Airplane Defense Logic

[LargeDiameterBomb]

C3k,

your post answering CWReese just seemed very flippant at first look and I missed your earlier post here, or rather I thought Chickensim wrote it.

For me it was perfectly obvious that DWReese was being serious and you starting your post by asking if he was sarcastic seemed (Out of the context of your earlier post) like you thinking his opinion was so crazy that it didn't even merit a serious reply, especially when you mentioned being a pilot for 28 years.

Now when I saw that it was you who had written that earlier post everything falls into place though.

I apologize for the misunderstanding.

[DWReese]

I know that you are busy, and I am not complaining, but when you get a chance can you look into the Airplane Defense Logic thing again? We (you/everyone) had some suggestions that I think would really improve play within the game.

I was just playing Broken Shield. I was well on my way to a complete, absolute, and total Israeli victory when two of my F-35s went downtown Damascus to take out the HQ. Suddenly, a lowly SA-3 emerged. The SA-3 has a 1 percent chance of killing my F-35s. But, my F-35s start evading, going into the old "spiral of death" corkscrew routine, where they continue to turn INTO (toward) the shooter. So, instead of being very safe on the outskirts of town, but F-35s in up right in the middle of the city with SAM shooters everywhere. Both eventually get shot down.

I know that we can turn off automatic evasion, but it would be so much better if you could apply some of your wizardry and come up with a workable solution to eliminate the "spiral of death" issue once and forever.

(BTW, the pilots ejected in downtown Damascus, so there was no chance to rescue them. I would hate to be them.)

Thanks.

Keep up the great work.

[ARCNA442]

One easy addition I would like to see made is a "min altitude when evading" doctrine setting. A few times I've had planes attempt to evade a fairly non-threatening radar SAM by diving down into a mess of extremely dangerous IR SAMs and AAA. And setting min altitudes is a very common real-world tactic to avoid precisely this situation.

Personally, I feel like the more doctrine settings we have the better. They do a great job allowing you to customize your force's reactions without having to manually micromanage everything.

[DWReese]

The altitude issue is another great point. Not all defense settings are the same for the attacking planes each time. Often it is necessary to go low, and at other times (as you pointed out), it is better to remain high. Having some adjustable parameters available in the doctrine would be a great way to avoid various threats.

[Frostitute]

So the problem isn't actually a mistake by the airplanes, its an error in missile code. Pilots will indeed turn 90 degrees to the threat as it poses the greatest problem to the missile. Its helpful that it becomes difficult for pulse doppler radars but the more important aspect here is that the closer to 90 degrees the aircraft is, the more the missile will perceive the aircraft's change in direction thus making it easier to pull ahead of the seeker. The problem here is that the missile flies towards the aircraft's current location making it end up somewhat behind the aircraft. Thus the aircraft keeps it at 90 degrees, and the result is that the defender flies towards the threat. In the real world the missile will fly a straight line towards the point at which the missile meets the aircraft for the shortest distance (surprisingly easy to calculate with basic analogue systems) and so there is no turning effect, and the pilots will often keep 90 degrees from the threat site because it is so close to the missiles direction that there is little problem.

This is not to be confused with other techniques mentioned in other posts here. There are a number of specific techniques to evade specific missiles or specific ranges - for instance flying away or at an angle greater than 90 degrees might be used as a suboptimal technique where gaining distance is more important, or it might be used when a straight line away puts the aircraft out of range of the threat anyway. Other techniques like flying towards the missile and diving it into the ground are also interesting but not particularly relevant at this point. The generic missile evasion technique any time you don't have a specific instance and technique to use would simply be flying 90 degrees from the missile, manoeuvring in the vertical to bleed missile energy if you have time, and then remaining with the missile at 90 degrees as you pull hard in the hopes that the missile either runs out of energy and can't follow or that the seeker can't follow. In this way the aircraft are following a fairly accurate defence to an inaccurate missile trajectory.

If I were to make a suggestion, it would be firstly to fix the missile guidance logic even if the guidance check is only done occasionally at longer distances to make sure the game isn't dragged down too much. One could even apply this as a feature with datalinked missiles only occasionally altering course as per the datalink updates. Then the aircraft could run a calculation to see if a turn and burn would get them out of the missiles range or not. Mentioned earlier was a point about not knowing where the missile is and who it is from - if an aircraft has an RWR then the majority of attacks particularly predating active missiles will be known both the type of the attack and a rough/accurate location of the missile depending on if the defender has spotted the missile or knows the SAM site location. If these aren't met (no RWR, IR missile etc) then default to the standard defence tactics. With an accurate missile model other techniques could be included such as attempting to burn missile energy though frankly I'm not sure much more needs to be added beyond just dive and beam for a reasonably good tactical simulation.

[DWReese]

If you read from the beginning of the thread then you would see that an annoying tactic that the plane does it continue to turn back toward the SAMs shooting at it. So, a plane that was at 10 miles out and flying away, will begin turning, and then turning again, and then turning again, and before long it finds itself 4 miles out because it keeps getting closer to the SAM, not further away.

Turning off the Automatic Evasion will get the plane out of harm's way sooner, but will also subject it to little defensive moves.

Hence, the need to come up with a method of continuing to move away from danger, but still maintaining some kind of defensive posture. Some really good ideas came out of this discussion.

Altitude was also mentioned. Often, these planes begin diving down in areas to escape high-flying SAMs, only to find themselves now be shot at by AA guns and low-firing SAMs.

So, it would be nice if the planes took a somewhat less-predictable, and more efficient, method to flee from danger.

The devs will come up with something.

[Sardaukar]

This is one of my pet peeves too.

[LargeDiameterBomb]

I had an Idea after I wrote that long IF-THEN post with some other comments a while back in this thread. In that comment, I touched on the subject that it is unfortunate that DECM systems doesn't work in the same way as OECM systems, in other words like a jammer capable of jamming one or perhaps a couple of emitters (Depending primarily on how modern the DECM system is but also on crew size and thus pilot load, where a two person crew, one of them being a WSO/CSO/RIO might be able to jam more emitters) with a low jamming power that just might break lock for SARH SAMs in the same way a OECM system on an aircraft X is able to do so if turned on when a hostile fighter Y, using it's onboard radar for illumination, has just launched a SARH AAM at aircraft X just after the radar has gotten within range to be able to successfully lluminate the target.

I understand this would take a lot of developmental resources to implement though, but I have an idea that might be much easier to implement. In the missile endgame calculations, an aircraft that has a DECM system automatically lowers the chance of a missile hitting the aircraft, how much depending on the difference in "tech level" between the DECM system and the SARH seeker. Let's say a very modern DECM system is coming up against an old SARH SAM system, which would probably give the missile a 35 % penalty to it's chance to hit.


But what if that calculation was performed not in the endgame, but instead a few seconds after the launch has been noticed by the targeted aircraft and the pilot/WSO activates the DECM system (Here the already implemented concept of the OODA loop for automatic missile evasion might come in handy)?
So, if the penalty to the missile hitting would be 35 %, a "roll of the dice" would then happen when the DECM system is "activated" and if it comes up 35 or lower, the DECM system/self protection jammer is considered to have broken lock and the launched missile will disappear in 1-2 seconds, which represents the missile veering strongly off course.

How to do with ARH missiles dependant on mid-point updates is another thing, but I have some ideas there too, but let's leave that aside for now.

I don't know if this idea alone will solve the problem DWReese has brought to our attention, but together with a doctrinal choice to primarily dive and run from a missile while taking evasive action only at the last second if the missile comes very close to the aircraft (Whatever that means), it might make old SAM systems at least less likely to become these "fly traps" because when lock is breaked, the aircraft is already running away instead of beaming and keeping a constant distance to the SAM system.

Also, this seems like something that could be implemented relatively easily (Please notice that I said relatively [:)] ).


---------------------------


Also, this is just a guess on my part and I do not in any way claim any definitive knowledge here, but it seems like very modern DECM systems, like late 2010 ones gives too small a penalty to a missile, equipped with a very early seeker, chance's to hit. ECCM was, as I understand it, almost an unknown concept, at least in field use, in the 1960s, since ECM was just starting to appear.

To exemplify, when a F/A-18E with a AN/ALQ-214(V)4 self protection jammer is being attacked by a SA-3b missile with an early 1960s SARH seeker, I would guess that the penalty should be more than the present 35 %, which I think is the most severe penalty I have noticed in game. I have gotten the impression that jamming systems of late has become much more effective, because of what I guess is an exponential increase in computational power and cumulative advances in solid state electronic performance in general. Of course, this is not something I as a civilian can find any hard data on.

With an even bigger chance of breaking lock with a state of the art jammer vs an old seeker the solution outlined above would become even more effective.

Can anybody with some specialist knowledge on this give any comment, without saying too much? Maybe c3k for instance, your input and opinion would be very much appreciated.

[DWReese]

I kind of like your "simplistic" approach as a potential solution. It would, of course, need lots of testing, but it sure seems simple enough and would, at the very least, offer up some relief to this situation.

[LargeDiameterBomb]

Thanks for your feedback. Glad to hear that you think the idea is of some value.

I think this idea has it's strength primarily in that it would be relatively easy to implement.

In the best of worlds, it would work quite nicely.

Instead of an engagement between a F/A-18E from 2018 and a paltry SA-3b leading to certain death,
the jet would immediately turn and start to dive and run (and since it is already diving it doesn't have to turn into the missile before the missile comes within maybe 2 nm of the aircraft). A third of the time the missile would also "just disappear" and the jet would have a substantially longer time to run, and when the next missile is fired after the first missile "just disappeared" the jet is also also from the start pointing roughly away from the SAM site and is also already at or close to maximum speed, while if it first had avoided a missile by defensive maneuvering it would first have to turn approx 120-150 degrees and then accelerate to it's maximum speed after coming out unscathed from the missile endgame maneuvering.

But as you say, of course it will have to be tested, but considering the things above I think there is a chance that that 1/3 or 1/4 of the time that the missile "just disappears" gives the jet a fair chance to eventually get outside the firing envelope of older SARH SAMs.


-----------------------------


Also, a note on my thoughts on jammers in the previous post. I know many users strongly dislikes when other users say that so and so equipment "seems" to be less effective in CMANO than it is in the real world without providing hard data. I said I was guessing, but anyway, let me expand my thoughts a little before you group me with the grug-brains.

I think a good case can be made for that a late 2010 level DECM systems should be having a higher chance to break lock on a missile from the early 60s than 35 % or at least that the difference between the penalty a missile incurs should be much bigger between a late 2010 DECM system and an early 1960s seeker vs the same jammer and a missile with a late 1980s seeker, even if we have no data to go by.
Now the penalty on chance to hit that a missile accrues due to the difference in tech level between seeker and DECM system in CMO now seems to be linear (I am not 100 % sure on this though), but to me it intuitively seems like it instead should be exponential.

The reason for this is that when technology is dependent on computational power, there is a kind of threshold effect, so when a large enough amount of computational power suddenly becomes available old theoretical concepts that have already been formulated can almost instantly be translated into technology.
This is the same phenomenon we see with AI today, where AI research was for a long time, during the 80s and 90s, in a "winter" where almost no progress seemed to be made, but once a threshold was passed, especially in 2010 when a paper was published that showed that a common graphics card could replace 10 ordinary CPUs in certain AI applications (Machine learning) with the same performance, theoretical concepts that were before impossible to get to work in practice suddenly went from being fiction to reality and as a result for instance machine vision has seen an explosion in performance and the same could be said for machine translation or conversational AI and a host of other technologies that have been enabled by the leap in available computational power during the 2010s, even though the theoretical concepts used are to a large extent from the so called winter of AI where little progress was made and many scientists became quite pessimistic about the whole field.

I think the same type of case could be made for many types of military technology, but especially electronic warfare, for instance jammers, where technologies such as DRFM and the type of jammers that technology enable. DRFM, which was first described as a theoretical concept in the mid-1970s if i remember correctly, was a kind of technological threshold that was overcome precisely by the availability of large amounts of fast computational power and which enables new jamming tactics that were only theoretical concepts a couple of decades back, and which some people in the know today vaguely seems to regard as a real threat even against early/mid versions of the AMRAAM - so I think it's fair to say that the effect on a 1960s seeker would probably be much more forceful.

But this is probably not the right place to further discuss this subject anymore since the thread has been moved to tech support.

[DWReese]

I know that you are super busy, but any chance that this topic can be addressed sometime in the near future?

I do believe that some good ideas came out of this discussion. Perhaps some of these ideas could help in developing a new, useful approach. Having to constantly micromanage (manually switch back and forth from Yes to No to Yes....gets a little tiresome. <G>

Thank you so much.