Is losing Radar Lock still a thing?

[jojo]

Thanks for the info, it is nice to know it :thumbup:

 

So I was wrong, it's not Doppler filtering but side lobe filtering ?

 

I would say that the doppler effect is used to filter:

- ground returns : closing velocity = ground speed

- side lobe : closing velocity = 0

 

Unless we're proved wrong :smilewink:

[myHelljumper]

I would say that the doppler effect is used to filter:

- ground returns : closing velocity = ground speed

- side lobe : closing velocity = 0

 

Unless we're proved wrong :smilewink:

 

Thanks, I will look a bit more into side lobe filtering :thumbup:

[sedenion]

So it's not bug, it's a feature, and not specific to M-2000C = side lobe filtering

 

Maybe ! But, could we have some confirmation from developers ? I think communication about current radar implementation, programmed rules about ground clutter, atmosphere, closing speed, etc, could definitely dismiss some doubts and sterile debates and help everybody to clearly state and recognize "Yes, it is regular, it is feature, it is coded as this". I don't think my request is groundless considering the amount of debates about radar lock loss.

[jojo]

Maybe ! But, could we have some confirmation from developers ? I think communication about current radar implementation, programmed rules about ground clutter, atmosphere, closing speed, etc, could definitely dismiss some doubts and sterile debates and help everybody to clearly state and recognize "Yes, it is regular, it is feature, it is coded as this". I don't think my request is groundless considering the amount of debates about radar lock loss.

 

Ask ED the radar code then...

 

M-2000C is a module of DCS World, not a stand-alone sim.

[p1t1o]

OK. You described the main point of the doppler radar, and what you said about it is correct regarding the radar main lobe.

Now you have to learn about secondary lobes / side lobes.

Annnnd there it is. You get the reason why "iso-speed" (relative to the aircraft) targets are a pain in the ass, as are "stationnary" (relative to the ground) targets.

 

I would advise Sedenion to look at this too, since he's "probably" :D not gonna see the code anytime soon ;)

 

++

Az' :)

 

PS: That said, I don't rule out the possibility of an un-described bug to still be there. I'm also advocating for short tracks to be provided, as it makes the discussion far more concrete. And AFAIK, there is still one (known/unresolved) bug afflicting "PID lock + aircraft maneuvering frankly" situations.

 

 

Besides which, in my example mission, the target is a slow-mover which i made no effort to match speeds with, its unlikely that I ever spent much time at zero doppler shift. Certainly not often enough to explain the difficulty in maintaining a lock.

 

Not for nothing, but when Im in an M2000, I am very rarely matching speeds with a 250kt C-130 target. And still dropping locks. I was never really talking about an "iso-speed" engagement, only answering queries about approaching from the rear quarter.

 

I'd also be very surprised to learn that "side-lobe filtering" was simulated in DCS, given the simplicity (no criticism) of the simulation of things like ECM, countermeasures etc.

 

 

Any chance, to avoid me having to learn an entirely new field and perhaps acquiring the wrong information, you could explain in a few sentences how "side lobe" = dropping locks from "iso speed" targets?

 

Cheers! :)

[Azrayen]

One of the side lobe is directed below the aircraft. Hits the ground. Comes back.

This creates an echo seemingly staying at iso-distance (assuming ~ level flight) = iso-speed.

This has to be filtered out.

 

You may now be surprised all you want :)

[sedenion]

Ask ED the radar code then...

 

M-2000C is a module of DCS World, not a stand-alone sim.

 

Which is not an argument since we don't know the ED's API, how Razbam used it and how much. If Razbam can't tell because something is in the ED side, so he said "Sorry, i don't know, this is on the ED side"...

[TomCatMucDe]

Which is not an argument since we don't know the ED's API, how Razbam used it and how much. If Razbam can't tell because something is in the ED side, so he said "Sorry, i don't know, this is on the ED side"...

 

We are customer's of Razbam on the Mirage not ED. If somebody has a question he goes and ask his supplier, not the supplier of the supplier.

[p1t1o]

One of the side lobe is directed below the aircraft. Hits the ground. Comes back.

This creates an echo seemingly staying at iso-distance (assuming ~ level flight) = iso-speed.

This has to be filtered out.

 

You may now be surprised all you want :)

 

Colour me surprised! Makes sense.

 

Is this based on knowledge of events/equipment, or is it an attempt to explain the behavior we are seeing in DCS?

 

It does seem like a pretty YUGE weakness of air-to-air radar - so much that Im surprised that I dont drop MORE locks - that I've never heard of before, it would seem significant enough to be a major problem in pretty much all modern air combat then? Is it a well characterised problem for those people who work in the field? I dont mean side lobes, I mean the filters making it hard to maintain locks.

 

I think i said before, I dont see it as very likely that DCS is simulating side lobe effects.

[GGTharos]

Multi-path issues (That's what this is) have been a 'thing' since radar as a thing, and so have been solutions to them.

 

I would be very surprised if the RDI isn't able to just blank out the side-lobe without blanking out the real target. I don't believe that specific closures are used in this process; it sounds more like a doppler blind-speed which is actually eliminated by applying some parts of ECCM techniques already used to blank out the side-lobe returns.

[jojo]

Multi-path issues (That's what this is) have been a 'thing' since radar as a thing, and so have been solutions to them.

 

I would be very surprised if the RDI isn't able to just blank out the side-lobe without blanking out the real target. I don't believe that specific closures are used in this process; it sounds more like a doppler blind-speed which is actually eliminated by applying some parts of ECCM techniques already used to blank out the side-lobe returns.

 

And what is your theory about the same issue on DCS F-15C ?

 

Because we aren't discussing real life radar, but DCS simulation of AA radar...

[GGTharos]

My 'theory' is that they all use the same radar code, tuned to represent the issues of of the Su-27 radar - and that's very, very old news.

 

The F-15C shouldn't have any serious issues with doppler blind speeds since it used pulse compression/chirping and staggered PRFs to eliminate all these issues - essentially these are basic radar techniques ... nothing new or special. You just need enough processing power for them.

 

So my conclusion is that the AA radar code isn't configurable enough at this time to represent differences between radar sets other than setting a couple of basic notch gates and detection ranges.

[red_coreSix]

Also the F-15C has its guard antenna mounted on the gimbal system, where as they're fixed on the Su-27, for whatever reason. IIRC the effect of that is actually represented in the sim, with the flanker loosing targets when rolling above a certain limit at low altitude.

[jojo]

My 'theory' is that they all use the same radar code...

 

Great, finally we agree. So what I said...ask ED :music_whistling:

[Azrayen]

@p1t1o:

 

I dare to think I'm a somehow knowledgeable person :smartass:

 

And no, I don't believe the radar code actually modelizes the whole complexity of the side lobes. Just one of their main noticable characteristics (operationnaly), which is a 2nd threshold of target speed where you'll get a hard time seeing/locking it.

 

This also all depends on PRF, and RDI offers its full capacity (range, lock ability) in HPRF. But this is also the setting where the "blind spots" are the most sensible.

Plus it cannot lock (STT) in ILV or LPRF modes (this is specific to the 2000, not inherited from elsewhere; and accurate).

[GGTharos]

Not being able to lock in ILV or LPRF sounds strange. The radar would just select the best PRF according to parameters, but always try to transition to HPRF for guidance. Is there some sort of explanation for not being able to lock while in these other modes? Just curious.

 

As for blind speeds, as I said ... I'd be surprised if the RDI lacks PRF pulse manipulation to deal with it ... and if it doesn't, then the blind speeds aren't represented correctly. They'd be a harmonic - a multiple of the lowest speed at which the blind speed occurs.

[Vatikus]

Since there are dozens of bugs regarding radar display, we cannot be sure if these are isolated only to the display representation.

[jojo]

Ok, since everyone seem to doubt about sidelobe effect modelisation in DCS radar, here is an extract of DCS F-15C manual:

 

Pulse-Doppler radar thus depends on "closure" (i.e. target approach velocity) to discern low-altitude

targets against the background of the earth. Aircraft on the defensive can often break a pulseDoppler

radar lock by a tactic called "beaming" or "flying the notch," which consists of flying on a

trajectory perpendicular to the hostile radar beam. The defensive pilot observes the threat radar on

the aircraft's radar warning receiver (RWR) display and flies to place that threat at the "three-" or

"nine-o'clock" position. The defensive fighter is then flying neither towards nor away from the threat,

and its closure is the same as that of the surrounding terrain in a look-down geometry, or any

deployed chaff countermeasures in a look-up geometry.

 

The rate of closure of the surrounding terrain effectively generates a primary "notch" in the radar's

sensitivity, due to ground-reflected signals ("clutter") received along the axis of the main radar beam.

Target signals in this "look down clutter notch" are rejected by filtering as if they were ground clutter,

allowing beaming targets to break a radar lock. Antenna focusing is never perfect, however, and

some transmitted energy also spills out in unintended directions called sidelobes. This energy can

also be reflected from the ground, and re-enter the antenna from the sidelobe directions. If a fighter

is flying at low altitude, signals reflected from the ground may enter the radar and appear on the

scope as additional clutter, with a closure equal to the fighter's rate of climb or descent, and a range

equal to the fighter's altitude. If the fighter is in pursuit against a fleeing target travelling at the same

speed and range, the target signals may become lost in sidelobe clutter, breaking the lock. This can

create a secondary "notch" in the fighter's radar sensitivity.

 

Sidelobe clutter is usually filtered out ("compensated") with the help of a small "guard" horn antenna.

The guard antenna is designed to be more sensitive than the main antenna in sidelobe directions, but

less sensitive along the axis of the main beam. Signals received on the main and guard channels are

then compared and rejected as sidelobe clutter if they are stronger on the guard channel.

The guard horn is attached to the slotted array in flat-plate radar antennas like the APG-63 and scans

together with it for good compensation in all scan directions. In Russian Cassegrain radars like the

N019 and N001 however, the guard horn is not attached to the scanning reflector but is rather fixed

and aimed in a downward direction. Banking the fighter at low altitude during a radar lock on a

fleeing target can thus rotate the compensation horn away from the ground, degrade the sidelobe

compensation and break the lock due to ground clutter. During normal scanning operation in search

mode, the entire radar Cassegrain antenna housing is roll-stabilized on a rotating gimbal to keep it

oriented with the horizon. In this mode, search targets can be lost from the scope if the fighter roll

exceeds the limits of the rotating gimbal (110-120 degrees angle of bank). MiG-29 and Su-27 pilots

thus need to make careful decisions about operating altitude during an engagement, since high

altitudes reduce sidelobe clutter to maximize their radar performance, but also allow look-down

targets to more easily break lock by notching. F-15C pilots enjoy fewer restrictions in radar

performance, and might make such decisions based instead on the effect that altitude has on missile

performance.

[jojo]

Since there are dozens of bugs regarding radar display, we cannot be sure if these are isolated only to the display representation.

 

A few ones but not that much, rather some people who are chocked to discover that a radar isn't a perfect sensor. :music_whistling:

Edited June 17, 2017 by jojo

[Vatikus]

A few ones but not that much, rather some people who are chocked to discover that a radar isn't a perfect sensor. :music_whistling:

If the fighter is in pursuit against a fleeing target travelling at the same speed and range, the target signals may become lost in sidelobe clutter, breaking the lock. This can create a secondary "notch" in the fighter's radar sensitivity.

I fully understand that radar should not be all seeing eye, etc. Thing is, in dcs m2k lock drops as well in high aspect angle with plenty of closure.

[jojo]

In TWS maybe, that's the remaining bug and already known.

I never encountered the issue in STT.

[Azrayen]

Thing is, in dcs m2k lock drops as well in high aspect angle with plenty of closure.

The dev (an me, because I'm curious) would probably need a track to see what was the target doing, and what was the fighter doing at the time of the drop lock.

 

Then precise assessment (i.e.: bug or feature?) can be made.

 

Unfortunately, I haven't found the time to do specific tests about that. I imagine a multiplayer test so that maneuvers are briefed (target & fighter).

 

Regards,

Az'