Noise rules

The working title for the air grid is “Air Warning System”. I suggested “Airly Warning System” but I don’t think Gophur got the joke.

Objectives: Kill skirting, Don’t kill strat bombing, Air combat attractor (dateline for fighter jocks).

Some of you will immediately insist that the 3rd isn’t needed, that you fly to the frontline and strafe opels or laffys until enemy planes come. Thank you for making the point for me :)

I’ve got to work on “Theseus” first (working title for the under-the-hood work) and a few client-support tasks (supporting a new system for managing spawn objects to let troopers spawn at AFs, trucks/atgs/etc spawn at depots, etc).

We’re going to start with something pretty close, as it turns out, to what I’d previously portrayed: a global, strat-independent, low-fi, unilateral, “blip” system based on fixed map squares of probably 16x16km with at least 60 second intervals between updates.

An important element of this concept is noise. Someone suggested that when everyone has radar the map will be less noisy than shown in my previous depictions. I don’t think that’s true. There’ll still be people bombing bridges, people going to specific towns, etc, etc – and on top of that different pilots will choose different enemy blips to fly towards or intercept, plus pilots returning to base, afk, etc. If you sit and watch for a few minutes, you might get an idea of where planes are at, but you’re going to need to watch it for a good 3 minutes before signal begins to emerge from the noise.

23 Comments

god i can’t wait for this.

ATGs from depots will be real nice too.

It’d be nice if the presentation could be a user preference… I know some people like the “dots in corner” from AW, but I prefer the “semi-transparent color overlay” from the mockup.

Oh, I also vote for “AAG” – Air Activity Grid… to me warning implies some kind of correlation to targets that must be defended, and an active alerting. Not that it really matters, though. ;)

It was me who said there would be less noise. What I actually said was pilot behavior will change and there might be less noise. I’m going to stand by that point, but preface it by saying any reduction in noise will only be marginal. Other new activities will take place that will offset it but it will be more geographically dispersed.

16×16 sounds much better. 20×20 was definitely too low rez based on the sample grid you provided.

When can AWS be expected to go live?

He’s got to spec it out and see what he thinks the effort will be. Slate it with the 1.28 release and see if it can make the 1.28 time line with everything else being added to the 1.28 release.

This should be easier than TOE’s ;)

Surly Warning System?
Gurly Warning System?

Trucks spawning @ Depots would make HAAC obsolete, perhaps it is time to allow truckers to create their own mini spawn list of equipment.

Trucker Load-Outs could be an interesting concept.

Zheriz said: 16×16 sounds much better. 20×20 was definitely too low rez based on the sample grid you provided.

Me said… 8×8 sounds much better… due to the 4Km limit of visible range in the 3D space… considering that just flying 5K behind you, this game denies visual contact.

8x8km allows lone, single fighters to find bomber formations with ease. 16×16 means sending several aircraft to find them without simply handing them over.

Again: Not radar.

I agree with you Cid. In a previous post I made an argument for a smaller grid as well. Based on what KFS1 said about fast binary calcs, the options are 8×8 (8192×8192) or 4×4 (4096×4096), the two next smallest sizes. If the delay is going to be 60 seconds, then higher resolution will definitely be required. The way I see it, the AWS delay time and resolution are inversely related. The longer the delay time, the sharper the resolution will have to be and vise versa. If not, it won’t be terribly useful. The reason is as follows:

The Pilot vis limit is 4km, as you said. So at 100% visibility, the pilot can see 360 degree spherical volume of 150km at any given time under optimal conditions (no clouds, etc). In reality, the pilot’s vision is reduced by clouds, ground clutter, system config, human eye perception reasons, lag, etc. In other words, the pilot actually sees less than the 4km limit. So let’s assume a modest visual degradation of 15%. The pilots visible volume drops from 268km to 228km.

If we compare that value against the search blocs, we find:

At 100% vs search volume:
4x4x10 grid is 168** cubic km vs 268km pilot visibility, so 4×4 is clearly too small.
8x8x10 grid is 671** cubic km. So the pilot sees 40% of the volume at a time.
16x16x10 gris is 2,684 cubic km. So the pilot sees 10% of the volume at a time.

**using the binary value

At 85% vs search volume:
8x8x10: 34% of volume.
16x16x10: 8% of volume.

The Rats plan to use 16×16. Not including the 60 second delay, noisy environment, and motion of the target, which will all make the odds significantly worse, at 16×16 the pilot has a 1 in 12 chance of finding the target under perfect conditions. Considering everything, the real odds will more likely be 1 in 30 or worse. At those odds most pilots will come to feel that AWS is not particularly helpful and it won’t have the effect the Rats are looking for…except for stopping skirting.

Solution?
Now I’m not saying I’m absolutely positively right because there are a lot of assumptions, but they’re reasonable assumptions. So the outcomes would be:

* The 4×4 solution would bring us into AWACS territory. We’d be looking at near perfect intercepts. Clearly not the way to go.

* The 8×8 solution would give at best 1 in 3 odds, realistically more like 1 in 10. This is much more reasonable and AWS would have some utility for pilots. 8×8 is probably the best way to go, especially if the Rats use a long 60sec delay.

* The 16×16 would result in very low intercept odds and won’t have the desired effect. But the Rats can obviously start here and go to 8×8 as needed based on actual results.

I’m betting they’ll end up with an 8×8 solution unless other issues like bandwidth, etc, make it impractical.

Regards,

You say the Grid will cover the entire map (From what I read here I understand that to mean exactly what it sounds like that the AWS will cover to all four corners of the moldeled world).

As far as the grid naming will go (I know presentation question and not your department) i.e n-s = letters e-w= numbers (or somthing like that).

My concern is that the areas of the map that are going to be referenced most will be somewhat odd designations ( we will proably never see anyone refence row a,b,c or column 1,2,3 unless some bombers are on a serious side trip)

Any thoughts on this?

Also any plans on tying this into the current text based warning system.

It would be really cool if players could pick which grids they wanted to get text chat alerts about and probably the only way to do it without inundating the players text channel

I think those probabilities are much smaller than reality. It only calculates the probability that the two planes would see each other if they were sitting stationary in random places in the same cell. But they are not stationary, they move. That means that the chance of the paths intersecting are larger.

It is a mean free path problem. With two planes in a cell of 16x16x5 km the plane density is 0.0016 km^-3. The cross-section of detection is pi*3^2 = 28 km^2. The mean free path of planes in the volume is then 1/(0.0016*28) = 22.6 km, comparable to the size of the cell.

This means that on average you will detect the other plane just flying through the cell, and this assumes random motion. A smart pilot with some idea of what’s going on (likely altitude and heading, for example) will probably have a much higher detection efficiency. Anything smaller than 16km will be a giveaway, it seems to me.

Surly Warning System?

Winner!

At 16 x 16, a ‘picket line’ of five aircraft, near max vis distance from each other, should be able to completely visually cover a 16 km line.

…first, the viz limit (radius) is 3km or 4km?? Either way Bloo, your math is a bit off. =P If the radius is 3km, then the diameter is 6km. It would take 2.67 aircraft to picket a 16km line.

Lutorm,

It makes for a fun problem to solve, doesn’t it? :) I admit, particle physics isn’t my thing. I hadn’t heard about ‘mean free path’ until now. But after reading the Wiki article, I get it.

There’s definitely a ton of assumptions involved. You assume a 5km Y-axis (Y being altitude). I assume 10km, because the target could be a bomber. Once AWS is involved, bombers will be grabbing max alt to make interception as hard as possible. I did some research, and 10km is too high for our current set of aircraft. 10km would require a B-17 Flying Fortress, which is the alt number that was jumping out at me from the back of my memory. Our current bunch of aircraft have operating ceilings around 8.5km. If it’s fighters, then 5km is also valid.

There is a 30-60sec time delay, which would essentially act to make the mean free path volume larger than the static volume, reducing density, cross section, etc. Noise, target *evasion* and spoofing also can’t be accounted for in mean free path or in my solution either. And you made a good point about smart pilot optimization.

I’m not saying either of us is right or wrong based on the math. The answer is probably somewhere in between. It was just fun to put some numbers to it. We’ll only know by observing the actual results.

The math does drive us to one irrefutable mathematical conclusion, the Rats need to add B-17 bombers. *grin*

These are just order of magnitude estimates, it’s true. I actually think things are going to be even better, because while the bombers are going to go as high as possible, the fighters know that, too. Noone will go hunting an inbound bomber at 2k alt, because they know they will be at 5-8k.

Bottom line is, I think 16km will be perfectly adequate. Remember, we WANT the fighters to have to search a little. I don’t think 5-10 minutes looking for an inbound flight is out of the question.

My math isn’t off. If you can’t see the next plane in the picket, you can’t be sure your picket is still even when using the eyeball (you can use the map, but then you’re going to miss the enemy).

3km – plane – 3km – plane – 3km – plane – 3km – plane – 3km – plane – 3km = 18 km wide picket (overlap, naturally, as 4 could only cover 15 km).

Erh, yes, the diameter is 6km – that’s the distance between the plane on your left and your right. However, you’re half way between them both, which would be whatever 6 divided by 2 is. Fearing my math skills, I’ll leave that to you.

Since you can see circles out to 3km maybe you would allow 4km but any further and its not a picket line.

You’re assuming the picket planes need to see each other in order to form a perfect picket, I’m not making that assumption. The picket doesn’t have to be perfect. The whole grid can be swept in a minute or so, there’s no need for such a tight nit picket. 5 planes is harder to assemble than 3, so 3 is more ….yada yada yada…. who cares, not gonna happen.

burp

Gophur says “grid A1 will be the upper most corner of the entire theater.”

A1? Urgh, don’t you have any coders working there? A0, please!

Coders gotta be kept in check. Otherwise it will be grid E100101 600km east, or even worse, grid EF0xEF. :-)

Damnit, I wanna be in grid reference 0xDEADBEEF

Black Bishop-3 to Queen-5! Black Bishop-3 to Queen-5!

damit maverick, engage!

Leave a Reply

Name and email address are required. Your email address will not be published.

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

You may use these HTML tags and attributes:

<a href="" title="" rel=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <pre> <q cite=""> <s> <strike> <strong> 

%d bloggers like this: