Author Topic: Sensor Model for C# Aurora  (Read 2646 times)

0 Members and 1 Guest are viewing this topic.

Offline Steve Walmsley

  • Moderator
  • Admiral of the Fleet
  • *****
  • Posts: 6508
  • Thanked: 851 times
    • View Profile
    • http://www.starfireassistant.com
Sensor Model for C# Aurora
« on: March 26, 2017, 09:41:34 AM »
As per the recent suggestions in the Aurora Changes Discussion thread, I am looking at a new active sensor model for C# Aurora. There is an issue that active sensor ranges become so huge with large size-50 sensors, that the standard tactic is to create a ship with such a sensor so that it can watch the entire inner system, taking away some of the fog of war. In addition, such extreme-range sensors allow ultra-long range missile combat, giving the race that possesses such sensors a major advantage. The following change is intended to create a situation where:

a) Multiple scouts or pickets become a serious alternative to one huge sensor.
b) Missile combat ranges are reduced
c) Fog of war is increased, leading to more interesting exploration and combat.

The VB6 sensor model is based on the following formula, which increases range in direct relation to sensor strength:

Sensor Range = Racial Sensor Strength * HS * Racial EM Sensitivity * SQRT(Resolution) * 10,000 km

The proposed C# model uses similar basics and leaves all the existing technology in place. However, the sensor strength now has to cover an area rather than a direct range, creating diminishing returns for larger sensors. In addition, the modifier for resolution has been adjusted from square root to the power of (1 / 1.5). Because of this formula, smaller, lower resolution sensors are now more effective than the VB6 equivalents (much more in some cases), making earlier detection of missiles and fighters possible for non-specialised ships. The new formula is:

Sensor Range = SQRT((Racial Sensor Strength * HS * Racial EM Sensitivity * (Resolution ^ (1/1.5)) / PI) * 1,000,000 km

The following screenshots are based on the Commonwealth in my current campaign, which has active sensor strength 21 and EM sensitivity 11.









 

Offline Zincat

  • Lt. Commander
  • ********
  • Posts: 234
  • Thanked: 18 times
    • View Profile
Re: Sensor Model for C# Aurora
« Reply #1 on: March 26, 2017, 10:05:07 AM »
Those numbers don't look bad to me. Though I wonder how high they would get with max-tech. Never really reached that.

I assume that planet based sensors will not stay the same they are now? In my opinion, they should change somewhat as well. Else I can just build 50 deep space tracking stations on my colony and see pretty much anything in the system anyway.

I would simply take away the linearity in adding more stations -> more sensor power. Makes no sense that 4 deep space tracking stations have 4 times as much sensor capabilities than one station.

Maybe you should use the same area-based formula for them as you use in this simulation.

Or, just a square root increase factor.
4 stations -> twice as much sensor strength as one station
9 stations -> three times as much sensor strength as one station
and so on.
 

Offline Steve Walmsley

  • Moderator
  • Admiral of the Fleet
  • *****
  • Posts: 6508
  • Thanked: 851 times
    • View Profile
    • http://www.starfireassistant.com
Re: Sensor Model for C# Aurora
« Reply #2 on: March 26, 2017, 10:07:50 AM »
Those numbers don't look bad to me. Though I wonder how high they would get with max-tech. Never really reached that.

I assume that planet based sensors will not stay the same they are now? In my opinion, they should change somewhat as well. Else I can just build 50 deep space tracking stations on my colony and see pretty much anything in the system anyway.

I would simply take away the linearity in adding more stations -> more sensor power. Makes no sense that 4 deep space tracking stations have 4 times as much sensor capabilities than one station.

Maybe you should use the same area-based formula for them as you use in this simulation.

Or, just a square root increase factor.
4 stations -> twice as much sensor strength as one station
9 stations -> three times as much sensor strength as one station
and so on.

I will be doing something similar with passive sensors as well (covering area not range). Just want to decide on the active model first and then I will use that as the basis for the passive model.
 

Offline Haji

  • Commander
  • *********
  • Posts: 347
  • Thanked: 19 times
    • View Profile
Re: Sensor Model for C# Aurora
« Reply #3 on: March 26, 2017, 10:19:53 AM »
The numbers look very good to me. I especially like the fact that the missile detection range is actually increased as on lower tech levels it's very difficult to actually launch a useful amount of interceptors against any given salvo. Personally I would change the "1 000 000 km" in the equation to "900 000" to make the ranges just a tad lower, but that's just a detail.

As for passive sensors. I agree with Zincat that tracking stations need additional work independent of the sensors in general as from the middle fusion era they become far too powerful even in very low numbers. Please remember it's not just the case of their sensitivity - sensors and engines become more powerful making them easier to detect.

As such I'd like to see three basic changes to the tracking stations on top of the coming changes to passive detection ranges. First, like Zincat proposed, the increase in detection range should be equal to square root of the number of stations, same as it's done with sector command. Second I'd like the stations to require manpower in order to prevent them from being spammed all over the system. The manpower requirement does not have to be high, but it needs to at least force a minimalistic habitat or a couple of points of infrastructure to support it. Third I'd like their basic detection capabilities to be reduced to equivalent of a 20 HS sensor (I believe they now provide detection equal to a 50 HS sensor) and to somewhat limit their improvements from technology.
 

Offline byron

  • Rear Admiral
  • **********
  • Posts: 880
  • Thanked: 28 times
    • View Profile
Re: Sensor Model for C# Aurora
« Reply #4 on: March 26, 2017, 10:42:50 AM »
I think I'm in favor of this.  It should work against my current strategy of 'build big sensor ships' and force a dispersed network of smaller sensor platforms.  Which is good.
This is Excel-in-Space, not Wing Commander - Rastaman
 

Offline Zincat

  • Lt. Commander
  • ********
  • Posts: 234
  • Thanked: 18 times
    • View Profile
Re: Sensor Model for C# Aurora
« Reply #5 on: March 26, 2017, 10:49:35 AM »
As such I'd like to see three basic changes to the tracking stations on top of the coming changes to passive detection ranges. First, like Zincat proposed, the increase in detection range should be equal to square root of the number of stations, same as it's done with sector command. Second I'd like the stations to require manpower in order to prevent them from being spammed all over the system. The manpower requirement does not have to be high, but it needs to at least force a minimalistic habitat or a couple of points of infrastructure to support it. Third I'd like their basic detection capabilities to be reduced to equivalent of a 20 HS sensor (I believe they now provide detection equal to a 50 HS sensor) and to somewhat limit their improvements from technology.

I am ok with lowering the base sensor strength of deep space tracking stations. I'm not certain about the infrastructure requirements.

I do agree with the idea that deep space tracking stations should not be easy to deploy on places where you do not have a colony, but not with the fact that it should be impossible to do so unless you have a colony.


Maybe the solution could be along some different lines. In order to avoid spamming deep space tracking stations everywhere, they could be 1) a lot bigger and so much harder to move around and 2) have a lot more of signature, and thus be much more vulnerable to enemies if not protected. Right now, they're very hard to spot.
 

Offline Haji

  • Commander
  • *********
  • Posts: 347
  • Thanked: 19 times
    • View Profile
Re: Sensor Model for C# Aurora
« Reply #6 on: March 26, 2017, 11:07:50 AM »
I do agree with the idea that deep space tracking stations should not be easy to deploy on places where you do not have a colony, but not with the fact that it should be impossible to do so unless you have a colony.

An alternative is to have the tracking stations pretty much the same way they are now (except of the coverage vs number issue) but have their price increased that that of the spaceport, shipyard or military academy (2400 BP in total). That will require significant investment in time, more than anything else, to build the things and as such making even putting a single one on every planet in the system a rather difficult option.

Now don't get me wrong I think there should be a way for a heavily inhabited planet to easily see a lot of what is going on in the system it is located in. The problem is that the tracking stations aren't really that expensive when you come down to it. Even with a relatively low start it is very easy to build 20-50 of them and even if you lower their detection range, putting 20 of them in various places in the system will be trivial and not even that much time consuming. So if the whole point of lowering detection range is to encourage patrols and pickets, than something has to be done to stop the tracking stations from being an answer. An alternative yes; a no-brain answer no.
 

Offline Bremen

  • Lt. Commander
  • ********
  • Posts: 282
  • Thanked: 12 times
    • View Profile
Re: Sensor Model for C# Aurora
« Reply #7 on: March 26, 2017, 12:54:48 PM »
I really like it  ;D. It both deals with the massive sensor issue and active sensor range scaling exponentially with tech. Additionally, it makes having multiple resolutions of sensors instead of a few larger ones much more practical.

As for populated systems... part of me wants to suggest building fields of sensor buoys (why yes, I do get most of my ideas from Sci-Fi I read!), but that would probably be too much micromanagement and performance overhead. I was going to suggest a special order for "drop buoy at each survey point" but realized that doesn't quite work since survey points are often either very close in or very far out compared to the real estate you care about. OTOH, it does mean that's where the jump points are, so maybe it does make sense to put the buoys within that range.

Even with the performance improvements in C#, though, having dozens of sensor buoys in each system might be a bit much. Detection is tricky stuff.
« Last Edit: March 26, 2017, 01:10:16 PM by Bremen »
 

Offline Iranon

  • Captain
  • **********
  • Posts: 473
  • Thanked: 32 times
    • View Profile
Re: Sensor Model for C# Aurora
« Reply #8 on: March 26, 2017, 03:42:43 PM »
Looks good to me. I also like how increasing resolution becomes more effective relative to increasing size; looks like it enables valid trade-offs between efficiency and EM signature.

I think techs would benefit from a small change though. Current situation:
Strength increases range, but also build/development cost and EM signature.
Sensitivity increases range by a comparable proportion, without increasing cost and signature, and benefits passive sensors.
Having cost scale with Sensitivity rather than Strength would open up viable options where currently there's "one correct way".
 

Offline Michael Sandy

  • Captain
  • **********
  • Posts: 486
  • Thanked: 15 times
    • View Profile
Re: Sensor Model for C# Aurora
« Reply #9 on: March 26, 2017, 05:20:21 PM »
I definitely agree that there needs to be some special orders for deploying multiple sensor buoys, going for either a detection line with a given thickness (to detect missiles crossing it), or a zone deployment.
 

Offline Zincat

  • Lt. Commander
  • ********
  • Posts: 234
  • Thanked: 18 times
    • View Profile
Re: Sensor Model for C# Aurora
« Reply #10 on: March 26, 2017, 05:31:32 PM »
Sensor buoys? No. Really, just no.

The main idea here is that one needs a sensor network with pickets and the like. Not 200 micro satellites dispersed over a large area. Which would just be a huge pain to manage, slow down the game to a crawl, and be a repeat of the sorry state of minefields as of current versions of the game. Not to mention all the interrupts.

Micromanagement hell...


I may be a crazy loon for playing Aurora but there is a limit to madness. I refuse to manage a sensor buoy network in 40 systems or so.
« Last Edit: March 26, 2017, 05:40:06 PM by Zincat »
 

Offline alex_brunius

  • Rear Admiral
  • **********
  • Posts: 854
  • Thanked: 22 times
    • View Profile
Re: Sensor Model for C# Aurora
« Reply #11 on: March 26, 2017, 06:09:15 PM »
Going by area covered instead is solid IMO. Real sensors monitoring a set area of space should require 4 times the strength to detect the same object twice as far away ( think of it as monitoring a 4 times larger area on the inside of an imagined sphere of double size around your sensor).



What I would like to see in sensors is narrow focused sensors. A sensor that covers only the forward 90 degrees should need significantly less strength to achieve same resolution, but at the cost of leaving the flanks vulnerable.

I also in general feel that passives are pretty underwhelming since they are useless for all warfare except "early" warning (which often tends to become late warning after your actives already are tracking them). Along this line Id love to see missiles with mostly passive guidance or passive launched ones actually work better to allow more sneakiness, even if missiles use on-board actives for final approach. If I can track where an enemy craft is good enough for ships to intercept it within beam range using only passives I see no logical reason why missiles couldn't do the same until they are within range of their actives.

Or perhaps you could have passives which can spot stuff further out, but also have a large "probably area" or "range of possible enemy emission strengths" in what they spot, so they work better in their intended early warning role.

I don't know what is needed, but currently I don't find passive sensors that fun or useful, although in reality passive sensors are super important, especially Anti Radiation Missiles


Something that probably needs to be addressed anyways is AMM balance since missile agility scaling increase their interception chance to unreasonable levels after a few tech levels. If you are boosting missile detect range of HS1 sensors by almost x4 AMM ships can be made smaller, cheaper and more redundant so it might be even more important.

Two things I would like to see is:
1.) Missiles need progressively larger "guidance packages" the further away from ships missiles are supposed to operate ( they need to keep signal with the ship FC for guidance ), and/or some sort of minimum size frame as mentioned. This should also help against the commonplace AMMs as an offensive swarm weapon strategy to overwhelm PD/Enemy AMMs.
2.) Missile armor scales with techs ( such that AMMs need more then strength 1 warheads on later techs even against missiles with way less then 1MSP armor ).
 

Offline Haji

  • Commander
  • *********
  • Posts: 347
  • Thanked: 19 times
    • View Profile
Re: Sensor Model for C# Aurora
« Reply #12 on: March 26, 2017, 06:48:57 PM »
Sensor buoys? No. Really, just no.

The main idea here is that one needs a sensor network with pickets and the like. Not 200 micro satellites dispersed over a large area. Which would just be a huge pain to manage, slow down the game to a crawl, and be a repeat of the sorry state of minefields as of current versions of the game. Not to mention all the interrupts.

Micromanagement hell...

I may be a crazy loon for playing Aurora but there is a limit to madness. I refuse to manage a sensor buoy network in 40 systems or so.

In my current campaign I have tracking stations in 200 systems or so. Most of them are uninhabited but the threats are many and early warning saved several hundred million lives already. But I digress. I just wanted to say this perfectly reflects my own attitude about the buoy issue.

What I would like to see in sensors is narrow focused sensors. A sensor that covers only the forward 90 degrees should need significantly less strength to achieve same resolution, but at the cost of leaving the flanks vulnerable.

This seems like an interesting idea... until you remember there is no reason why a ship can't do 4 ninety degrees turns every couple of minutes, allowing it to effectively function as a typical 360 degrees sensor.

I also in general feel that passives are pretty underwhelming since they are useless for all warfare except "early" warning (which often tends to become late warning after your actives already are tracking them).

That is half true. Passive sensors on most of the ships are effectively a waste of space. However tracking stations (which are passive) are overpowered, fleet scale (25HS+) sensors on flagships are very potent and having mid-sized sensors on your survey ships may be a difference between life and death as enemy will usually turn on at least one very easy to detect active sensor as soon as he begins looking for your. By and large however it is true that passive sensors aren't really that useful at the moment.

What I think passive sensors need is tied not so much to their own sensitivity (which will be changed soon, hopefully addressing the power of the tracking stations) but rather the fact that ships emit far too little. Active sensors cannot be detected if they are not being used and thermal signature is tied purely to engine power with rather cheap technology to limit the signature. What I think we need is for ships to emit something (may be thermal) that will advertise it's position to even midsized passive sensors from a billion kilometres or so away which would make not only passives but also stealth much more useful.
 

Offline Zincat

  • Lt. Commander
  • ********
  • Posts: 234
  • Thanked: 18 times
    • View Profile
Re: Sensor Model for C# Aurora
« Reply #13 on: March 27, 2017, 12:52:34 AM »
In my current campaign I have tracking stations in 200 systems or so. Most of them are uninhabited but the threats are many and early warning saved several hundred million lives already. But I digress. I just wanted to say this perfectly reflects my own attitude about the buoy issue.

I think I did not explain my point well, thus this misanderstanding. I do also use a very large network of early warning stations. But with satellites, it would be different.

A sensor buoy network using small sensor satellites to monitor thing would, considering reduced range and all the possible routes enemy ships use, and the fact you don't want your sensor satellites to be seen, require a lot of satellites PER SYSTEM. Not just one or two or five. I said 200 just to write a big number, but realistically in important systems with many jump points you'd probably need 30-50 sensor satellites per system

Thus you would have , for those 200 systems, 6000-10000 satellites or so? Which have to be put down manually. And joy oh joy once they start to break down due to maintenance and you have to replace them. In my opinion this scenario would be complete madness.


I also in general feel that passives are pretty underwhelming since they are useless for all warfare except "early" warning (which often tends to become late warning after your actives already are tracking them).

I use passive sensors extensively with deep tracking stations in systems far out. And on my scouts, when I explore new systems or enemy systems. Stealth exploration is extremely important. I do feel it needs to be made more viable though, as my stealth scouts often end up being dead scouts. But they do manage to give me intel before dying so...
 

Offline alex_brunius

  • Rear Admiral
  • **********
  • Posts: 854
  • Thanked: 22 times
    • View Profile
Re: Sensor Model for C# Aurora
« Reply #14 on: March 27, 2017, 03:34:05 AM »
I use passive sensors extensively with deep tracking stations in systems far out. And on my scouts, when I explore new systems or enemy systems. Stealth exploration is extremely important. I do feel it needs to be made more viable though, as my stealth scouts often end up being dead scouts. But they do manage to give me intel before dying so...

Well I still find that if my scouts had a minimum size active sensor instead I would get more useful intel in alot of cases before dying though... Then I know the tonnage of any enemy beamships that killed it as well as more info on missiles that hit it. If I stumbled on an enemy colony which I am not at war with I also learned the tonnages of all their ships and shipyards in orbit.

Nothing prevents them from carrying all of the sensors but besides finding large enemy colonies (with their huge emissions) I haven't found much use from passives where having larger actives instead wouldn't have served me better.


This seems like an interesting idea... until you remember there is no reason why a ship can't do 4 ninety degrees turns every couple of minutes, allowing it to effectively function as a typical 360 degrees sensor.

I still think there could be some way to make it work, maybe for anti missile/fighter sensors that only can be targeted in known threat vectors ( directly towards a known enemy passive or active signature for example )?


That is half true. Passive sensors on most of the ships are effectively a waste of space. However tracking stations (which are passive) are overpowered, fleet scale (25HS+) sensors on flagships are very potent and having mid-sized sensors on your survey ships may be a difference between life and death as enemy will usually turn on at least one very easy to detect active sensor as soon as he begins looking for your. By and large however it is true that passive sensors aren't really that useful at the moment.

What I think passive sensors need is tied not so much to their own sensitivity (which will be changed soon, hopefully addressing the power of the tracking stations) but rather the fact that ships emit far too little. Active sensors cannot be detected if they are not being used and thermal signature is tied purely to engine power with rather cheap technology to limit the signature. What I think we need is for ships to emit something (may be thermal) that will advertise it's position to even midsized passive sensors from a billion kilometres or so away which would make not only passives but also stealth much more useful.

What I am after mostly is also some way to have passives work or at least be more useful in combat situations, thus my link to the Anti Radiation Missiles which have no active signal except the one the enemy is emitting to go by. And it's far from the only weapon system to rely on passive guidance, torpedoes are often guided only or mainly by their target engine/propeller noise, and we have heat seeking missiles for air-air use as well. Camera/Optically guided missiles are steered towards moving targets using passive visual or IR emissions as well.

Stealth/Emissions could be way more interesting and dynamic if you could go by either tonnage(active), heat or EM emissions for targeting instead of only one of them.
« Last Edit: March 27, 2017, 04:13:06 AM by alex_brunius »
 

 

Sitemap 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51