Mad_Shell

While I agree with what you say, I hope the elements I provide above, which given the context of the publications are almost certainly describing how the AIM-120C deals with target beaming, will be considered.

After some research, I'm pretty sure I've found THE solution to the problem (no, it's not clickbait lol): 2 patents (now in the PUBLIC DOMAIN), deposited by Hughes (manufacturer of the AIM-120 at the time), in 1995 and 1996 (the AIM-120C deliveries began in 1996!). I join the patents PDF files. The patent EP0747723A2 was deposited first, then the following year the patent US5748140A was deposited, and uses the 1st one to refine the method. Both patents explain how to specifically track a target beaming the missile to hide in the ground clutter. Here is how they describe the situation they claim to solve: 7 That picture at the bottom reminds you something, no? Here are just a few highlights of the system and algorithms, and the differences with the new missile API by ED: - the new ED's API uses what is called "constant false alarm rate" (CFAR), which means detection occurs when the signal/noise ratio is high enough. Problems arise when the target signal has not enough range/Doppler separation with the ground clutter (as we see in DCS currently). In these patents, it is claimed that: That means the radar can determine the angles from which the signals (target and clutter) come from! That also means that this angle information can be used to help discriminate the target (the whole process is describes in the patents, way too long and complicated to write here). Basically, that huge ground patch producing an enormous noise and described by Маэстро earlier, with this system the vast, vast majority would not hamper target detection as the angle of the returns are different enough from angle the target returns. - here are the results they obtain: These results show that when the beaming target has a return that is barely superior to the clutter return, and the clutter comes from an angle with 3° difference compared to the target return, with the old method (used by ED) the detection probability was almost 0%, while with the new method it's already 15%. At signal/noise ratio where the old method has a probability of detection of just 50%, they now obtain between 65% and 90% of detection. IMPORTANT: Keep in mind, those results are for when the clutter comes from a single big source, with an angle slightly different from the angle of the target return. In most DCS situations, the clutter comes from a way bigger area, moderately reflective. In those situations, almost all the clutter will be filtered out by the new method, since most of the clutter will come from angles very different from the angle of the target return. in the US5748140A patent, it is explained that if the missile seeker has a lock on the target, but loses the target fort a short period, the lock gates will continue to follow, for a period of time, along the predicted target angle, Doppler and range. That allows the missile to not break lock immediately when the target isn't detected anymore, and to continue to fly toward the predicted intercept point. If the target is lost for too long (a few seconds?), the lock is lost and the radar enters search mode again I'll also add some personal remarks on the modelling as explained by ED and Маэстро: - to your remark "it assumes unambiguous range" I quoted at the top, it seems you don't model blind Doppler and blind ranges. Now, most MPRF radars indeed use several frequencies (generally 8 ) to avoid those blind regions, but also to solve range and Doppler ambiguities. I don't quite understand why you assume target range is ambiguous, when most MPRF radars solve that problem. - target detection probability should be a thing for all radars in DCS. The Mirage, and in the last update the JF-17, implement this now, and we see how it's way more realistic than "5sqm RCS = detection at 26 miles exactly", like in all other modules. - We see how important it is in real life that the target RCS is way bigger when seen from the sides and from above/bellow. Jets RCS are tens or hundreds of time bigger than when seen from front aspect, and that plays a huge role in being able to still detect them while they're beaming a radar. That's not modelled in DCS, except for the new missile API. Stores increasing the RCS would be really nice too. - the patents authors here mainly refer to a missile, but you can be sure that means that similar systems are used on the most recent radar jets (F-16, F-18, JF-17), and that beaming those radars should probably be really quite hard (most probably VERY hard). EP0747723A2.pdf US5748140.pdf

The stinger version we have in DCS has no proximity fuse irl, only impact fuse. The Army began to add a proximity fuse only 2 or 3 years ago, for better effectiveness against UAVs.

Well that's a simplification, since both HPRF and MPRF (not even sure this one is modelled for the AMRAAM in DCS) are ambiguous in distance, there are several range gates around several possible distances, but yes basically it's that. And it wouldn't surprise me if the AMRAAM would use 2 different frequencies to solve the range ambiguity irl... (but we'll probably never know lol)

Sorry, but since the target is at high altitude, and closer to the missile than any clutter source (ground), and no chaff is present, how do you explain that the missile is not able to follow the target in the notch using signal/noise ratio? The range gates should allow excellent signal/noise ratio in this situation, and the absence of chaff means no secondary target that can confuse the missile.

Here are a few suggestions to make the new AI perform better and more realistically in BVR: - the AI should crank and dive as soon as it has fire its missile, but only if it is in range of enemy missiles. For example, lets take a situation with a F16 vs a M-2000C. Currently the F16 (with AMRAAMs), will crank and dive as soon as it fires its missile. That means it gives up its altitude for nothing, since it's not even inside enemy missile max range. The AI should remain high and nose hot on target as long as outside enemy missile range. - the AI should use more tactics, especially to be able to face opponents with longer range missiles. Currently the AI remains high and nose hot on target as long as it hasn't fired its missile. That leads them to always eat an enemy missile in the face if the enemy has longer range missiles. The AI with lower range missile should use tactics such as baiting the opponent into taking a shot, turning cold for a few moments while remaining high, turning hot again, notching the enemy plane radar, etc... everything to close a bit the distance. Ideally using mountains too, but I recon it's probably really hard to program an AI to do that.

Some planes with very good high altitude performance, like the F-15, can launch an AMRAAM at very long distances. However they don't seem to take into account the range limitation due to missile battery duration. That leads the F-15 to sometimes fire its 1st missile from too far, wasting it, and cranking and dropping altitude too soon. That makes for situations like the F-15C almost always losing a BVR fight against a MiG-29S, despite having a superior missile. bug_BVR_AI_AMRAAM_battery.trk bug_BVR_AI_AMRAAM_battery_2.trk

Ran some tests, and here are the results: - 1 vs 1: the AI cranks and drops altitude just after firing its missile (correct behaviour) - 1 (blue) vs 2 (reds): each red AI fire a missile then cranks and dives. The blue AI launches 2 missiles, but stays nose hot and only defends once enemy missiles go active (track bug_BVr_AI_1vs2.trk) - 2 (blues) vs 2 (reds): each red fires a missile against an enemy plane, and remains nose hot, only defending once enemy missile goes active. Same for blues. (track bug_BVr_AI_2vs2.trk) bug_BVR_AI_1vs2.trk bug_BVR_AI_2vs2.trk

I agree, that delay seems a bit ridiculous, as it leaves very little, if any, time to react to a missile. I don't know why a delay has been introduced in the lua, why would you delay a threat warning? I hope it's fixed soon, it makes CMWS quite hard to use effectively in manual mode, as you have to look at the quadrant at all time.

If after several years there is nothing to report on stuff like DTC for example, I'm honestly pessimistic it'll ever come in DCS... And that's the problem to me and some more people... as a helicopter player mainly, and despite being almost a new player to DCS in the grand scheme of things (about 2 years), I'm already finding DCS so, so boring to play... I love the aircraft, learning the systems, procedures, and then, what? Always killing the same dumb static AI (my main grip)? Having to spend hours in the mission editor to make interesting missions because you have to tell the AI everything they have to do and their reactions? Spend even more hours to try to randomize them because otherwise there is no surprise during the mission? I, and I think many more people, am eagerly awaiting more news on those development items because I quite frankly struggle to enjoy DCS when it's so repetitive and "lifeless" (compare with that other F-16 sim).

The problem is that despite the weekly newsletter, we've had no news on the DTC in years. We've had no news on the dynamic weather system and AI seeing through clouds since the new clouds were released a year ago. We've had no serious news on the development of the dynamic campaign since quite some time too, while they hoped for open beta test Q2/Q3 2021. Remember that 3rd party talking about a pretty IADS? No news since a year. And so on...

It's been updated in 2.7.11 but, like many changes, it's not in the changelog.

PROBLEM 1: Displaying the TADS on a MPD can be useful to ID targets as a pilot, and George keeps using FLIR only, often with non optimal luminosity and contrast, while during daytime TV has better magnification and allows better ID. SOLUTION: a way to ask George to switch between TV and FLIR with the TADS. PROBLEM 2: impossible to tell George to fire missiles on several targets. Each time you have to go in the target list, selec a single target, tell George to fire, and begin again for the next target. That's not only a chore, that also prevents tactical possibilities such as rapid engagement of several targets by firing several missiles in quick succession then lasing another target as soon as the one lased is destroyed, which considerably increase the rate at which the apache can destroy targets, and decrease the exposure time to enemy defenses. It also prevents George to fire 2 or more missiles at a target within a few seconds. For example that's a problem for engaging Tor on excellent level, as they'll intercept Hellfires if they're fired one after the other. SOLUTION: a system to select several targets in the target list. RIGHT SHORT press would select a target as #1 (a small "1" could be displayed next to the target in the target list), then going over another target, a RIGHT SHORT press would select it as #2 target (with a small "2" appearing next to it), and so on... A RIGHT LONG press on a target would mean it's the last selected target and close the target list. A RIGHT SHORT press on an already selected target would deselect that target. If George is in "fire at will" mode, he'll fire a missile at each target in the selected order. If for example targets #3 and #4 are close enough, he'll not wait that the missile fired at target #3 impacts it. He'll fire a missile at target #4 about 8 or 10 seconds after firing the missile at target #3. After impact on target #3, George will immediately lase target #4. If the missile fails to destroy target #3, George keeps lasing target #3. If George is in "wait for manual authorization to fire" mode, he'll fire a missile each time the manual authorization to fire is pressed, even if the precedent missile is still in the air. Each time a target is destroyed, George will immediately lase the next selected target in the list. If a missile fails to hit a target, George will keep lasing it. EDIT: okay, an example of engagement so people can see some advantages of this system! You hover 300 feet above the ground, and spot 3 enemy tanks close to eachothers, at 6 km in front of you! CURRENT GEORGE SYSTEM: select a tank, tell George to fire. The missile flies for about 30 seconds, kills the tank. Select the next tank, the missile flies toward the target for about 30 seconds, kills the tank, etc... Time to destroy 3 tanks from the moment you fire the 1st missile? Almost 2 minutes! SUGGESTED SYSTEM: select the 4 tanks in the target list. Give George consent to fire. George lases the 1st tank and fires a missile. Wait 10 seconds and tell George to fire another missile. Wait again 10 seconds and tell him to fire another missile. 1st missile kills tank 1, George automatically lases tank 2. 2nd missile kills tank 2, George automatically lases tank 3. 3rd missile kills tank 3. Time to destroy 3 tanks from the moment you fire the 1st missile? About 50 seconds!

Max range is 8000 meters, not feet. Ground forces, including US ones, use the metric system. So the Apache uses it too for almost everything (except altitude, speed, aviation stuff) to facilitate coordination and communication with boots on the ground. Same remark. Unless you changed the unit system somewhere in the settings in one on the MPDs pages, those are kilometers.

I suspect that if it's a VRAM leak problem, that may cause problems for cards with 6 or 8 GB of VRAM, but not for your 3090. Would not surprise me if it was the reason you can't reproduce it.

It's a very heavy configuration, and high temperatures have a big effect on torque required to fly for helicopters. Also, the Delta model is known to be quite overweight for its engines.

Warhead effectiveness is about the same for the 2 missiles, but it's not everything... The Hellfire allows LOAL mode, remote designation, ripple fire...

I disagree. This behaviour mimics a human CPG having the pilot's LOS set as his acquisition source, then slaving the TADS to it, and deslaving once it's in the right direction. If you have a human CPG, you won't tell him "Look at my LOS with the TADS", then move your LOS around before he's looking where you want. Edit: typo

I agree, and made a similar post a few days ago: That would really help with target ID. FLIR has worse magnification, and it isn't always as easy to identify shapes and objects compared to DTV.

That's what I did, and it worked temporarily, but now several others AH-64D textures do the same. I ended up deleting all AH-64D releated textures then run a repair. Had a new "Hash Mismatch" error when the repair redownloaded the textures. And now my DCS updater won't launch! I get a "The application was modified" error, then nothing...

No mod installed, no script, nothing. I ran a full repair, without effect. Here is what I get for integrity check. Also, when I updated I got several hash mismatch errors.

The AGM-114K allows a variety of engagement modes and interesting tactics, that George is currently unable to perform. I propose the following improvements: - IRL the AGM-114K allows rapid fire, which means firing the missiles in quick succession (about 8 seconds interval for different targets, or less if several missiles are fired at the same target), before the precedent missile has hit its target. As soon as a missile hit its target, the CPG lases another nearby target, which the following missile will hit, etc... That allows faster and more lethal engagements, decreasing unmasking duration and time to react for the enemy. It also allows to fire several missile at a single target if necessary. For example in DCS, the Tor SAM set on excellent IA will intercept most Hellfires thrown at it. Currently George launches another missile only after the 1st one was intercepted (and the 2nd one will probably be intercepted too...). A smarter tactics would be to fire 2 missiles in quick succession so that the Tor can't intercept both in time. PROPOSITION 1: rework the manual "consent to fire" command. Each time it is pressed by the pilot player, George should fire a missile. If the 1st missile hits its target, George automatically lases the next target. If the 1st missile didn't hit its target (miss or intercepted), George keeps the laser on the 1st target. - the AGM-114K allows to use "RIPPLE" mode, allowing to fire 2 missiles programmed to track 2 different laser channels with less than 2 seconds interval. It allows to hit almost simultaneously 2 targets, 1 designed by own aircraft, 1 lased by a remote designator. Combine that with the ability to fire, like describe in 1st paragraph, the next salvo 8 seconds later while the precedent missiles are still in the air, and that gives the capability to strike 16 different targets in about 1 minute. PROPOSITION 2: add a "RIPPLE" mode to George.

When playing as pilot, with George as CPG, displaying the TADS video on a MFD can be useful for target ID and situation assessment. It's a bit frustrating that a long distance, during daytime, George only uses FLIR, and not DTV that allows a better ID, and with superior magnification. A command to ask George to switch between the 2 would be very welcomed.

Thx. Please report the G limit of the missile too. In DCS both MICA are limited to 24g manœuvers, while IRL they're advertised as having 50g manoeuvrability. That's a huge difference...

Flying along terrain features is a major aspect of helicopter flight. For example, flying just above a river allows to terrain mask efficiently. The ability to arrive next to a river, and just tell Petro/George "follow that river, direction North, altitude X ft", and having him do that would be a considerable improvement. With the current system, each change of direction has to be commanded, making it almost impossible to effectively follow such terrain features if you're a bit busy as a CPG.