[RESOLVED] F-14 RWR missile gives launch tone incorrectly

[Tomsk]

Loving the F-14 Heatblur, you've done an amazing job. The level of detail and work that's gone into it is clear, and this is clearly step above anything that's come before. Heatblur has raised the bar for everyone else to follow.

 

So with that said, one issue that I've noticed is that the F-14 RWR seems to be much too sensitive. In particular, it seems to give a missile launch warning if there is a SAM vaguely near (like within 50 miles) that is launching a missile at someone, even if you are nowhere near the target the SAM is launching at, even if you are completely the opposite side of the SAM.

 

This (I believe) does not accurately reflect how SAM guidance (and RWRs) work, and is not how RWRs work in other modules (for example the F/A-18C and A-10C). Most radar guided SAMs use SARH and guide missiles using a tight radar beam. As I understand it RWRs detect the missile launch by detecting that tight guidance beam. Because the beam is tight it general means only the aircraft that is being guided to (or one that is very close to it) will see the guidance beam and get the RWR launch tone. The F-14 RWR is currently giving a missile launch tone when it seems there is no way it would be receiving enough power from the guidance beam.

 

EDIT: oops excuse the title should say "F-14 RWR gives missile launch tone incorrectly", apparently I can't edit it now.

 

EDIT: Now with actual concrete measured examples https://forums.eagle.ru/showpost.php?p=3864406&postcount=12

Edited April 13, 2019 by IronMike

[Knock-Knock]

Might want to read this post by IronMike (Heatlbur dev) on the subject.

https://forums.eagle.ru/showpost.php?p=3841440&postcount=11

[Tomsk]

Might want to read this post by IronMike (Heatlbur dev) on the subject.

https://forums.eagle.ru/showpost.php?p=3841440&postcount=11

 

Yes, I've seen that. I'm saying it's not working properly :) I've been getting RWR warnings for the SAM launching on an aircraft that was more than 10 miles away. Where my bearing to the SAM is 090, and the aircraft being launched on is at bearing 135 from the SAM.

[Knock-Knock]

Oh, was actually the follow up post that contained the info I had in mind. I guess this falls into that?

 

The geometrical relations.

The main beam lobe width of a tracking radar can be as large as 10-15 degrees. Plus there are side lobes. The same applies to the missile guidance signal. In DCS, the emitter owner (module) decides which objects are withing the beam geometrical limits, and sends information about being painted by the radar to those objects. Maybe some systems are a bit too generous in informing about being locked, but it's not something we can control.

 

https://forums.eagle.ru/showpost.php?p=3847831&postcount=18

[Tomsk]

Oh, was actually the follow up post that contained the info I had in mind. I guess this falls into that?

 

https://forums.eagle.ru/showpost.php?p=3847831&postcount=18

 

Ah I hadn't seen that though. Yes, it's just I've spent a lot of time in the F/A-18C and have noticed the F-14 gives launch warnings when the F/A-18C definitely would not. The Hornet's interpretation seems more accurate in this case, given that the F-14 gives warnings when I was definitely nowhere near (in distance or angle) the aircraft that was being launched on. In one mission this resulted in my RWR constantly saying I was being fired on by an SA-6, due to a fight the AI was having 40 miles away.

Edited March 29, 2019 by Tomsk

[Victory205]

It is awesome to watch Heatblur model the vagaries of real systems. Complex gear doesn't work perfectly in aviation. Even airliners that havé been flying for decades with regular updates have constant glitches and annoyances that have to be solved.

 

Tactical systems are being spoofed, blanked, countered and jammed. This is just scratching the surface, and it is fantastic.

[Tomsk]

It is awesome to watch Heatblur model the vagaries of real systems. Complex gear doesn't work perfectly in aviation.

 

I can appreciate this argument, in real life things aren't as "clean" as in a sim. On the flip side, there is clearly a point where the system is just not modelling the reality very well. A RWR that just screamed "SAM launch" the moment you turned it on, is just modelling a broken RWR.

 

After flying the Tomcat regularly for the last few days and having many missions where the RWR seems to be constantly on fire telling me of SAM launches even though I've never been launched on even once, I'm kind of skeptical how realistic that modelling is. However, you might know better than me :)

[1Shot1KiLL]

I can appreciate this argument, in real life things aren't as "clean" as in a sim. On the flip side, there is clearly a point where the system is just not modelling the reality very well. A RWR that just screamed "SAM launch" the moment you turned it on, is just modelling a broken RWR.

 

After flying the Tomcat regularly for the last few days and having many missions where the RWR seems to be constantly on fire telling me of SAM launches even though I've never been launched on even once, I'm kind of skeptical how realistic that modelling is. However, you might know better than me :)

 

getting launch warnings 100miles out to see when a sam fires at a friendly in land. launch warning for friendly SA10s ( maybe normal dont know but starange if it is normal ) and on top of this, jester is constantly yelling missile missile 2 o clock brake left break right when friendly sams go off. jesters dumb as xxxx :)

[Super Grover]

That's the world of the EM warfare. The RWR is not a crystal ball or a magical box that knows which missile carries a death wish with your name on it. And radar antennas are no lasers.

 

To detect a target, a radar has to emit quite significant energy in the form of an electromagnetic wave. The wave has to travel to the target, and then back to the radar antenna. It means that the signal that returns to the radar is much much weaker than the signal at the target - the signal that the RWR can receive. And that weaker signal must be still intensive enough for the SAM to detect the target. So the signal at the RWR, in a typical situation, is quite intensive even at longer distances. However, there are situations when the RWR can receive faint signatures from a lethal threat. So choosing which potentially lethal threat ignore and which keep is very difficult. And usually, it's better to be conservative and warn about more threats that aren't lethal than to miss that one missile that wants to hunt you down.

 

Add to that that no antenna is perfect, and it has multiple side lobes - and the RWR can't tell if a signal received comes from the main lobe or a side lobe. As Victory205 wrote, this is just scratching the surface.

 

Now let's return from those theoretical considerations to our DCS World. The good news is that we found a way to emulate the main/side lobe shape for the SAM threats using some fancy in-house developed techniques . Additionally, we added some additional custom corrections to the guidance signal strength which will improve the simulation at longer distances. Now it's as real as it gets.

 

Does it make the RWR more selective? At very long distances, it makes the situation a bit clearer. But the closer you are to a SAM, the more prone to fake launches the RWR becomes - and the result it's similar to the current state you have in the sim. Because the SAMs are just huge glowing electromagnetic lanterns.

[Eldur]

I just don't get the issue and why it should be F-14 related and the F-18 not having it. I've got a map with the carrier group SW of the small islands in the middle of the Gulf. I was flying around, triggering a flight of Su-24s coming up from the base at Bandar Abbas, going into the direction of the carrier group. When they came into vicinity, I was already south of the CG, going towards Al Dhafra and suddenly I got lock up and missile warnings popping up from a lot of "AE"s thinking WTH at that moment. Checking directions I knew it was the CG, but I kept wondering why they tried to shoot me out of a sudden. Then I checked F10, F2 and F6 just to see they engaged the Fencers coming from ~60ish whereas I was at 150-160 from the CG, more than 90 degrees of off what they attacked, getting full missile launch warnings by my ALR-67 from the "AE"s. TBH I still don't know what that stands for as I write this up as it doesn't correlate in any way to John C. Stennis, Oliver Hazzard Perry or Ticonderoga class "Normandy". But it still gives me nightmares... this was in the 18 ofc, long before the cat was here.

[SDsc0rch]

fascinating insight ---- but i think there's actually a bug in DCS where everyone on a side gets an RWR lock indication if ONE person on that side gets locked - it wasn't like that, and then the bug popped up and... that's the way it is now ---- would be nice if ED could take that on

i'm sure its on the list...

[Tomsk]

Because the SAMs are just huge glowing electromagnetic lanterns.

 

So my understanding is they are more like electromagnetic flashlights. They direct radar energy at the target they are painting to guide the missiles. This makes sense from a tactical point of view, the guidance beam shoudn't be much larger than needed to guide the missiles or you just make the SAM more vulnerable to ARMs. Indeed you say seem to say that the guidance beam is narrow (10-15 degrees) in your other post:

 

The main beam lobe width of a tracking radar can be as large as 10-15 degrees.

 

However you currently seem to be modelling a beam that's more like 75 degrees wide or more. Here's some examples from tacview that prove my point:

 

You can see my F-14, you can see an SA-6 (which is 25nm away from me) has just launched on number of nearby friendly F-16s. In this example I have made a hard break turn and turned cold because my RWR is telling me that I am being launched on. However, the angular difference between my bearing to the SAM and the actual target being tracked is 30 degrees (measured using a paint tool). It's as though the SAM is projecting a 60 degree wide tracking beam, which as I understand it isn't at all realistic.

 

Here's another example:

My squad mate has just been launched on by an SA-2, I'm turning away because my RWR is telling me I've been launched on. It keeps telling me this for an extended period so I continue to break. Again you can see the angular difference is greater than 30 degrees. As I understand it I shouldn't be being told I'm being launched on here.

 

Another example:

Here I'm hard breaking after my RWR is telling me I'm being launched on (20nm from the SAM). Again you can see the aircraft that is actually being launched on and the difference is more than 30 degrees.

 

Another example:

Again a friendly is launched on, again I'm turning away hard here because my RWR is telling me I'm being launched on, but the angle is 20 degrees different.

 

Indeed, my RWR is still screaming that I'm being launched on here, at 37 degrees (this missile has gone for chaff):

 

Another example:

I don't turn away here because I've wised up to it at this point and can see the launch visually, but my RWR is still saying I'm being launched on.

 

I could probably keep going and find even more examples in my recent tacviews, but these are probably enough to make the point: it seems like the guidance beam is being modelled much larger than it really is, as a 75 degrees wide beam (or more). This is not a weird one-off, it's very repeatable. I have not observed this problem in other modules, only with the F-14.

Edited March 30, 2019 by Tomsk

[Victory205]

Ah I hadn't seen that though. Yes, it's just I've spent a lot of time in the F/A-18C and have noticed the F-14 gives launch warnings when the F/A-18C definitely would not. The Hornet's interpretation seems more accurate in this case, given that the F-14 gives warnings when I was definitely nowhere near (in distance or angle) the aircraft that was being launched on. In one mission this resulted in my RWR constantly saying I was being fired on by an SA-6, due to a fight the AI was having 40 miles away.

 

Really? The Hornet seems more accurate to reality? Lay out your logic here please. How do you determine which one is accurate.

 

The RWR picks up all sorts of spurious energy from all sources. Sidelobes from friendly radars, wingman, your own leaking gear.

 

Sitting on the deck of an aircraft carrier would pinball all sorts of systems due to proximity to the ship's emitters. If Heatblur modeled that, you all would be screaming bloody murder about the module being "broken".

 

The standard can never be "it doesn't work the way I want it to work..."

[Tomsk]

Really? The Hornet seems more accurate to reality? Lay out your logic here please. How do you determine which one is accurate.

 

As in my post I just posted, Super Grover says they are modelling the tracking beam as being 10-15 degrees wide, which seems reasonable. However, the results in the simulation show it is working as a beam that is 75 degrees wide or more. Hence why I suggest it's inaccurate.

[BlackLion213]

As in my post I just posted, Super Grover says they are modelling the tracking beam as being 10-15 degrees wide, which seems reasonable. However, the results in the simulation show it is working as a beam that is 75 degrees wide or more. Hence why I suggest it's inaccurate.

 

Have a look at how emissions take place with a focused EM beam. Even if the beam is tightly focused there are additional emissions that encompass 360 degrees. Less power than the main lobe, but certainly significant.

 

https://en.m.wikipedia.org/wiki/Side_lobe

 

-Nick

[Victory205]

As in my post I just posted, Super Grover says they are modelling the tracking beam as being 10-15 degrees wide, which seems reasonable. However, the results in the simulation show it is working as a beam that is 75 degrees wide or more. Hence why I suggest it's inaccurate.

 

"Seems reasonable". Got it. I'm slayed.

[Tomsk]

Have a look at how emissions take place with a focused EM beam.

 

Sure, I am familiar with the concept of sidelobes on directed antennas. However, the side lobes have a lot less power than the main lobe right. For the 'typical' example shown on wikipedia the strongest sidelobe is -15 decibals relative to the main lobe. Decibals is a logarithmic scale, so that's POWER(10, -15/10) = 0.03 so in that wikipedia example the strongest side lobe has at most 3 percent of the power of the main beam. It drops off quickly as well, so at around 15 degrees off axis the strongest sidelobe is around -23 decibals, which is 0.5 percent (i.e. one two hundredths) of the power of the main beam. At 30 degrees off, where I was getting lock indications, the example on wikipedia suggests the strongest sidelobe might be around -35 decibals, or 0.03% (3 hundredths of one percent) of the power of the main beam.

 

It seems to me that it is ... unlikely that this would trip the RWR into believing it had been launched on. Particularly given that in some of my examples the SAM is over 40 nautical miles away.

 

In the end, the modelling of directed antennas isn't my field of expertise. So if Heatblur look into this issue and decide the simulation is accurate that's fine. However, I'm raising it as a potential issue because based on some observations and some simple maths it looks very suspicious. It would also be an easy mistake to make I imagine, for example forgetting to apply a logarithmic scale to a power output based on decibals.

Edited March 30, 2019 by Tomsk

[probad]

behold the power of simple maths lmao engineers and scientists can suck it

[FSKRipper]

For the 'typical' example shown on wikipedia the strongest sidelobe is -15 decibals relative to the main lobe.

 

You made me speechless pal, you found the no. 1 source for information on military radar properties :lol:

 

Our EW guys picked up every little spike in a 80nm radius to find the Blue Fox from our "enemies". Even when they locked something up 10nm apart from us it was possible to detect their emissions. The electronics were also not interested if a lock signal was weak or strong...

Edited March 30, 2019 by FSKRipper

[Tomsk]

You made me speechless pal, you found the no. 1 source for information on military radar properties :lol:

 

Actually it wasn't my example, it was Black Lions :) Of course I'm sure the real data on the exact properties of military radars is highly classified, so I doubt heatblur is using that to model their systems. However, on the grounds that one directed radar antenna is probably similar to another and that you would want to minimize sidelobes as much as possible in such a system it doesn't seem a terrible approximation.

 

Our EW guys picked up every little spike in a 80nm radius to find the Blue Fox from our "enemies". Even when they locked something up 10nm apart from us it was possible to detect their emissions. The electronics were also not interested if a lock signal was weak or strong...

 

Right but different tool, different purpose. I'm sure you know more about this topic than I do so correct me if I'm wrong but the role of EW in this case would be to detect and locate enemy radars, so you'd want that tool to be as sensitive as possible. Whereas ideally you'd want an RWR to only show "you are being launched on" if you actually are, or if it was at least fairly likely. You ideally wouldn't want it to tell you "you are being launched on" if it's actually someone else that's nowhere near you who's being launched on. False positives are better than false negatives, but you would try to filter out as many false positives as you could. Different tool, different function, different design right?

Edited March 30, 2019 by Tomsk

[FSKRipper]

Actually it wasn't my example, it was Black Lions :) Of course I'm sure the real data on the exact properties of military radars is highly classified, so I doubt heatblur is using that to model their systems. However, on the grounds that one directed radar antenna is probably similar to another and that you would want to minimize sidelobes as much as possible in such a system it doesn't seem a terrible approximation.

 

 

 

Right but different tool, different purpose. I'm sure you know more about this topic than I do so correct me if I'm wrong but the role of EW in this case would be to detect and locate enemy radars, so you'd want that tool to be as sensitive as possible. Whereas ideally you'd want an RWR to only show "you are being launched on" if you actually are, or if it was at least fairly likely. You ideally wouldn't want it to tell you "you are being launched on" if it's actually someone else that's nowhere near you who's being launched on. False positives are better than false negatives, but you would try to filter out as many false positives as you could. Different tool, different function, different design right?

 

Similar tools but more sophisticated in our version as it was handled by 3 persons. As you already said false positive is better than false negative and exactly this happens. Strength of emission can be used for an educated range guess, even from an Exocet missile.

The point is when you receive a lock from a fire control channel (not relevant if on your unit or just painted from a side lobe when the enemy radar is tracking) you treat it as a lock on your unit and prepare for immideate missile defense even when you are only painted intermittent.

 

I have no idea how the more automatic systems in planes work but since the pilot has no time to eveluate the threat over some minutes cause he is busy with flying, fighting... I would expect the RWR to expect the worst and recommend defensive measures.

Edited March 30, 2019 by FSKRipper

[Super Grover]

Hey Tomsk,

Thanks for your involvement and all your effort in explaining your problem. I appreciate that you want to make our product better.

 

However, you got some points wrong, and you come from some false assumptions which make your conclusions wrong.

 

First, I've never written that we model the beams as 10-15 degrees wide. We don't. At some point, I might have written that some of the SAM systems use quite wide beams (true for some older systems), as wide as those numbers.

Currently, the beam size is governed by DCS or module owner, and usually, it's wider than that, and we cannot control it. Over the next weeks, we will introduce our custom enhanced emulation of the main and side lobes which should improve the beam modelling.

 

As stated in the preceding posts, the electromagnetic wave generation and propagation isn't simple to understand. The radars have to work at really long distances, where some of the equations are inversely proportional to the square of the distance or even the fourth power of the distance. The amplitudes change drastically over the range, and the numbers of decibels (note: decibels not decibals) you mentioned are easily reached when compared something that is close and something that is far. It's really complicated.

 

I'd love to bring you our algorithms and explain the mathematical models that stay behind them so that you could verify them on your own, but unfortunately, I can't because of the obvious confidentiality reasons. Probably, it would also be quite exhausting and boring to get through them for most of the readers. Hence, I must ask you to trust us that we really know our job and that we have enough education and skill to do it right.

 

Nevertheless, you're right with one thing. We should never stop chasing for perfection. And that's why we're preparing an update or a series of updates containing our custom beam properties modeling that I mentioned before. I hope you will appreciate them, because our main goal is making the RWR as realistic as possible, given all the limitations of the platform. And realistic quite often means imperfect.

 

Best regards

[Tomsk]

Nevertheless, you're right with one thing. We should never stop chasing for perfection. And that's why we're preparing an update or a series of updates containing our custom beam properties modeling that I mentioned before. I hope you will appreciate them, because our main goal is making the RWR as realistic as possible, given all the limitations of the platform. And realistic quite often means imperfect.

 

Thank you for your time in answering my query. As I say it was the wide angular range over which you get lock indications that surprised me, rather than the distance: I'll well believe the distance aspect is complicated. My observation was simply that your RWR works very differently to the other modules RWR, and that the way it works seems very surprising particularly with regards to angles. This makes more sense if you say DCS is modelling much wider beams than would really be the case, and I really look forward to your updated logic.

 

Once again I'd like to congratulate you on the great work you've done with the F-14 and for your first rate community interactions. I realise we can be a difficult bunch at times :)

[Beamscanner]

The HEATBLUR ALR-67 is leaps and bounds more accurate than any other RWR in DCS.

 

OP doesn't understand all the factors at play. It is very realistic for an RWR to trip from a SAM 40nm away. Even if its only receiving sidelobe energy.

 

The RWR is dumb, and can only make assumptions based on what it measures.

 

Here are some things to consider:

- All antennas have sidelobes. Radar antennas, RWR antennas, all of them. No antenna is perfect, they all leak energy in unintended directions.

- RF energy is light. Light scatters, reflects, diffracts, refracts, etc.

- SAM radars emit very powerful signals so they can ensure target detection. Especially old SAMs

- RWRs are designed to detect extremely weak signals so they can inform the pilot to emitters even far away.

- RWRs have a benefit over radar receivers in that the light has only traveled one way, where as radar receivers have to detect light that traveled out and back in (two way path). Thus, compared to a radar's receiver, and RWR will more easily detect signals at range.

 

 

The DCS RWRs (besides the HEATBLUR ones) do not simulate noise, amplitude comparison (AOA) and have unrealistic blind spots. They use relatively simple logic like

 

'When SAM XXX locks Client XXX, trigger RWR detection'

 

A real RWR doesn't know it been locked on to. It listens to a very wide range of frequencies, detects and identifies the signal, compares the amplitude (or phase) between it's antennas, and outputs the info to the pilot.

 

At no time does the SAM communicate who its intended target is (well besides in a secure datalink to it's allies). Enemy SAMs are non-cooperative. Thus, whether or not you're the one being targeted is unknown.

Edited March 30, 2019 by Beamscanner

[Ignition]

getting launch warnings 100miles out to see when a sam fires at a friendly in land. launch warning for friendly SA10s ( maybe normal dont know but starange if it is normal ) and on top of this, jester is constantly yelling missile missile 2 o clock brake left break right when friendly sams go off. jesters dumb as xxxx :)

 

 

Yeah, Jester freaks out by missiles fire by own side and friendly locks.

 

 

Its really annoying specially because the rwr spikes every time by every single radar on the map including friendlies.