maneuvers, which is non trivial.

Look, i'm not saying that the maneuver per se is unrealistic, it is possible to use this maneuver to defeat some missiles, like older SAMs, as it happened IRL with an F-16. However the results we are seeing in the game for the amraam certainly aren't realistic. Otherwise, why pilots would invest countless hours in learning BVR tactics, timelines and opt to stay as far away as possible from enemy aircraft? just do a barrel roll and defeat any missile. 

I remember another frustrating issue in DCS regarding missiles, not too long ago, the amraam would try to intercept enemy air to air missiles almost every time, even if you were targeting the aircraft launching them, not their missiles. Up to this day, if you are fighting an F14, the amraam will regularly stop tracking the tomcat and start tracking the Phoenix if the combat is in the 10-15nm range. Also, you could easily detect air to air missiles with the radar in the F/A-18 and deliberately target the enemy missiles, to the point that an enemy flanker could launch all of his 10 missiles and you could intercept them all and merge with it. 

Now, i don't doubt that radars can see air to air missiles in some situations, however, the way that it was implemented in DCS gave extremely unrealistic and arcadey-type results. So, just because some math is right, doesn't necessarily means that the final result of the simulation will be correct. The same thing is happening here, right now.

Edited December 9, 2024 by Xhonas

[GRY Money]

Problem contributing to this is that a violent high G roll is not affecting the pilot's fatigue level. 

People can spin around at 7g forever and there's no downside to doing it. If you perform such a maneuver in Il-2 the pilot gets tired with heavy blurred vision and controls won't work like they should, giving you partially randomized control stick inputs. 

Edited December 9, 2024 by GRY Money

[Mad_Shell]

1 minute ago, Xhonas said:

Look, i'm not saying that the maneuver per se is unrealistic, however the results we are seeing in the game for the amraam certainly are. Otherwise, why pilots would invest countless hours in learning BVR tactics, timelines and opt to stay as far away as possible from enemy aircraft? just do a barrel roll and defeat any missile. 

I remember another frustrating issue in DCS regarding missiles, not too long ago, the amraam would try to intercept enemy air to air missiles almost every time, even if you were targeting the aircraft launching them, not their missiles. Up to this day, if you are fighting an F14, the amraam will regularly stop tracking the tomcat and start tracking the Phoenix if the combat is in the 10-15nm range. Also, you could easily detect air to air missiles with the radar in the F/A-18 and deliberately target the enemy missiles, to the point that an enemy flanker could launch all of his 10 missiles and you could intercept them all and merge with it. 

Now, i don't doubt that radars can see air to air missiles in some situations, however, the way that it was implemented in DCS gave extremely unrealistic and arcadey-type results. So, just because some math is right, doesn't necessarily means that the final result of the simulation will be correct. The same thing is happening here, right now.

I agree with you here. But just saying to ED "just fix it" won't help. Because if they try to fix it with some bandaid (unrealistic proximity fuse value, magical terminal guidance,...) it will only create other problems and people will complain that now the missiles are unrealistically effective in some situations. That's why finding some article describing a realistic solution is needed imo.

[Xhonas]

6 minutes ago, Mad_Shell said:

I agree with you here. But just saying to ED "just fix it" won't help. Because if they try to fix it with some bandaid (unrealistic proximity fuse value, magical terminal guidance,...) it will only create other problems and people will complain that now the missiles are unrealistically effective in some situations. That's why finding some article describing a realistic solution is needed imo.

I'd rather have a band-aid fix until a more realistic solution can be found than have what we have right now. 

[M0ltar]

I’ve been away for almost 2 years and it’s strange to see that this is still an issues. I’ve flown for a couple days and it’s to the point that I’m contemplating stopping again as it’s just as bad as it’s always been, if not worse since it seems more people know about the exploit and are using it on public multiplayer servers.

When we did SATAL issues such as this were a huge problem and something that I constantly had to police when running that event. The roll that was an issue was more of a spaz roll where people would input and induce tons of rapid instant AoA that should break or deteriorate an aircraft in real life that would cause a missile to desync and miss. We banned anyone found to be doing this. Traditional barrel rolling, as in flying your plane like you’re flying around a barrel, was fully allowed, and if the missile was slow enough, would cause it to bleed all of its energy and miss. Valid tactic!

I think ED is misconstruing the 2 here in 2 different scenarios. You should be able to run a missile out of energy if it’s already slower, losing the ability to pull the needed AoA to kill the player. This primarily is an option at lower altitudes with thicker air where missiles slow rapidly. The issue most people here are talking about is what can happen at high altitude at huge speeds where the missile still has plenty of AoA to hit its target. No maneuver by a plane should be fast enough for a missile track to lose you. If it has the energy, and therefore ability to pull the needed AoA, it should kill you, especially at higher altitudes.

Players abuse this head on exploit mechanic to cause missiles to miss when they have no reason to and I assume it’s still due to desync. Chaff should have almost ZERO effect on a C, especially if the aircraft is not in a notch, and even then should have virtually nil cause on the missiles tracking, but for some reason still works. Continuing to bring up the barrel rolling maneuver because it’s real world is just evidence, to me at least, that there is still a misunderstanding of what the issue is. While that maneuver does work in a certain situation, it’s not supposed to work where people are using and abusing it. 
 

 

Edited December 10, 2024 by M0ltar

[Xhonas]

19 hours ago, M0ltar said:

I’ve been away for almost 2 years and it’s strange to see that this is still an issues. I’ve flown for a couple days and it’s to the point that I’m contemplating stopping again as it’s just as bad as it’s always been, if not worse since it seems more people know about the exploit and are using it on public multiplayer servers.

When we did SATAL issues such as this were a huge problem and something that I constantly had to police when running that event. The roll that was an issue was more of a spaz roll where people would input and induce tons of rapid instant AoA that should break or deteriorate an aircraft in real life that would cause a missile to desync and miss. We banned anyone found to be doing this. Traditional barrel rolling, as in flying your plane like you’re flying around a barrel, was fully allowed, and if the missile was slow enough, would cause it to bleed all of its energy and miss. Valid tactic!

I think ED is misconstruing the 2 here in 2 different scenarios. You should be able to run a missile out of energy if it’s already slower, losing the ability to pull the needed AoA to kill the player. This primarily is an option at lower altitudes with thicker air where missiles slow rapidly. The issue most people here are talking about is what can happen at high altitude at huge speeds where the missile still has plenty of AoA to hit its target. No maneuver by a plane should be fast enough for a missile track to lose you. If it has the energy, and therefore ability to pull the needed AoA, it should kill you, especially at higher altitudes.

Players abuse this head on exploit mechanic to cause missiles to miss when they have no reason to and I assume it’s still due to desync. Chaff should have almost ZERO effect on a C, especially if the aircraft is not in a notch, and even then should have virtually nil cause on the missiles tracking, but for some reason still works. Continuing to bring up the barrel rolling maneuver because it’s real world is just evidence, to me at least, that there is still a misunderstanding of what the issue is. While that maneuver does work in a certain situation, it’s not supposed to work where people are using and abusing it. 
 

 

There are 2 major issues that plagues the 120 in multiplayer right now and one of these issues affects all dcs missiles. First issue is, if you do a loaded roll, barrel roll, the 120 will not be able to keep up with you and will miss. Check the video in the thread below. It was performed in single player, current DCS version.

The 2nd issue that affects the 120 and all the other missiles in MP is teleporting. When a player starts doing a loaded roll, if he rolls the jet too fast, like in the video in the thread below, it will cause the player to teleport (desync), thus making the missile miss, this is a problem for all missiles and is an issue that is being neglected for a very long time now.

 

Please ignore the notching part of the video, lets just focus on the barrel roll problem for this thread.

 

Edited December 10, 2024 by Xhonas

[Hobel]

vor 21 Stunden schrieb Mad_Shell:

which no realistic value of proximity fuse will solve

So which value is realistic?  

There are references to the aim120 that it could also be 15m, but let that not be valid enough.

And 15m has already been tested by the community, it reduces abuse more than enough.

The warhead is at least enough to do some damage.

So my question here would be why should the value be limited to a fixed 9m when the fuze would be perfectly capable of detecting targets up to 15m, plus the seeker and Fuze can interact with each other to find a better solution. However, it was decided not to go beyond 9m, the fuze sees the target but does not trigger because?

What is the reason or philosophy that the value is fixed or so low?

Edited December 10, 2024 by Hobel

[Mad_Shell]

58 minutes ago, Hobel said:

So which value is realistic?  

There are references to the aim120 that it could also be 15m, but let that not be valid enough.

And 15m has already been tested by the community, it reduces abuse more than enough.

The warhead is at least enough to do some damage.

So my question here would be why should the value be limited to a fixed 9m when the fuze would be perfectly capable of detecting targets up to 15m, plus the seeker and Fuze can interact with each other to find a better solution. However, it was decided not to go beyond 9m, the fuze sees the target but does not trigger because?

What is the reason or philosophy that the value is fixed or so low?

I would prefer we go over the dedicated thread for discussing the warhead, this one is for the guidance issue. 

[Hobel]

vor 27 Minuten schrieb Mad_Shell:

I would prefer we go over the dedicated thread for discussing the warhead, this one is for the guidance issue. 

The topic name is.

Zitat

Aim120 can be trashed with a barrel roll and chaff headon

And this includes supposed solution ideas that can be discussed. And that's exactly why PF was invented in case the missile doesn't hit directly.

And that is precisely the case here, among other things.

 

[GRY Money]

Am 10.12.2024 um 20:09 schrieb Hobel:

So which value is realistic?  

There are references to the aim120 that it could also be 15m, but let that not be valid enough.

And 15m has already been tested by the community, it reduces abuse more than enough.

The warhead is at least enough to do some damage.

So my question here would be why should the value be limited to a fixed 9m when the fuze would be perfectly capable of detecting targets up to 15m, plus the seeker and Fuze can interact with each other to find a better solution. However, it was decided not to go beyond 9m, the fuze sees the target but does not trigger because?

What is the reason or philosophy that the value is fixed or so low?

In multiplayer having bigger fuzes will have funny results i believe.

Btw the Aim-120 is the only missile in the game that has this issue. The Sparrow cannot be barrel rolled and neither can the R-77. 

[Hobel]

vor 14 Minuten schrieb GRY Money:

In multiplayer having bigger fuzes will have funny results i believe.

Nothing different from missiles that already have 11-15m PF.

 

 

vor 19 Minuten schrieb GRY Money:

Btw the Aim-120 is the only missile in the game that has this issue. The Sparrow cannot be barrel rolled and neither can the R-77. 

Other missiles such as the SD10 have more PF, otherwise the SD10 etc. would also miss more often.

Test it out At the same energy state. How far misses the SD10 and then Aim120.

The Aim120 has also been tested with a higher PF and the results are very similar to other missiles.

[Muchocracker]

22 hours ago, GRY Money said:

In multiplayer having bigger fuzes will have funny results i believe.

Btw the Aim-120 is the only missile in the game that has this issue. The Sparrow cannot be barrel rolled and neither can the R-77. 

As maestro has already stated in this thread. The amraam is the only one directly effected by this because it's the only missile on the new API (besides the PL-12)

Said new API seems to simulate almost every possible source of noise in the missile. It eventually adds up. 

- slight imperfections in the fin servo's that create inaccuracies in the readout of the position of the fin or in executing those commands (over/under deflects)

- imperfections in the seeker gimbal servo's for the same reasons

- inaccuracies from the seeker for any number of reasons like the range resolution, imperfection of monopulse's angle measurement, digital signal processing lags, automatic tracking lags like the range gate loops. There's a ton here that could contribute, we'd be here a while if i listed every one of them.

- scintillation or otherwise known as glint phenomena that can radomly cause the seeker to think the target has suddenly moved 200 feet to the left from one pulse to the next. 

The Kalman filters that maestro mentions are there to smooth all of those noises out before it sends the final control inputs through the guidance autopilot. But those filters can also introduce their own lags depending on how its tuned and what weights it's given. It can be made super stable and resistant to noise but then it's sluggish to react to the target's true movements when they are rapid. Or it could be made more reactive but then those noise's have more effect.

For better or worse the new API is what it is. It's excrutiatingly detailed in every possible way you can and can't think of. Like maestro has repeated it's not as simple as adjusting the kalman filter and the problem is gone entirely. And we'll just have to take his word for it that it'll never be.

I think hobel and others are right that the proximity fusing is where the focus should be at. I for one don't understand why the proximity distance has to be so rigidly within the 100% kill radius of the warhead. A zero radial velocity trigger if the missile passes within a certain extended miss distance would be enough to at least cause some damage.

[MARLAN_]

I don't understand why ED is trying to reverse engineer a multi-billion dollar AMRAAM project instead of creating a reasonably realistic simulation AMRAAM instead...

We don't need a full simulation, especially if it doesn't work. Its a nice to have sure, but it needs to work. If something can be made more realistic that's great, but if it causes the missile to be defeated by hilariously simple maneuvers like this, how about we just stick with a reasonably realistic simulation instead. I don't care how much depth the AMRAAM has in the internal code that I can never even read anyway, if it feels real, that's good enough.

For example, if ED decided to implement background cosmic ray radiation causing random errors in the sim... I mean.. great, I guess, I can't argue with something being more realistic... but if their code adds a bunch of ridiculous bugs it wasn't worth it.

Edited December 16, 2024 by MARLAN_