AIM 120C-5 looses lock and turns away

[skywalker22]

I know we have been through this countless times, but this really needs some fix. It's kind of that default examples where missile is being (almost) notched. But it was far from being perfectly notched, so this should not be the reason to loose the lock. Anyway, my point is, missile turns away, instead of at least following the trajectory which was on just before lost lock.

 

 

[henshao]

"Ewwww, is that a target?"

[skywalker22]

1 hour ago, henshao said:

"Ewwww, is that a target?"

Sure it is, F-14.

[Moonshine]

There might be some weird stuff going on with the Tomcat when its jamming, of course this is just a suspicion

Over the last months i had multiple tailchases (within 5nm), shot an amraam and it just never tracked at all. This while shooting in STT. Missile just flew past the target, barely maneuvering at all. In all of these cases it was a tomcat with its jammer turned on. 
 

[Noctrach]

4 hours ago, Moonshine said:

There might be some weird stuff going on with the Tomcat when its jamming, of course this is just a suspicion

Over the last months i had multiple tailchases (within 5nm), shot an amraam and it just never tracked at all. This while shooting in STT. Missile just flew past the target, barely maneuvering at all. In all of these cases it was a tomcat with its jammer turned on. 
 

I guarantee it's not jamming as I see it happening all the time when I'm flying Tomcat singleplayer without a jammer.

8 out of 10 AIM-120 shots can be evaded with a lazy turn because they just seem to miscalculate the intercept and actually just fly straight past me with plenty energy to spare.

[Moonshine]

okay, even weirder. i just remember that all of them were jamming at this point so there might have been a chance it would affect it. i did notice the amraam also missing other aircraft when shot from rear aspect at close range (within 5nm), barely pulling Gs, yet it seemed less frequent than against tomcats.

[Flappie]

@skywalker22The latest OB got an AMRAAM fix. Is it any better now?

Quote

• Weapons. AIM-120 AMRAAM - improved position estimation and reacquisition when target pases through notch. Missile Kalman filter was adjusted to significantly improve missile performance against manoeuvring targets. Many other improvements.

 

[skywalker22]

I didn't test it out yet, will do it today. 

[skywalker22]

Hmm, I did only 1 run, and look what has happened to AMRAAM. Seems not to be fixed for sure.

Tacview-20220922-163559-DCS.zip.acmi

[Moonshine]

Yeah some weird things seem to happen (focus on the missile going for the left target). In final stage of flight as the bandit passes the notch, the missile goes crazy, wobbling around terribly and then going straight up. 
 

so still some weird things going on, however the missile feels better than before the patch. Credit to that.

Tacview-20220921-180256-DCS-TWS_TEST.zip.acmi

[skywalker22]

There seems to be some work done on AMRAAM, Bogey Dope explains well what:

 

[DCS FIGHTER PILOT]

I just ran a couple of tests. It seems to be if the radar track is lost, even for a moment, the missile fails to go active on the target and as it passes it, it goes nuts (i.e. the 40g turn and wobbling). Interestingly enough though, the Hornet still seems to be able to make the missile go active even if the track is lost briefly. In the Viper and Eagle though, as I mentioned, if the track is lost for just a second, it does not go active even though it appears to be homing in on the target (See the tracks).

 

AMRAAM No Pitbull.trk AMRAAM No Pitbull2.trk AMRAAM No Pitbull3.trk

Edited September 22, 2022 by DCS FIGHTER PILOT

[henshao]

2 hours ago, DCS FIGHTER PILOT said:

I just ran a couple of tests. It seems to be if the radar track is lost, even for a moment, the missile fails to go active on the target and as it passes it, it goes nuts (i.e. the 40g turn and wobbling). Interestingly enough though, the Hornet still seems to be able to make the missile go active even if the track is lost briefly. In the Viper and Eagle though, as I mentioned, if the track is lost for just a second, it does not go active even though it appears to be homing in on the target (See the tracks).

 

AMRAAM No Pitbull.trk 356.78 kB · 1 download AMRAAM No Pitbull2.trk 357.08 kB · 0 downloads AMRAAM No Pitbull3.trk 402.29 kB · 0 downloads

 

 

great so now the AIM-54C is more advanced than the AIM-120

[Маэстро]

12 hours ago, DCS FIGHTER PILOT said:

I just ran a couple of tests. It seems to be if the radar track is lost, even for a moment, the missile fails to go active on the target and as it passes it, it goes nuts (i.e. the 40g turn and wobbling). Interestingly enough though, the Hornet still seems to be able to make the missile go active even if the track is lost briefly. In the Viper and Eagle though, as I mentioned, if the track is lost for just a second, it does not go active even though it appears to be homing in on the target (See the tracks).

 

AMRAAM No Pitbull.trk 356.78 kB · 4 downloads AMRAAM No Pitbull2.trk 357.08 kB · 3 downloads AMRAAM No Pitbull3.trk 402.29 kB · 3 downloads

 

Missile always go active at certain range to INS cue, this cue also defines targeting data for seeker - assumed target range, closing velocity, LOS angular rate etc. When you lose lock you drop datalink. Because target does maneuver in tracks(yes it does - it accelerates) seeker does not lock target.

[DCS FIGHTER PILOT]

13 hours ago, Маэстро said:

Missile always go active at certain range to INS cue, this cue also defines targeting data for seeker - assumed target range, closing velocity, LOS angular rate etc. When you lose lock you drop datalink. Because target does maneuver in tracks(yes it does - it accelerates) seeker does not lock target.

Should the missile seeker still see the target (as it does not deviate much) and still go after it? Before this update that is what was happening. 

[coolneko]

I've added a track file of the same problem. Before the latest patch I was able to launch a cheapshot 120 and it would hit almost every time with the AIs limited maneuvering after launch. Now it seems like either the radar cone of the 120 is smaller or the time that it goes active is later, missing the target completely.

pitbull.trk

[DCS FIGHTER PILOT]

44 minutes ago, coolneko said:

Before the latest patch I was able to launch a cheapshot 120 and it would hit almost every time with the AIs limited maneuvering after launch.

pitbull.trk 910.84 kB · 0 downloads

Heck I was able to do this most of the time against real players. Now if you drop the track for just a second (at any point in the missiles flight), you run a very high risk of the missile missing. 

[GGTharos]

On 9/23/2022 at 6:30 AM, Маэстро said:

Missile always go active at certain range to INS cue, this cue also defines targeting data for seeker - assumed target range, closing velocity, LOS angular rate etc. When you lose lock you drop datalink. Because target does maneuver in tracks(yes it does - it accelerates) seeker does not lock target.

Maestro, even the old sparrow would open up the doppler gates if it couldn't detect a target using pre-launch information.  This is the real problem with dropped tracks, the missile will drop certain constraints and you may have unexpected results, ie. it now gets to choose what it locks onto - with enough intelligence it may be able to choose one among many targets that best fits the criteria, or just the first thing it sees.  AIM-120 has a search mode; I know it's not represented in the game, but I want to point out that CHEAPSHOT tactics are real, and they have a place (maybe not quite how DCS players use them, but they exist).

In addition, there is an 'infamous' British experiment where they did not want to buy the data-link for the 120, and the Pk ended up being about the same as sparrow.   With datalink it is twice as good as sparrow (this is 120A/B era).  I know Pk is pretty nebulous but at least it gives an idea that something more is going on in there, so I think any target in a 120's search basket should be in danger.

Let me know if you need the relevant info as proof - again I just have sparrow docs, not 120 ... but we're talking old sparrow (7F) and you might imagine a 120 would do much better.

[Маэстро]

On 10/2/2022 at 8:49 PM, GGTharos said:

Maestro, even the old sparrow would open up the doppler gates if it couldn't detect a target using pre-launch information.  This is the real problem with dropped tracks, the missile will drop certain constraints and you may have unexpected results, ie. it now gets to choose what it locks onto - with enough intelligence it may be able to choose one among many targets that best fits the criteria, or just the first thing it sees.  AIM-120 has a search mode; I know it's not represented in the game, but I want to point out that CHEAPSHOT tactics are real, and they have a place (maybe not quite how DCS players use them, but they exist).

In addition, there is an 'infamous' British experiment where they did not want to buy the data-link for the 120, and the Pk ended up being about the same as sparrow.   With datalink it is twice as good as sparrow (this is 120A/B era).  I know Pk is pretty nebulous but at least it gives an idea that something more is going on in there, so I think any target in a 120's search basket should be in danger.

Let me know if you need the relevant info as proof - again I just have sparrow docs, not 120 ... but we're talking old sparrow (7F) and you might imagine a 120 would do much better.

Hi GGTharos,

First of all our model does "open" velocity gates if it can not find target near INS cue. But this opening is gradual and based on remaining TOF(to INS cue). So target acquisition is not guaranteed in case of cheap shot and depends on target motion params and interception geometry. Target may simply come out of missile FOV before velocity search range became wide enough to find it. There is also similar restriction on range search.

Second thing is real AMRAAM does not do any gate opening. I have a research paper which claims that. We implemented gradual gates opening just for game pourposes.

AIM-7 indeed have velocity search pattern that initially performs narrow and then wide sweeps, but that's a bit different weapon concept.

[GGTharos]

On 10/4/2022 at 6:20 AM, Маэстро said:

Hi GGTharos,

First of all our model does "open" velocity gates if it can not find target near INS cue. But this opening is gradual and based on remaining TOF(to INS cue). So target acquisition is not guaranteed in case of cheap shot and depends on target motion params and interception geometry. Target may simply come out of missile FOV before velocity search range became wide enough to find it. There is also similar restriction on range search.

Second thing is real AMRAAM does not do any gate opening. I have a research paper which claims that. We implemented gradual gates opening just for game pourposes.

AIM-7 indeed have velocity search pattern that initially performs narrow and then wide sweeps, but that's a bit different weapon concept.

Thanks for the reply.  Is this research paper available publicly?   In any case, the 120 does something IRL if not changing gates, otherwise IRL experience cannot be reconciled.

Edited October 5, 2022 by GGTharos

[Маэстро]

On 10/5/2022 at 4:24 PM, GGTharos said:

Thanks for the reply.  Is this research paper available publicly?   In any case, the 120 does something IRL if not changing gates, otherwise IRL experience cannot be reconciled.

 

You are welcome. You may find some info in ADA355385 p.4-5. I do not see any big contradiction with reality. Cheapshot tactics in DCS works too. If target does not maneuver such tactics works well but quite not well if target does maneuver.

[KlarSnow]

1 hour ago, Маэстро said:

You are welcome. You may find some info in ADA355385 p.4-5. I do not see any big contradiction with reality. Cheapshot tactics in DCS works too. If target does not maneuver such tactics works well but quite not well if target does maneuver.

All that document says is that the missile searches an uncertainty volume around the cue. It does not state how that volume is defined. It also does not state that the volume is static. It very much stands to reason that if there is a loss of update, that the missile should expand its parameters to find something.

And just to be clear Cheapshot is not snipping the missile before it is active. Cheapshot is snipping the missile at HPRF active, so Cheapshot tactics implies the missile is HPRF active and searching for or has found the target. 

US Army Brevity link

CTRL-F Cheapshot in that document.

Without HPRF, or HPRF indications in the cockpit in DCS it is currently Impossible to do Cheapshot tactics.

There is no terminology for snipping the missile before HPRF active, that would be a trashed or snipped shot, or just not supported to active.

Edited October 16, 2022 by KlarSnow

[Маэстро]

35 minutes ago, KlarSnow said:

All that document says is that the missile searches an uncertainty volume around the cue. It does not state how that volume is defined. It also does not state that the volume is static. It very much stands to reason that if there is a loss of update, that the missile should expand its parameters to find something.

And DCS missile do this expand because this is very possible when datalink is lost. But if datalink is present it looks like missile does not do such expansion. From document:

The AMRAAM searches a volume, known as the uncertainty volume, around the data link provided point. If the targeting data is bad enough the target will never be seen by the missile, even if it's own robust Electronic Protection capabilities would allow it to track and guide on the jamming target.

This means what if target inside of missile FOV but outside of uncertainty region it will be ignored.

35 minutes ago, KlarSnow said:

And just to be clear Cheapshot is not snipping the missile before it is active. Cheapshot is snipping the missile at HPRF active, so Cheapshot tactics implies the missile is HPRF active and searching for or has found the target. 

US Army Brevity link

CTRL-F Cheapshot in that document.

Without HPRF, or HPRF indications in the cockpit in DCS it is currently Impossible to do Cheapshot tactics.

There is no terminology for snipping the missile before HPRF active, that would be a trashed or snipped shot, or just not supported to active.

Good to know.

[GGTharos]

7 hours ago, Маэстро said:

You are welcome. You may find some info in ADA355385 p.4-5. I do not see any big contradiction with reality. Cheapshot tactics in DCS works too. If target does not maneuver such tactics works well but quite not well if target does maneuver.

Thank You Maestro,

I am reading the paper and I have a couple of comments regarding how it is interpreted - like beauty I guess, it is in the eye of the beholder because:

1) They do not quantify what 'bad enough data' means ... how far off do you have to be?  ECM can produce target information that is well outside of the target's physical maneuver capability

2) They are concerned with ECM, not a maneuvering target.  Targets maneuver, so if the command link is lost, I would expect the missile to attempt to find an exact parameter match and then open up the parameters if it does not, based on the amount of time it was flying in inertial mode

3) Do you more some ability of the missile to select different targets in unresolved groups? (I know this doesn't completely apply to DCS, but if you were to shoot 120's without DL at a close group, they should have a good change of picking a different target.  My guess, the 'seed' would be the datalink channel and the decision would be taken at the moment the missile's search mode can resolve the group when close enough)

4) The DCS F-15C TWS mode drops the track the moment a target sneezes through the notch.  Given how you've implemented the AMRAAM, could you guys please fix this TWS mode?  I know it's an FC3 bird but I don't think this is an unreasonable request.

I'll summarize how I interpret this:  I think the paper is concerned with ECM which can produce Vc and Range (and possibly other effects) that varies quite extremely compared to what an aircraft can physically do.  The paper does not seem to concern itself with maneuvering and does mention what happens when DL is lost, but rather it is concerned with the type of information the DL provides to the 120 before the missile goes active.   Consider here how TWS builds an 'oval' to consider the next hit for a track - this would be your uncertainty volume and you need about 1 sec to correctly capture a 12g target, 2 sec interval for 6g etc. If the interval is 10 sec, the volume becomes very large and it becomes much more likely to capture the wrong target.  But here we do not consider ECM, only maneuvering and so the parameters/logic would be different between 'I have uncertainty due to target maneuver' and 'I have uncertainty due to ECM' which could easily place target data as being 'outside of the capture volume'.

[MARLAN_]

On 10/16/2022 at 8:14 AM, Маэстро said:

And DCS missile do this expand because this is very possible when datalink is lost. But if datalink is present it looks like missile does not do such expansion. From document:

The AMRAAM searches a volume, known as the uncertainty volume, around the data link provided point. If the targeting data is bad enough the target will never be seen by the missile, even if it's own robust Electronic Protection capabilities would allow it to track and guide on the jamming target.

This means what if target inside of missile FOV but outside of uncertainty region it will be ignored.

Good to know.

For your SA: In the F/A-18C we have in DCS, we should have a solid "A" on the ATTK page to indicate HPRF Active (Husky) and then change to a flashing "A" to indicate MPRF Active (PItbull) - currently in DCS we are receiving only a solid A (Note that it is not a requirement to go from HPRF to MPRF, depending on target aspect you can go directly to MPRF). (DCS F18 HUD is correct as is, assuming it is indicating active (HPRF or MPRF), the HUD does not differentiate between HPRF/MPRF only if it is active.) If ED doesn't have evidence for this I can indicate a page/index number for it, it's either in the FRM or 742 (as well as the other book)

Also, does that mean HPRF active is implemented in DCS? We only lack the indications? Great news if that's the case, the indications I guess wouldn't be too hard to add.