Radar modeling

How is DCS radar actually modeled. Is there any info out there on it?

At present it's just a hit/scan system (ie line of sight at set ranges, with modifiers from aircraft for their "visibility" and if there is terrain blocking LoS)

There is supposed to be a new radar system in dev that is ray-based and a more realistic emulation of the real thing.

So I see pulse rep rates are "modeled" and people hint at things like doppler notching. Is that modeled?

So I see pulse rep rates are "modeled" and people hint at things like doppler notching. Is that modeled?

Oh yes, it is indeed!

Oh yes, it is indeed!

How well is it modeled? Is it a general thing or are there specific differences in radars currently in game?

There are specific differences. They're close enough together in practice though.

Generally speaking: The older the radar, the more susceptible it is for notching.

How is actual RCS handled? Is it actually modled from an aspect point of view or it more like an average number? I.e. a head on mig21 vs a side on mig21, versus the same for B52?

There is a hardwritten RCS value for each aircraft, but I don't know exactly how it is being handled. You can check the RCS value in the "entry.lua" file which you can found in the folder of each aircraft.

It's just a number. So yes, a B-52 and a MiG-21 will have a different RCS, but it it aspect invariant.

How is actual RCS handled? Is it actually modled from an aspect point of view or it more like an average number? I.e. a head on mig21 vs a side on mig21, versus the same for B52?

How is actual RCS handled? Is it actually modled from an aspect point of view or it more like an average number? I.e. a head on mig21 vs a side on mig21, versus the same for B52?

I would imagine that it would be too taxing to continuously calculate aspect variant RCS values and using a fixed RCS makes sense from a computing point of view. I do know, but it would not be that taxing to have a function of RCS with range. RCS of an aircraft doesn't change, but it's detection should be a function of range with respect to a radar type. Maybe someone here knows if indeed RCS is a function of range with respect to radar detection.

GGTharos : I like the photo in your signature block.

F104 for the win? :)

Anyway, no need to calculate varying RCS - you could have a 3D shape representing it instead. That's probably too complicated for this sim, not to mention not easy to obtain data, but you could have a base RCS processed for aspect-based variance with a reasonably simple calculation :)

As far as detection range goes, the game seems to use the basic radar equation.

It's just a number. So yes, a B-52 and a MiG-21 will have a different RCS, but it it aspect invariant.

Thats a shame, that should be really easy to do. Just run a RCS modeling on your mig beforehand (offline) from (all or at least most aspects) then just do a quick aspect check in game and look at a table for what it should be. You could also do it at different freqs if you wanted more "Realism" or do it in 5-10 degree increments if perfromance is a concern.

I would imagine that it would be too taxing to continuously calculate aspect variant RCS values and using a fixed RCS makes sense from a computing point of view. I do know, but it would not be that taxing to have a function of RCS with range. RCS of an aircraft doesn't change, but it's detection should be a function of range with respect to a radar type. Maybe someone here knows if indeed RCS is a function of range with respect to radar detection.

 

 

GGTharos : I like the photo in your signature block.

You don't calculate the RCS dynamically in game you do it offline, in game you just use a bit bigger lookup table based on aspect, and the table can be "rough" in 5-10degree or whatever increments make sense. You can do it at different freqs too if you want more realism. Range wise thats a simple RF power calc.

F104 for the win? :)

 

Anyway, no need to calculate varying RCS - you could have a 3D shape representing it instead. That's probably too complicated for this sim, not to mention not easy to obtain data, but you could have a base RCS processed for aspect-based variance with a reasonably simple calculation :)

 

As far as detection range goes, the game seems to use the basic radar equation.

That may be a better approximation. Cylinder + something that looks like wings. But maybe too simple.

Sure, I guess they could but I'm not convinced there's a real need for it :)

RCS is important in the real world but I would agree that a table lookup where RCS is cross referenced with range would be a simple and quick way to do it. The datum would be probability of detection as a function of radar type.

It would be nice to have aspect as factor though. And just have that as part of a lookup function.

This is old but I’ve always been really curious about it. What information would I need about a specific radar system to model it adequately in DCS? What are the minimum known pentameters you need to know for say a FC3 radar? What about something more complex?

10 hours ago, F-2 said:

This is old but I’ve always been really curious about it. What information would I need about a specific radar system to model it adequately in DCS? What are the minimum known pentameters you need to know for say a FC3 radar? What about something more complex?

The more you know the better. For a minimum implementation that just simulates basic A2A functions like searching and locking you need at least the info like radar type, output power, main antenna dish size, gimbals, patterns, additional antennas placement, angles, beam widths, basic info on capabilities, radar modes, PRFs, filters, how it behaves in relation to EW, radar displays, controls, fire control, weapon support... uhm, Everything!

On 7/21/2022 at 1:13 AM, F-2 said:

This is old but I’ve always been really curious about it. What information would I need about a specific radar system to model it adequately in DCS? What are the minimum known pentameters you need to know for say a FC3 radar? What about something more complex?

For the absolute minimum?

Instrumented range, scan zone(s), rough detection ranges for different sized targets, modes and presentation.

AI radars in DCS are incredible simple: a general type (air search/surface search/multirole), max detection ranges in upper and lower hemispheres, instrumented range, minimum radial velocity, distance at which the radar transitions from search/acquisition to track & vice versa and update rate is all they consider in their definitions.

For something more complex like the Mirage 2000, about as much as you can possibly get, draconus covers pretty much all of it, the only thing missing would be sidelobe distribution.

3 minutes ago, Northstar98 said:

AI radars in DCS are incredible simple: a general type (air search/surface search/multirole), max detection ranges in upper and lower hemispheres, instrumented range, minimum radial velocity, distance at which the radar transitions from search/acquisition to track & vice versa and update rate is all they consider in their definitions.

I'd add from memory: scan zones, ECM burn through range, simulated earth curvature horizon.

7 minutes ago, draconus said:

ECM burn through range

Only thing I'll say is that mimimum SNR to resolve a contact is probably a better metric, as more powerful jammers will produce shorter burn-through ranges. Fairly sure this kind of thing is already somewhat simulated, in that different aircraft will have different burn-through ranges.

8 minutes ago, Northstar98 said:

Only thing I'll say is that mimimum SNR to resolve a contact is probably a better metric, as more powerful jammers will produce shorter burn-through ranges. Fairly sure this kind of thing is already somewhat simulated, in that different aircraft will have different burn-through ranges.

That I did not know as far as implementation goes. Of course it should be like that but is it really in DCS? Ex. is Su-27 with two jamming pods better in ECM than MiG-29S with internal one?

Just now, draconus said:

That I did not know as far as implementation goes. Of course it should be like that but is it really in DCS? Ex. is Su-27 with two jamming pods better in ECM than MiG-29S with internal one?

I'm fairly sure something like it is implemented.

I doubt it goes into detail like the 2 L-005 pods on a Su-27 vs the internal L-201 on the MiG-29S (FWIW these are both DECM systems IRL, probably focused more towards track-breaking. Not sure if they have a noise jamming mode), but on initial inspection, there seems to be some differentiatiation (and I don't think RCS accounts for the change by itself).