Radar Warning Receivers

Background: RWRs track a wide spectrum of radar-band frequencies and use various signal processing techniques to determine what is and is not a military search or fire control radar. The signals are processed and analyzed against known threats. The RWR will then return a warning along with the type and threat level based on this processing.

From the point of view of a player with no classified knowledge of EW, but a background in physics and computing, there are areas where DCS's RWR modeling seems to fall short:

1. Radar wash and radar limits - aircraft often receive an FCR lock ("spike") indication, FCR guidance (missile launch) indication, or active radar missile homing indication when the source radar is of no threat to the aircraft in question or is well outside reasonable parameters. The beam width of a typical FCR is generally less than 1°. As a rule of thumb, this means the width of the beam is less than 1 NM for each 60 NM the radar energy travels. At ranges of less than 10 NM, the beam width is on the order of 0.1 NM.

Currently in DCS, lock and guidance indications, which occur in a situation where the radar antenna is trained on a specific target and no longer sweeping, appear on the RWR of other nearby aircraft with an extremely high frequency, even when the aircraft is significantly offset outside of the normal beam width. Additionally, for threats such as active radar homing missiles, the active radar missile warning indication appears for a significant period of time after the missile is no longer a threat. In fact, the aircraft can be behind the missile's seeker head and still receive the threat indication.

2. Radar strength and rejection - It stands to reason that RWRs have logic filters that would eliminate physical impossibilities from being displayed. Radio emissions follow the inverse square law. RWRs can detect the strength of the emission and determine the range, and thus threat level, of the source radar based on a threat catalog. Likewise, this catalog would know the maximum possible detection and launch ranges of the platforms involved. This mean an RWR should, for instance, reject a launch/guidance warning from a source radar that is beyond a reasonable distance to engage the aircraft. For example, an F/A-18C that launches an AIM-7 would generate a launch warning for any aircraft that received sufficient energy to indicate that the F/A-18C was at a reasonable range to guide a semi-active missile. Allowing for error and unknown factors (such as missile improvements), this range would likely be 40 NM at most (since a reasonable range for the missile is at best, 20 NM).

Additionally, the known capability of the FCR radar itself would be taken into account. If the radar is only capable of detecting fighter-sized targets at approximately 40 NM. It would be reasonable for the RWR to reject any radar guidance warnings of a strength that indicates that the particular radar is outside of 40 NM.

Currently in DCS it is possible to receive missile guidance and lock warnings from platforms well outside of their detection range let alone that particular platform's best missile Rmax. This physical impossibility should be filtered out by the RWR.

3. Warning decay times - RWRs must balance probability of warning (POW) against false alarm rates (FAR). This means that there is some signal processing time that must occur for the RWR to determine whether or not the received radar energy is in fact a threat radar and then to classify that threat. Higher confidence reduces false alarms, but increases the processing time. However, in aerial engagements, this increased processing time also comes at another price, the inability of an aircraft to know when he is "naked". In other words, how quickly the pilot is made aware that the enemy has stopped locking, guiding, or searching for his aircraft. This information is vital in beyond visual range missile engagements.

Currently in DCS, warning decay times are excessive. Multiple seconds elapse between a threat disappearing and the warning stopping. At best, 1-2 seconds is the longest period a threat should remain after the source signal is lost.

All of these issues means that RWR usage in DCS is difficult at best and worthless at other times. A DCS pilot simply cannot rely on the RWR to produce accurate, useful warnings in a busy battlespace. Only in small-scale, set-piece air battles can a pilot expect to be able to use the RWR to a reasonable degree.

The above is based on my personal testing and observations in the sim and my own reading and knowledge on the various physical principles involved. I would be interested to know what other people think or if there is other evidence that demonstrates that DCS's RWR modelling is actually accurate.

This isn't research, this is a bunch of incorrect assumptions.

This isn't research, this is a bunch of incorrect assumptions.

He’s actually pretty correct, especially regarding point #1, which has been a pet peeve of mine for some time in game.

Wrong section of the forum, perhaps.

He's certainly wrong about #1. Most FCR beamwidths are over 1 degree for one unless you have a fairly large antenna ... and that's with modern SAMs. Onboard missile radars have much larger beamwidth but again, minor detail.

As for missiles appearing on RWR well after they should no longer do so, sure - there are absolutely issues with these 'air mines' floating around doing nothing but bothering your RWR.

Background: Currently in DCS, lock and guidance indications, which occur in a situation where the radar antenna is trained on a specific target and no longer sweeping, appear on the RWR of other nearby aircraft with an extremely high frequency, even when the aircraft is significantly offset outside of the normal beam width. Additionally, for threats such as active radar homing missiles, the active radar missile warning indication appears for a significant period of time after the missile is no longer a threat. In fact, the aircraft can be behind the missile's seeker head and still receive the threat indication.

missiles in dcs will turn around and kill you, so that RWR to the rear is accurate in game.

DCS does generally do a poor job at modeling RWRs, but your points are wrong.

1. Way off on FCR beam widths. And because of the inverse square law, RWRs can detect much more than the main beam of the FCR. But that's dependent on range. Also, multi-path can cause detections.

2. DCS F-15 and SU-27 RWRs will straight up indicate the targets range. IRL, RWR ranges are very course estimates that do not account for atmospheric attenuation or the noise of the local EM environment. The RWR may not even know if it detected a main beam, sidelobe or multi-path reflection. Or perhaps you're in a climb and the signal came into your RWR antenna's sidelobe and were thus weaker.

Ignoring Missile launches "beyond 40 miles away" may result in ignoring a missile launch 20 miles away too...

3. A short decay time, say 2 seconds, means dealing with new threat indications every time the enemy radar scans off and back on you. (ie F/A-18 symbol pops in and out of ur RWR scope becuase his radars scan is longer than your RWRs decay time)

Its just not practical. Especially when trying to correlate "new hits" to the same RWR threat.

HB's F-14 RWR is modeled very well.

DCS does generally do a poor job at modeling RWRs, but your points are wrong.

 

1. Way off on FCR beam widths. And because of the inverse square law, RWRs can detect much more than the main beam of the FCR. But that's dependent on range. Also, multi-path can cause detections.

 

2. DCS F-15 and SU-27 RWRs will straight up indicate the targets range. IRL, RWR ranges are very course estimates that do not account for atmospheric attenuation or the noise of the local EM environment. The RWR may not even know if it detected a main beam, sidelobe or multi-path reflection. Or perhaps you're in a climb and the signal came into your RWR antenna's sidelobe and were thus weaker.

 

Ignoring Missile launches "beyond 40 miles away" may result in ignoring a missile launch 20 miles away too...

 

3. A short decay time, say 2 seconds, means dealing with new threat indications every time the enemy radar scans off and back on you. (ie F/A-18 symbol pops in and out of ur RWR scope becuase his radars scan is longer than your RWRs decay time)

 

Its just not practical. Especially when trying to correlate "new hits" to the same RWR threat.

 

 

HB's F-14 RWR is modeled very well.

I didn't think about sidelobes, but that makes sense.

For 2, I get launch warnings from aircraft that are 100+ NM away. Pick whatever number you want, eventually there should be a cut off where the RWR can't distinguish it from background noise.

For 3, I wasn't clear, but I'm talking about missiles, launches, and spikes. All things you would know have dropped off quickly. I do agree with your point about nails. One of the big offenders is the missile radar indication will continue for 15+ seconds after the missile has passed you by and is no longer pointed at you at all (180° behind the missile).

All nice and dandy but do you have any idea of the current implementation of the workings of ED's DCS RWR piece of software code?

Having been in the navy and worked with ESM equipment, the whole RWR thing is classified and nobody will tell you exactly how it works. Sure, plain physics can be used but I think you're up for a big suprise if you found out how ED has implemented the RWR.

My opinion on the current RWR implementation is that it's a simplified model of emission and detection based on a few parameters which, when true to some statement, gets shown on the display. I don't think it takes PRI/PRF frequency, EIRP or any other physical quantity into account but path loss.

Everything above is my own opinion, YMMV.

[

2. Radar strength and rejection - It stands to reason that RWRs have logic filters that would eliminate physical impossibilities from being displayed. Radio emissions follow the inverse square law. RWRs can detect the strength of the emission and determine the range, and thus threat level, of the source radar based on a threat catalog. Likewise, this catalog would know the maximum possible detection and launch ranges of the platforms involved. This mean an RWR should, for instance, reject a launch/guidance warning from a source radar that is beyond a reasonable distance to engage the aircraft. For example, an F/A-18C that launches an AIM-7 would generate a launch warning for any aircraft that received sufficient energy to indicate that the F/A-18C was at a reasonable range to guide a semi-active missile. Allowing for error and unknown factors (such as missile improvements), this range would likely be 40 NM at most (since a reasonable range for the missile is at best, 20 NM).

The RWR has no distance information whatsoever. It just has signal strength.

What it then can do is, match the signal against a database of known radar types guess, which radar type it is. And once you know it's say an AN/APG-65(5) you might have the data about emission strength in relation to distance in your RWR db and this estimate how much of a threat it is.

But highly classified and no real idea. But there must be a reason why business or ELINT jets are always shadowing radar / missile tests :-)

BR,

Deadpool

All nice and dandy but do you have any idea of the current implementation of the workings of ED's DCS RWR piece of software code?

The RWR in game resolves the type of radar from its type library which simply relates the symbol to an in-game object (ie. if it's an F-18, it will show '18') and if course, if there's no entry, it shows unknown.

There is zero RF-type processing. There is a basic signal strength calculation based on the distance between the two objects.

I didn't think about sidelobes, but that makes sense.

 

For 2, I get launch warnings from aircraft that are 100+ NM away. Pick whatever number you want, eventually there should be a cut off where the RWR can't distinguish it from background noise.

100nm isn't that cut-off.

For 3, I wasn't clear, but I'm talking about missiles, launches, and spikes. All things you would know have dropped off quickly. I do agree with your point about nails. One of the big offenders is the missile radar indication will continue for 15+ seconds after the missile has passed you by and is no longer pointed at you at all (180° behind the missile).

That's a good point, these types of signals should be dropped quickly. But this isn't necessarily the RWRs fault either - all it does is receive a 'signal type' flag from the offending object, and there's a radar memory for that object that may continue to trigger the RWR.

This is a little inside of how it's done in F-14 module:

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

The RWR in game resolves the type of radar from its type library which simply relates the symbol to an in-game object (ie. if it's an F-18, it will show '18') and if course, if there's no entry, it shows unknown.

 

There is zero RF-type processing. There is a basic signal strength calculation based on the distance between the two objects.

Actually HB do some "ish" processing based on signal strength and antennas. They had a big post in the F14 forum on how they are doing it.

Not like thats classified, I have fricking USB dongle that does it...

ETA, and beat by one post!

Discussing about inverse square law doesn't work when you have a beam shaping/focusing done.

If you have a simple radiating point, like a light bulb. Then inverse square law explains the light behavior correctly.

But if you place a parabolic mirror behind that light bulb, beaming that light like a spotlight. Inverse square law doesn't anymore apply, you will have much stronger emission on distance and far less drop in power per distance.

Radars are not pointy radio emitters. But shaped by various means, and inverse square law doesn't apply as is. As well inverse square law doesn't apply to energy that is reflected.

Yet, the RWR systems are designed to be far more sensitive than the threat radars would be at their maximum power to detect the aircraft with the RWR. Example the SPO-15 is designed to be 125% sensitive to that radar maximum power, requiring the radar to receive and echo to detect the aircraft. So RWR in hypotheses has 25% more range to detect an radar before radar can detect it.

And as we don't have radar emissions in most aircrafts. The RWR and radars (including missiles radars) are in most modules "instantaneous" by knowing what is where. This is as well problem with missiles radars that are instantly knowing where targets are and what they do without any search patterns or logic to observe targets first.

The inverse square law applies for radars as well as for any(?) other antenna.

It is just that you always have to be at the same azimuth/elevation in the antenna diagramm when you make the simple power calculation.

inverse square law applies to all radiation. radars included, its a universal rule, and no fancy antenna lets you get away from it, AESA or otherwise. For those specific focusing antennas, all they do is up the gain and focus the power in the beam, IE make less of its energy spill out, it still loses power at the same rate due to the inverse square law.

Best way to think of it is if I put 1 unit of power through an omnidirectional (non focused) antenna, then if I measure the power 1 unit of distance away in each cardinal direction I would get 1/360th of the power output over each degree (since its spreading all that power over 360 degrees)

If I put 1 unit of power through a 1 degree focused antenna I then should get 1 unit of power 1 unit of distance away (all the power is in 1 degree now) the inverse square law still applies (2 units of distance away I will get 1/4th power, 3 units of distance 1/8th, etc...)

This is beamwidth in a nutshell (which also isnt perfect, beamwidth is just an arbitrary definition, there will always be spill over, sidelobes, and backlobes, no perfect antennas)

Radars also dont lose power to the inverse square of the range, they actually lose it to the inverse 4th power of the range, this is because the radar energy doesn't just have to get to the target, it also has to get back.

Just google radar range equation

here https://copradar.com/rdrrange/ Look down at the bottom of the page, lays it all out, and this 100% applies to every radar in existence, no fancy antenna lets you get away from this.

if you want to derive it https://www.radartutorial.eu/01.basics/The%20Radar%20Range%20Equation.en.html

I read the entire thread because, as anyone who's flown into big battles knows, the RWR is almost useless for reasons stated in the OP. As always there is a balance between "realistic Vs classified Vs known vs practicable coding"

In my humble opinion, range/signal strength aside, a very useful change would be to reduce the emitters locked radar beam width, and/or reduce the RWRs tolerance to class off-beam (say >3') detection as a spike.

I'm am sure there are complex real world examples where a radar lobe reflects off the ground and some other unrelated aircraft got spiked, but this is not going to be simulated, along with every molecule here. I I'd an aircraft at bulls has locked up an aircraft bulls 360 for 50, then a 3rd aircraft, bulls 050 for 100 ideally wouldnt be getting spiked (ideally imho). Can argue all day if it should be 1' wide or 3' or 5' but 90'+ even 180' is a little excessive imo

Agreed the tolerances do need looked at but certain extreme cases are probably not worth simulating as other things are just more important.

The RWR is hardly useless rn. Can it be improved? Sure, but its far from useless.