Terminal attack

[CV32]

Ever since the Israeli Navy’s Z class destroyer Eilat fell victim to an SS-N-2 Styx in 1967, and even more so since the loss of the Royal Navy’s Type 42 destroyer HMS Sheffield to an Argentine Exocet in 1982, the prospect of sudden and devastating attack by a previously undetected and hostile missile has been a major headache for ship captains.

 

Reducing the potential warning time to mere seconds was the principal means by which early anti-ship missiles sought to guarantee their terminal approach to an intended target. This was achieved primarily by “sea skimming”, i.e. flying at such a low altitude in the terminal phase that enemy sensors (usually radar and/or the Mk 1 eyeball) could not distinguish the approaching missile from the background clutter of the ocean’s waves until it was too late to do anything about it.

 

In keeping with steadily improving soft kill (e.g. electronic countermeasures) and hard kill (e.g. missiles and guns, and no doubt, eventually directed energy weapons) defences aboard ships, the terminal phase of anti-ship missiles has become more and more sophisticated since those early days.

 

The original Harpoon missile (the Italian Otomat is another example) used a terminal "climb and dive" (often described as “pop-up”) maneuver where upon the missile climbed sharply at a certain distance from the target (said to be 1,830 meters) and then dove at a 30 degree attack angle. A high apogee pop-up attack was ideal against small or fast, maneuvering targets such as submarines and missile craft. Subsequent versions of the venerable Harpoon have eliminated the pop-up maneuver, improved upon it, and even added multiple terminal maneuver options that can be selected by the launch platform operator.

 

Many Soviet era anti-ship missiles (e.g. AS-4 Kitchen) employed a terminal dive at such a high speed (transonic or supersonic) and from such a high altitude (as much as 230,000 feet) that the attack phase was nearly vertical, and so, while the defender could often see the missile coming, he may have in any event been hard pressed to do anything about it.

 

Most recently, terminal maneuvers have progressed in complexity to such options as “s-turns” or weaving attacks, or in other cases, to an increased terminal attack speed. Both are, as always, aimed at making engagement by point defence systems as difficult as possible.

 

The SS-NX-26, for example, employs both terminal maneuvers and a terminal attack speed that transitions from 1,650 knots up to 2,310 knots (Mach 3.5). Similarly, the SS-N-27 increases from a subsonic cruise speed of 528 knots to an attack speed of 1,380 knots in the final 11 nm of its approach.

 

It should, of course, be noted that terminal maneuvers are no longer limited to pure anti-ship missiles in the modern era. Modern, stealthy cruise missiles like the AGM-158 JASSM and Tactical Tomahawk employ selectable terminal maneuver options to make it ever more likely that they will reach their target intact.

 

How does all of this translate into Harpoon Classic Commander’s Edition ?

 

Not well, at least not yet. (We’re always the optimist here at HG). At present, while there does exist a “terminal guidance” flag, there is no “terminal maneuver” or “terminal speed” flag.

 

In its most basic form, a “terminal maneuver” flag might be as simple as adding a further calculation to the engagement process, either from the attack or defence side. In other words, a missile with terminal maneuver capability might have an increased hit probability in the terminal phase, or conversely, a defender might have reduced chances of success in trying to knock down a missile equipped with the flag.

 

A “terminal speed” flag seems, on its face, potentially more complex. But, as Tony has reminded me, the code already has the ability to model dual speed capability in torpedoes. (Something not possible in H3ANW, so chins up, people. )

 

It might be possible (and I’m speaking as someone who knows nothing about coding) to translate this into a workable model for missiles.

 

Having said all that, this thread is for discussion of ideas, concepts, options, and pipe dreams for incorporating a terminal attack model into HCCE. Discuss …

[TonyE]

No takers on this yet so I'll throw in my thoughts as they relate to HC...

 

The most basic solution is to modify hit chance up or down depending upon the capabilities of the missile. I think that is too restrictive and is the procedure in use already.

 

Next comes the generational idea. A generation 1 missile against generation 3 defenses doesn't have nearly the chance as a generation 3 missile against generation 1 defenses. As part of the abstract generation number would be an allowance for terminal maneuvers. I tend to like generational approach to attack vs defense.

 

Increasing the level of complexity we get to lists of terminal maneuvers a weapon can perform (each weapon could have multiple maneuvers highlighted) and a list of maneuvers a defensive mount or weapon can counter. It is tempting then to let the player choose which terminal maneuver to use (to keep it manageable, limit each group of weapons to only using one terminal maneuver type). I would say, don't let the player decide, if the player has an exact fix on the target, then have the game choose the 'best' terminal maneuver against the known defenses, otherwise use a default or randomly chosen terminal maneuver).

 

Complexity past that in a model probably means scripting of some sort with the PE for the offensive terminal maneuvers and a few flags for defensive weapons capability (ex. fast bearing change capable, fast altitude change capable, sea skimmer capable, etc). I don't see this as adding much of anything beyond the lesser complexity models except for the more detailed information that would then be part of the database as a reference work.

 

Thoughts? Different approaches?

[Akula]

The whole gererational idea is probably a good thing to implement, even if not dealing with terminal guidance. Basically a manually aimed 40mm Bofors (Generation 1, WWII era) flak battery is pretty well useless against a fast moving jet.

 

The next generation is the Vietnam era SA-2 Guideline (aka the Flying Telephone Pole) which while it does have a guidance system, is pretty well useless against a TLAM skimming along the surface.

 

Finally you move up to generation 3, the patriot battery. It has the best guidance system, and best chance of hitting. Older aircraft and missiles would pretty much be a 99% intercept chance, but...against a TLAM, I imagine it would still have issues, as the TLAM is designed to fool air defenses.

 

By making each generation have a 'to hit bonus' against lesser generations, you can get a relatively complex model, without having to change the fact that a SS-N-2C Styx still has a 60% chance to hit its target if it locks on. If it does manage to get past the defenses, you still have the same hit/miss ratio. The air defense check done by generation simply ensures that a 40mm Bofors will have a hard time stopping it, and the Patriot will probably get it most of the time.

 

We could simply define generation by guidance capabilities

• Generation 1 is the Mk 1 Eyeball guidance package, 100% reliable, 90% likely to miss.
  
• Generation 2 is a rudimentry guidance system without LD/SD capabilities (IE hard to track very low level targets)
  
• Generation 3 Modern tracking systems, able to lock on and engage at all altitude bands.
  

You can even do the same with the offensive weapons and aircraft, defining the generation by the tracking and evasion capabilities.

[divefreak]

Next comes the generational idea. A generation 1 missile against generation 3 defenses doesn't have nearly the chance as a generation 3 missile against generation 1 defenses. As part of the abstract generation number would be an allowance for terminal maneuvers. I tend to like generational approach to attack vs defense.

 

 

Thoughts? Different approaches?

 

 

Sounds Cool.

 

And as Akula states it will represent the different capabilitys for weapons.

 

Such as the simple 40mm/70 Model 1948 vs the 40mm/70MK3 vs the CIWS Version aka Sea Trinity.

 

Or the Capability of the 76mm Oto-Melara with HE-MOM Ammunition against missiles vs the Ancient 76mm MK25.

 

the older Weapons won´t see a hit against fast moving targets due to manual aiming and loading. And this can´t be modelt by PH values only...

 

So the optimum from my point of view:

 

1. all Mounts respect arc limitations

2. Weapons get a offensive/defensive generation weighting aka Flag or Value

3. Target speed and course is taken into account for the final ph calculation

[noxious]

My first question : what's wrong with H4.1 rules on Terminal guidance ?

 

Which are simply hit modifiers, with some generational data influx.

Have they been decried by the Pooner community as wholly unrealistic ?

IMHO, it's good stuff.

And you could inspire yourself from the "free shot" for point defenses (when a missile would go through a weapon's envelope between turns, if you get my drift ) to do the reverse, and prevent any point defense if a T flagged missile is going faster than 1000 knots and within 1, 2, 3, 6 or whatever time delta will work of target, then no point defense rolls at all, go straight to missile hit resolution.

I don't know how the various time scales are handled internally in HCE, but the worse case means you have to craft an algo for each time step, or set up gate values for each time step or group of timesteps if mechanics don't change for every setting (the whole gamut from 1 sec/real time to 5 mins per sec)

 

If you want to go fancy, beyond the 4.1 stuff, using the generational attack vs defense (which I also like ) plus terminal maneuver flags and terminal maneuver defense flags, I would look at something like Finite State Machines as a mechanism to expose/modify on the fly weapon/platform parameters : your SSM would be in varying states from launch to hit/miss defined by transition events (classic FSM stuff Tony, hehe), and its internal data. So when Terminal maneuvers are engaged (with type of maneuver possibly as a parameter), it switch to a particular terminal state.

Similar construct for ship borne weapon platforms, which can then switch states as they change modes, targets, fire, etc.

When the missile comes in-range, states interact through methods, which leads to a final missile state that determines Hit percentiles while taking into account defender's capabilities.

 

That said, in line with my religion of KISS (Keep It Stupidly Simple, if you didn't know ) I'd go for generational interaction, something simple and along the lines of H4.1 rules. Or any updates to said rules through sitreps, etc.

My recently strongly devalued 0.02 Can dollar 8p

[CV32]

Some further thoughts emanating from discussion on IRC ...

 

<Brains> CV32_clone: That reminds me... Did you see the manuever profile question I had asked the other day or are you implementing a policy of strategic disengagement from such a question? *grin*

...

<Brains> <@Brains> CV32: DBish kind of question... How much do we really know about the movement profiles of weapons in open literature and

<Brains> how much do they really come into play? Enough to model? Does a terminal wiggle mean much difference from a terminal

<Brains> dive, etc.

<Brains> (And I'm not really too interested in the TMs as I lean to them being mostly lumped into the ATA/DATA mechanic, I'm mostly interested in the hi-lo or high diving or whatnot profiles of the weapons over their entire range.)

...

<CV32_clone> Brains: we know some things about some weaps .. the HCE DB doesn't take it into account at all, but the code does, somewhat ... one of my standing wishlist items is for a better 'terminal maneuver' model

...

<Brains> CV32_clone: Do we know roughly enough to class them? I mean, some are

<Brains> CV32_clone: Do we know roughly enough to class them? I mean, some are cruisers, some are zoom/divers, etc? And what are the holes you see in the TM setup?

<CV32_clone> yes, i'd say we know enough to class most

<TonyE> do we know enough to class which defenses can defeat each maneuver?

<Brains> Do they break down into broad classes or is there a lot of variation?

<CV32_clone> T can add more, but the code only looks at things like crossing targets, diving targets, high speed (>1300 kt)

<CV32_clone> broad classes would probably cover it, though some weaps would have elements of more than one class

...

<Brains> CV32_clone: Can you give some class examples as well as the ones you think would cross classes?

...

<CV32_clone> okay ... there's terminal dive, eg. several FSU/Russian ASMs, ASMP

...

<CV32_clone> high altitude followed by low altitude approach, eg. older Sov ASMs like AS-2 Kipper

<CV32_clone> terminal pop up, eg. -84B/C Harpoon, ASMP (selectable mode)

* Brains thinks that Pandora's use of the Music Genome Project (IIRC that is what it is called) is pretty cool.

<CV32_clone> high altitude followed by low altitude approach, eg. older Sov ASMs like AS-2 Kipper

<CV32_clone> terminal pop up, eg. -84B/C Harpoon, ASMP (selectable mode)

<CV32_clone> AGM-86D CALCM too has terminal pop up

<Brains> Would it make sense to separate the cruise profile from the TM type? Do all terminal popup missiles cruise at low altitudes and all terminal wigglers sea skim, etc?

* TonyE LOL at terminal wiggler

<CV32_clone> S weave, eg. Otomat Mk3

<CV32_clone> terminal high speed, eg. SS-N-27 Sizzler

<CV32_clone> Brains: generally, yes, i'd say most pop up types are also low alt cruisers, and the terminal 'wigglers' are also generally sea skimmers

<CV32_clone> there are some variation, though, like the Otomat .. goes from Low to Vlow with that S weave

<CV32_clone> C803 also has that boosted speed attack in the terminal phase

<CV32_clone> we can't presently model that

<CV32_clone> then there's reattack, sort of a terminal maneuver in itself

...

<Brains> Don't some missiles (I think I'm thinking of a Russian ASM) also have multiple cruise profiles that can be selected?

<CV32_clone> yep, like SS-N-19

<CV32_clone> High-Low or Low-Low

<Brains> Can you explain to me why the TM stuff can't be abstracted away like the generational compare or the simple flag system already in use? Are they that much difference in effect (even if the actions are much different)?

<CV32_clone> nope, i can't, really .. except maybe for the effect on detectability prior to engagement

<CV32_clone> the terminal diver, for example, might be spotted long before the skimmer

<Brains> Okay, moving back before the TM, how complicated are these "cruise" profiles? Is it something as simple as lo-lo, hi-hi, hi-lo?

<CV32_clone> i'd say its pretty simple (or easily simplified)

<CV32_clone> H4 gives a min alt, a max alt, and a trajectory type (cruise or ballistic)

<Brains> Are all the profiles set in stone at launch or is there some alteration allowed in the middle?

<CV32_clone> you mean like specific altitudes?

* Brains seems to remember that some can be altered in flight... Wasn't that a capability added (or about to be) to the Harpoon series?

<CV32_clone> well, i recall TASM had the sea skimming approach, a pop up before the terminal phase in order to get a PRH sniff or an ARH scan, followed by the Vlow terminal approach

<CV32_clone> sort of a Vlow-Low-Vlow

<Brains> Could any of the settings be changed in midflight?

<CV32_clone> in what sense? in HCE or real life?

<TonyE> real life since we know they can't in HCE

<CV32_clone> i'm sure you could do it with man-in-the-loop stuff like SLAM, otherwise it'd have to be preprogrammed or via datalink

<CV32_clone> Russians were pretty good at the latter

* Brains is thinking datalink.

<Brains> (Man-in-the-loop counts as datalink, pretty sure)

<CV32_clone> yes, but with much more direct control

<CV32_clone> SS-N-12 is a good example of the capability

<CV32_clone> lead missile transmits targeting info via datalink to the followers

<TonyE> that's pretty wicked

<CV32_clone> even more wicked .. if the lead gets shot down, another takes its place

<Brains> Which is why midcourse updates for Standards are nice....

<CV32_clone> yep, midcourse updates are nice all around

<Brains> How much do we know about the datalink sharing between missiles? The stuff I've read mentions a "swarm" but how the swarm is made up isn't mentioned. All missiles launched in the same salvo, all missiles with the right code (updateable in flight?), or what?

<CV32_clone> we know very little

<CV32_clone> it does usually mention a 'salvo'

...

<TonyE> a long winded report for the terminal maneuvers pipe dream would be nice though

<Brains> I'm thinking about how to model 'em if I ever get around to actually writing code...

<CV32_clone> i'd like to think it could be dealt with a simple piece of code that adds or subtracts pH, but on the other hand, some terminal maneuvers are more equal than others

<Brains> Which ones and how?

<Brains> Should it just be a TM ph reduction field on everything?

<Brains> Those without TMs get 0%, everything else gets a number with that number having suggestions in the Missile Making doc?

<CV32_clone> a terminal supersonic approach at Vlow, with S weave (a la SS-N-27) should be more effective than a subsonic pop up

<CV32_clone> some defensive weaps might be almost entirely incapable of dealing with that kind of TM, which makes the pH factor sort of inefficient

<Brains> Wouldn't the speed portion be already factored in aside from the TM?

* Brains thought there was a pH penalty against high speed missiles.

<CV32_clone> the >1300 kt factor

<Brains> So is the difference between the supersonic vlow s-weave and a subsonic popup more than the speed/alt diffs would make it?

...

<CV32_clone> well, i don't know, thats what needs discussion ... should my crew served 25mm Bushmaster have any chance at all?

<TonyE> the fast mod (> 1300 kts) docks PH by 30%

<CV32_clone> hehe, that probably removes the 25mm from the equation by itself

<TonyE> lol

<CV32_clone> a DB field would probably deal with all the variables, but a piece of code would be easier, no?

<TonyE> in a month or less we'll be open to entirely revamping the DB structure so I wouldn't let a DB field stop you.

...

* TonyE thinks we'll have to in tandem keep bug fixing the existing game but not sure that leads to improvements such as you are speaking of.

<CV32_clone> well, take the SS-N-27 Sizzler as a working example

...

<CV32_clone> Vlow cruise at 528 kt until 11 nm from the target, at which time it goes to 1,380 kt

<CV32_clone> interceptions at >11 nm should not be docked pH

<CV32_clone> how does one model that?

<TonyE> I always end up thinking equations, i.e. the terminal maneuver equation for this weapon entry is "if DistToImpact > 11 speed = 528, elxe speed = 1380".

<TonyE> trying to reach a thought closer to what Brad will bear

<CV32_clone> so a specific chunk of code for that DB entry?

<TonyE> the DB author would enter that little bit of code as part of the DB, like scripting (exactly scripting actually)

<CV32_clone> ok, in a revamped DB structure, obviously

<TonyE> yah

<CV32_clone> maybe a terminal speed field

<TonyE> yes, then you still have the question of what defines terminal distance and that's only for one maneuver.

<CV32_clone> right

<TonyE> maybe you end up with a whole annex of maneuvers, like mounts on a ship

<CV32_clone> so a specific chunk of code for that DB entry?

<TonyE> the DB author would enter that little bit of code as part of the DB, like scripting (exactly scripting actually)

<CV32_clone> ok, in a revamped DB structure, obviously

<TonyE> yah

<CV32_clone> maybe a terminal speed field

<TonyE> yes, then you still have the question of what defines terminal distance and that's only for one maneuver.

<CV32_clone> right

<TonyE> maybe you end up with a whole annex of maneuvers, like mounts on a ship

<CV32_clone> Harpoon's pop up, for example, is much closer .. something like 1-2 nm

<TonyE> that maneuver annex could have each of the fields, terminal distance (or time from target), terminal speed, maneuver type.

<CV32_clone> selectable method of terminal attack during play is another issue

<CV32_clone> yes

<TonyE> true, almost a separate issue

...

<CV32_clone> some of this could be solved by giving the player control of the missiles in flight

<TonyE> yes. was having that thought while you two were jabbering

<CV32_clone> also addresses things like SLAM and hunter-killer UAVs

...

<Brains> The last minute dash is something I think should be handled outside the TM stuff as an 11nm last minute dash isn't all that terminal. 1nm probably is...

...

<CV32_clone> 29 seconds is pretty terminal <g>

[donaldseadog]

On 10/15/2007 at 11:32 AM, TonyE said:

Next comes the generational idea. A generation 1 missile against generation 3 defenses doesn't have nearly the chance as a generation 3 missile against generation 1 defenses. As part of the abstract generation number would be an allowance for terminal maneuvers. I tend to like generational approach to attack vs defense.

I like this

(and the whole discusion)