Given some folks' reports of their airborne early warning & control (AEW&C) or AWACS aircraft proving vulnerable to 'sneak attacks' or generally not being able to spot enemy targets, I'd thought I would run a few tests through HCE 2009.050 and HCDB-110304 and post the results.

 

The premise being that AEW&C and AWACS are not the "all seeing eye of Sauron" that some might expect them to be, nor are all of their potential targets, foes and aggressors created equally.

 

I chose four different AEW&C platforms (with some similar and some varying capabilities), and a variety of radar 'targets' (again, with some similarities and some differences between them) for the testing.

 

First the AEW&C platforms:

 

1. E-2B Hawkeye (circa 1970),

w/APS-120 radar (AS (360 deg) type, SS (airborne) type, max range 238 nm, probability of detection 60%)

 

2. E-3B Sentry AWACS (circa 1977),

w/APY-1 radar (AS (360) type, SS (360) type, LD/SD, HF (3D), max range 350 nm, PD 70%)

 

3. E-3C Sentry (RSIP) AWACS (circa 2001),

w/APY-2 RSIP radar (AS (360) type, SS (360) type, LD/SD, HF (3D), max range 350 nm, PD 80%)

 

4. E-2D Advanced Hawkeye (circa 2010),

w/APY-9 radar (AS (360) type, SS (360) type, LD/SD, HF (3D), max range 400 nm, PD 80%)

 

And now the radar 'targets', starting with the airborne stuff and their radar cross section (RCS) values:

 

F-117A Nighthawk: RCS 95

Rafale C (F4): RCS 114

F-16C/D Block 50/52 CCIP: RCS 131

F/A-18C/D (USMC/CW): RCS 136

MiG-21MF Fishbed J: RCS 137

MiG-29 Fulcrum A: RCS 142

MiG-23MLD Flogger K: RCS 142

Tornado F.3: RCS 143

Su-24M Fencer D: RCS 148

F-15C Eagle: RCS 150

Su-27 Flanker B: RCS 151

Tu-22KD Blinder B: RCS 158

 

Some surface ships:

 

Fishing vessel: RCS 199

Sovremenny destroyer: RCS 215

Merchant: RCS 224

Supertanker: RCS 227

 

And finally, some land units:

 

SA-16 Team: RCS 0

155mm SP artillery battery: RCS 99

SAM SA-6a Bty: RCS 99

T-72 tank: RCS 99

Truck convoy: RCS 105

Bunker (surface) RCS 175

Small structure: RCS 175

Large structure: RCS 230

 

RESULTS (game results compared to GE formula calculations for theoretical maximum range):

 

Notes:

1. Ranges are given in nautical miles (nm)

2. All airborne targets detected at Low altitude except where noted

3. Theoretical maximum possible detection range may exceed the range horizon

4. Emitters (e.g. SAM batteries) typically have their detection range firmed up by ESM (where available) well before being detected by radar

 

Versus Airborne Targets:

 

E-2B Hawkeye performance

 

F-117A: 16 nm (maximum possible detect range 23.38 nm)

Rafale: none, missiles first (max possible 44.49 nm)

F-16: 49 (max possible 73.8 nm)

F/A-18: 39 (max possible 85.06 nm)

MiG-21: 57 (max possible 87.48 nm), target at Med altitude

MiG-23: 45 (max possible 104 nm)

MiG-29: 45 (max possible 104 nm)

Tornado: 46 (max possible 108.9 nm)

Su-24: 59 (max possible 134.76 nm)

F-15: 69 (max possible 146.46 nm)

Su-27: 74 (max possible 152.62 nm)

Tu-22KD: 89 (max possible 202.11 nm)

 

E-3B AWACS performance

 

F-117A: 30 (max possible 34.39 nm)

Rafale: 63 (max possible 65.43 nm)

F-16: 104 (max possible 108.53 nm)

F/A-18: 115 (max possible 125.09 nm)

MiG-21: 126 (max possible 128.65 nm), target at Med altitude

MiG-23: 152 (max possible 153.32 nm)

MiG-29: 152 (max possible 153.32 nm)

Tornado: 157 (max possible 160.14 nm)

Su-24: 174 (max possible 198.18 nm)

Su-27: 184 (max possible 224.43 nm)

F-15: 212 (max possible 215.38 nm)

Tu-22KD: 266 (max possible 297.22 nm)

 

E-3C (RSIP) AWACS performance

 

F-117A: 28 (max possible 34.39 nm)

Rafale: 61 (max possible 65.43 nm)

F-16: 103 (max possible 108.53 nm)

F/A-18: 123 (max possible 125.09 nm)

MiG-21: 126 (max possible 128.65 nm), target at Med altitude

MiG-23: 151 (max possible 153.32 nm)

MiG-29: 152 (max possible 153.32 nm)

Tornado: 157 (max possible 160.14 nm)

Su-24: 189 (max possible 198.18 nm)

Su-27: 186 (max possible 224.43 nm)

F-15: 200 (max possible 215.38 nm)

Tu-22KD: 269 (max possible 297.22 nm)

 

E-2D Hawkeye performance

 

F-117A: 37 (max possible 39.3 nm), target at Vlow altitude

Rafale: 71 (max possible 74.78 nm)

F-16C/D: 121 (max possible 124.03 nm)

F/A-18: 127 (max possible 142.96 nm)

MiG-21: 142 (max possible 147.03 nm), target at Med altitude

MiG-23: 172 (max possible 175.23 nm)

MiG-29: 171 (max possible 175.23 nm)

Tornado F.3: 182 (max possible 183.02 nm)

Su-24: 184 (max possible 226.49 nm)

Su-27: 182 (max possible 256.50 nm)

F-15: 240 (max possible 246.15 nm)

Tu-22KD: 327 (max possible 339.68 nm), target at Med altitude

 

Versus Ship Targets:

 

E-2B Hawkeye performance

 

Fishing vessel: 61 (max possible 63.84 nm)

Sovremenny destroyer: 90 (max possible 104.26 nm)

Merchant: 152 (max possible 152.62 nm)

Supertanker: 170 (max possible 172.41 nm)

 

E-3B AWACS performance

 

Fishing vessel: 93 (max possible 93.88 nm)

Sovremenny destroyer: 153 (max possible 153.32 nm)

Merchant: 220 (max possible 224.43 nm)

Supertanker: 240 (max possible 253.54 nm)

 

E-3C (RSIP) AWACS performance

 

Fishing vessel: 92 (max possible 93.88 nm)

Sovremenny destroyer: 153 (max possible 153.32 nm)

Merchant: 220 (max possible 224.43 nm)

Supertanker: 243 (max possible 253.54 nm)

 

E-2D Hawkeye performance

 

Fishing vessel: 107 (max 107.29 nm)

Sovremenny destroyer: 173 (max possible 175.23 nm)

Merchant: 245 (max possible 256.50 nm)

Supertanker: 245 (max possible 289.76 nm)

 

Versus Land Targets:

 

E-2B Hawkeye performance

 

SA-16 Team: none (n/a)

155mm Battery: none (max possible 4 nm)

T-72 tank: none (max possible 4 nm)

Truck convoy: none (max possible 6.01 nm)

SA-6 Battery: ?? (max possible 4 nm)

Small structure: 85 (max possible 92.5 nm)

Bunker: 91 (max possible 92.5 nm)

Large structure: 245 (max possible 486.09 nm)

 

E-3B AWACS performance

 

SA-16 Team: none (n/a), spotting visually at 1 nm

155mm Battery: 6 (max possible 6 nm)

T-72 tank: none (max possible 6 nm), spotting visually at 3 nm

SA-6 Battery: ESM detection firmed up at 185 nm

Truck convoy: 6 (max possible 8.83 nm)

Small structure: 132 (max possible 136.03 nm)

Bunker: 136 (max possible 136.03 nm)

Large structure: 246 nm (max possible 714.83 nm)

 

E-3C (RSIP) AWACS performance

 

SA-16 Team: none (n/a)

155mm Battery: 4 (max possible 6 nm)

T-72 tank: 6 (max possible 6 nm)

SA-6 Battery: ESM detection firmed up at 163 nm

Truck convoy: none (max possible 8.83 nm), spotted visually at 2 nm

Small structure: 133 (max possible 136.03 nm)

Bunker: 136 (max possible 136.03 nm)

Large structure: 246 nm (max possible 714.83 nm)

 

E-2D Hawkeye performance

 

SA-16 Team: none (n/a)

155mm battery: 6 (max possible 6.86 nm)

T-72: ?? (max possible 6.86 nm)

Truck convoy: 10 (max possible 10.1 nm)

SA-6: ESM detection firmed up at 183 nm

Small structure: 155 (max possible 155.46 nm)

Bunker: 155 (max possible 155.46 nm)

Large structure: 246 nm (max possible 816.95)

 

I may fill in the missing results above if I get an opportunity.

Given some folks' reports of their airborne early warning & control (AEW&C) or AWACS aircraft proving vulnerable to 'sneak attacks' or generally not being able to spot enemy targets, I'd thought I would run a few tests through HCE 2009.050 and HCDB-110304 and post the results.

 

The premise being that AEW&C and AWACS are not the "all seeing eye of Sauron" that some might expect them to be, nor are all of their potential targets, foes and aggressors created equally.

 

...

A quick perusal of those results brings out two points that are particularly of interest to me (please correct me if I'm wrong):

 

1. AWACS platforms are fairly ineffective at locating ground targets, particularly mobile units, from any reasonably-required stand-off distance... and, in fact, are not much good at suicidal ranges, either.

 

2. Typical AI fighter types that I have observed (or presumed) to sneak up undetected on my AEW assets, particularly on AWACS, should have been detected at distances well beyond the range of their own missiles, and thus the situations I've seen remain puzzling.

 

(Note: In case I haven't been clear about it, let me emphasize that the "sneak-up" situation does not happen anywhere close to a majority of the times, yet it happens frequently enough that it indicates that something unexpected is going on - as opposed to being some rare aberration. I can't give any actual data on frequency, but I'd estimate it to be in the realm of averaging once in every two games).

1. AWACS platforms are fairly ineffective at locating ground targets, particularly mobile units, from any reasonably-required stand-off distance... and, in fact, are not much good at suicidal ranges, either.

 

Well, first, think about the nature of that target. Your typical mobile ground target is buried in clutter, surrounded by intervening terrain and other features, often camouflaged, sometimes slow moving, etc. HCE doesn't model much of that, of course, but we try and simulate it through adjusting the detection probability (PD) and RCS values.

 

AEW&C (the more useful term, as AWACS is rather proprietary) platforms are not really designed to search for ground targets. (See, for example, the E-8 JSTARS project, for something that is).

 

That said, they do have some surface search capability. In some instances, their capability in this regard is superior to that found amongst many other aircraft types. In all cases, however, the raw capability of the radar and the RCS of the target will rule.

 

Looking at the results posted, we can see, for example:

 

1. The old E-2B from the 1970s has a surface search capability, but not the 360 degree type common to many other more modern AEW&C platforms. It has poor performance against small RCS mobile units like artillery pieces, tanks, and trucks. It lacks ESM too, so detecting SAM sites, even actively radiating ones, is going to be a chore. Against fixed structures with large RCS, however, say a bunker complex or cluster of buildings, it can pick them at roughly 100 to 250 nautical miles. Not too shabby.

 

2. The E-3B Sentry, an early version of the AWACS, likewise does not have great performance against small RCS mobile land units. The ESM feature, however, lets it detect and firm up the location of threats like actively emitting SAM sites at long range (approaching 200 nm).

 

3. Getting even better now, the RSIP upgrade to the E-3C AWACS is still having trouble detecting small RCS mobile units at distance, but can usually pick them up close to their theoretical maximum detection range. Against active emitters, the performance is generally the same as the E-3B, which is very good. Large RCS fixed ground targets are detected way out there.

 

4. The cream of the crop, the E-2D Advanced Hawkeye, can pick up mobiles pretty much at their maximum theoretical detection range. It has no trouble fixing the location of active SAM sites, and ground structures are picked up at maximum range.

 

2. Typical AI fighter types that I have observed (or presumed) to sneak up undetected on my AEW assets, particularly on AWACS, should have been detected at distances well beyond the range of their own missiles, and thus the situations I've seen remain puzzling.

 

So, to be clear, are you saying that your AI controlled fighters are not being detected at ranges typical to those posted here? Are the AEW&C and 'target' platforms shown here representative of the ones you are using?

Great work, Brad!

Great work ... And thinking on the subjectm Off topic: somebody will be tomorrow in Madrid in the antiterrorism seminar ?

Totally off topic:

 

The all-seeing eye of Sauron failed to see the most important thing - the location of the Ringbearer!

(after wiping the drool off the keyboard)

 

So it looks like all units (except the stealthiest ones) are detected by the AWACS (except the E-2B) at decent ranges. And you should not be trying to detect ground units this way. E-2B doesn't do very well at all!

 

Thanks, Brad, for doing all this work! (In addition to your other work.)

 

If I understand this correctly, you calculated the theoretical maximum range of detection, and then let a unit move towards the AWACS in the game until the alarms went off. Of course, a fast attacker can close more than a slow one from the first possible detection range.

Given some folks' reports of their airborne early warning & control (AEW&C) or AWACS aircraft proving vulnerable to 'sneak attacks' or generally not being able to spot enemy targets, I'd thought I would run a few tests through HCE 2009.050 and HCDB-110304 and post the results.

 

The premise being that AEW&C and AWACS are not the "all seeing eye of Sauron" that some might expect them to be, nor are all of their potential targets, foes and aggressors created equally.

 

I chose four different AEW&C platforms (with some similar and some varying capabilities), and a variety of radar 'targets' (again, with some similarities and some differences between them) for the testing.

 

.............................................

snip

..........................................

Small structure: 155 (max possible 155.46 nm)

Bunker: 155 (max possible 155.46 nm)

Large structure: 246 nm (max possible 816.95)

 

I may fill in the missing results above if I get an opportunity.

A very comprehensive test indeed.

If there's any thought that some people aren't getting the same sort of results, could you (Brad) post the scenario(S) and a few of us can run the same test and see how consistent are the results.

My 'feeling' is that your figures don't surprise me relative to what I see on my machine.

Don Thmas

E-2B Hawkeye performance

 

F-117A: 16 nm (maximum possible detect range 23.38 nm)

Rafale: none, missiles first (max possible 44.49 nm)

F-16: 49 (max possible 73.8 nm)

F/A-18: ?? (max possible 85.06 nm)

MiG-21: 57 (max possible 87.48 nm)

MiG-23: 45 (max possible 104 nm)

MiG-29: 45 (max possible 104 nm)

Tornado: 46 (max possible 108.9 nm)

Su-24: 59 (max possible 134.76 nm)

F-15: 69 (max possible 146.46 nm)

Su-27: 74 (max possible 152.62 nm)

Tu-22KD: 89 (max possible 202.11 nm)

 

I could imagine questions arising about the above results, specifically the Mig-23 and Mig-29, why are the actual detects much shorter than the theoretical as defined by the radar RCS spreadsheet? It goes back to the roll of the dice, radar runs once every 30 seconds, PD on the radar is only 60 so you have a bad string of luck at the dice rolling table and detect ranges get downright scary.

 

But, note the F-117 detect, a fortuitous roll!

Brad and I were chatting on IRC last night and he pointed out a finding that I didn't notice in his report. The E-2B detected a large structure at 245nm using its 238nm range radar .

 

That led to two findings:

1. MLUs (mobile land units) weren't properly integrated into the 'new' radar model so their detectability on radar was too huge. That is now fixed in 2009.053 and the current beta (see the Beta Testing general area for a debug build of 2009.053 that you can lay over your HUCE/2009.050 install).

 

2. The HCDB RCS values for the land units were outside of the acceptable ranges leading again to detects at too great of a range, Brad will be modifying the database.

 

So that's two for detects being more capable than they should have been, and zero for failure of detection.

1. AWACS platforms are fairly ineffective at locating ground targets, particularly mobile units, from any reasonably-required stand-off distance... and, in fact, are not much good at suicidal ranges, either.

 

Well, first, think about the nature of that target. Your typical mobile ground target is buried in clutter, surrounded by intervening terrain and other features, often camouflaged, sometimes slow moving, etc. HCE doesn't model much of that, of course, but we try and simulate it through adjusting the detection probability (PD) and RCS values.

 

AEW&C (the more useful term, as AWACS is rather proprietary) platforms are not really designed to search for ground targets. (See, for example, the E-8 JSTARS project, for something that is).

 

That said, they do have some surface search capability. In some instances, their capability in this regard is superior to that found amongst many other aircraft types. In all cases, however, the raw capability of the radar and the RCS of the target will rule.

...

Thanks for the detailed dissertation! All that, including the detection range information, is pretty much what I intuitively would have expected the behaviors and performance to be. And it generally supports the bottom line that AEW&C types are not particularly the right tool for the job of hunting ground-based AD units, fixed or mobile... which is all I was saying in this regard - although, it does leave the question open as far as what available types are suitable for the task. (Ah, if we could always have a couple of RA-5Cs or JSTARS, then ground-pounding life would be good! ) Unfortunately, those types aren't available in most cases, so it leaves me wondering what available types are useful. So far, it seems like the A-6E's are the best bet, although -a bit surprizingly- so are the Vikings. Unfortunately, both types are quite vulnerable to the SAMs that they are hunting, due to range considerations.

 

2. Typical AI fighter types that I have observed (or presumed) to sneak up undetected on my AEW assets, particularly on AWACS, should have been detected at distances well beyond the range of their own missiles, and thus the situations I've seen remain puzzling.

 

So, to be clear, are you saying that your AI controlled fighters are not being detected at ranges typical to those posted here?

Yes, most definitely! But, n.b. that it is not the general case, and only happens every so often, intermingled with "normal" detection behaviors... and it often seems to afflict certain scenarios more than others.

 

Are the AEW&C and 'target' platforms shown here representative of the ones you are using?

I believe so. Again, we're talking about the WestPac demo scenarios here, pretty much exclusively. I believe the bridge scenario has the least-capable Hawkeyes of any of these scenarios, but conversely, I believe The Backyard has some of the more-effective AEW&C types mentioned. And I believe the target types you mentioned pretty much cover all the target types in these scenarios, so we should be on the same page in that regard.

So, to be clear, are you saying that your AI controlled fighters are not being detected at ranges typical to those posted here?

Yes, most definitely! But, n.b. that it is not the general case, and only happens every so often, intermingled with "normal" detection behaviors... and it often seems to afflict certain scenarios more than others.

To me this hints that things are working correctly because every happening is modified by probability. Maybe keep a paper and pen handy with a little table for these 'happenings' and just put in a tick or cross to indicate one of these happenings being either as expected or not. As the numbers start to acculutlate you'll get a correct idea of how things are going, its very hard to make an accurate gauge form just seat of the pants feel.

DonThomas

To me this hints that things are working correctly because every happening is modified by probability. Maybe keep a paper and pen handy with a little table for these 'happenings' and just put in a tick or cross to indicate one of these happenings being either as expected or not. As the numbers start to acculutlate you'll get a correct idea of how things are going, its very hard to make an accurate gauge form just seat of the pants feel.

DonThomas

 

Better yet, use the radar logging and have the game spit out the numbers (granted you'll have to sift through the results to compile a nice spreadsheet).

To me this hints that things are working correctly because every happening is modified by probability. Maybe keep a paper and pen handy with a little table for these 'happenings' and just put in a tick or cross to indicate one of these happenings being either as expected or not. As the numbers start to acculutlate you'll get a correct idea of how things are going, its very hard to make an accurate gauge form just seat of the pants feel.

DonThomas

 

Better yet, use the radar logging and have the game spit out the numbers (granted you'll have to sift through the results to compile a nice spreadsheet).

Briefing, please! (Either I don't know how to do that, or else I only know "radar logging" by some other name... )

Briefing, please! (Either I don't know how to do that, or else I only know "radar logging" by some other name... )

 

Run launcher

Check -l for logging, check radar for radar logging

Run game, the ge.log in the game directory will be extra large thanks to the additional radar logging

 

You'll see a number of packets of entries for each radar attempt:

102671 Search.c:860 - Radar Emitter=F-15K Slam Eagle Target=F-4EJ Kai, TargetRange=240, Radar LOS = (BaseLOS x weather_mod + 128)/256

102671 Search.c:861 - Radar LOS of 328nm = (328nm x 256 + 128)/256 --> Radar LOS=107712/256

102671 Search.c:867 - A RCS=150 TName=F-4EJ Kai AName=F-15K Slam Eagle Range=240 Die=50 aPD=65

102671 Search.c:940 - B RCS=150 ARng=92 Arng=150 SRng=92 Srng=150

102671 Search.c:1017 - Not Detected F-4EJ Kai

 

First line lays out the players and the distance between them (240nm)

Second line looks at radar line of sight (328nm)

Third line shows the percent chance of detect for the radar and the die roll

Fourth line shows the radar used by the F-15K has a max 150nm range in both air and surface modes. Against the F-4EJ the F-15K radar has a max detect range of 92nm

Fifth line says the F-4 was not detected.

 

 

An example of a detect:

102671 Search.c:860 - Radar Emitter=Tsuiki Target=F-15K Slam Eagle, TargetRange=138, Radar LOS = (BaseLOS x weather_mod + 128)/256

102671 Search.c:861 - Radar LOS of 246nm = (246nm x 256 + 128)/256 --> Radar LOS=60644/256

102671 Search.c:867 - A RCS=154 TName=F-15K Slam Eagle AName=Tsuiki Range=138 Die=18 aPD=50

102671 Search.c:940 - B RCS=154 ARng=144 Arng=200 SRng=0 Srng=0

102671 Search.c:944 - Target was previously tracked.

102671 Search.c:999 - Detected F-15K Slam Eagle with Air Radar of range 144 with PD 50, Roll=18

 

That radar with a low PD of 50 gets a fortuitous roll of 18, radar range 200nm can detect F-15K at 144nm and is within radar LOS so the detection happens. A second note about this one, the "Target was previously tracked." has the effect of not needing a die roll