an AESA radar range calculator

[stealthflanker]

Well as title said. I tried to brew some sort of excel spreadsheet to calculate detection range of radar. Optimized for use on AESA radar and hopefully work good for fighter aircraft application.

Unlike other calculators, i tried to take account some factors like pulse integration, search sectors, pulsewidth, PRF's etc. Which for some reason not included in others. The most recent iteration i present here can automatically calculate required SNR (Signal to Noise Ratio) required for 90% detection probability of Swerling 1/2 target. Representative for maneuvering fighter aircraft.

The calculator also take account of the Aperture weighting algorithm. As we know ESA radar or even Slotted planar array radar radiator element arrangement does follow specific mathematically defined weighting scheme to shape its radiation pattern. The purpose of such arrangement is to reduce sidelobe which would be beneficial for the electronic warfare and radar signal processing. For now however the effect of aperture weighting only applies on width of the 3dB beamwidth of the radar, aperture efficiency and width of physical aperture the radar might have.

In later iteration i might try to include some Electronic warfare and ESM considerations, like calculating possible detection range of ESM to the radar and its burn through range against commonly used noise jamming. Deception jamming modeling is possible but quite limited to prediction of S/J (Signal to Jammer ratio) as i lack skill to actually model monopulse tracking to predict effect of sophisticated techniques like cross eye.

The following are some screenshots of the calculator :

The variables :

https://orig15.deviantart.net/3253/f/2017/240/c/e/screenshot_1_by_stealthflanker-dblkr0p.png

Drop down menu containing select-able antenna weighting algorithms

https://orig08.deviantart.net/0239/f/2017/240/0/1/aperture_weighting_by_stealthflanker-dblkr0s.png

The weighting algorithm table, for now there are only 8 weighting scheme available.

https://orig12.deviantart.net/c2c7/f/2017/240/8/d/weighting_by_stealthflanker-dblkr0c.png

The result coloumn :

https://orig09.deviantart.net/8b98/f/2017/240/1/d/screenshot_3_by_stealthflanker-dblkr0h.png

As seen there are 2 range predictions. the 50% and 90%.

The detection range at 90% probability states the range where the radar can likely lock-on. Thus engage the target. While at 50% probability is the detection range where the radar have some certainty that the target does exist and may follow up to track and perhaps guide another sensor or wingman to the target.

The spreadsheet can be downloaded through link below :

https://www.secretprojects.co.uk/threads/an-aesa-radar-range-calculator.29367/#post-385925


I am open to any correction or feedback to the excel spreadsheet. and of course, happy using. :-*

Edit :
adding the post where i put the latest version.

 

[stealthflanker]

Slight update to the calculator. Mainly to improve "usability" as some people apparently said that my calculator is rather confusing and cluttered with variables.

In this new version i trimmed down some variables and move most of the heavy calculations elsewhere in the sheet. Thus user can directly see where he/she can input variables.

As one can see the front page now only have few and hopefully easily "figured out" variables for user to fill in green columns, and directly see the calculated range in the lower yellow colored columns.

The second sheet contains the moved calculations and "heavy lifting stuff" As one can see variables like noise figure, system temperature and loss budget are there. User can edit the columns with their own value. However it might not be advised to do so.
The loss budgets and other pre-calculated variables are based on my literature study. Mainly from K. Barton's excellent radar book. Richard G Curry's Radar modeling book. Radar handbook 3rd Edition and several others.

The third sheet contain the table for antenna weighting algorithm as before. User may still use the drop down menu to change the antenna weighting algorithm as before.

For filling guide in case someone feel confused on the pulsewidth and PRF. Here is an excellent table from the Radar handbook 3rd Edition detailing radar mode parameters.

A2A modes :

A2G modes for sake of completion.

Download the improved Calculator :
https://www.mediafire.com/file/7wrkyslc1p4d36r/AESACalcTrial.xlsx

Feedback and critiques are appreciated. Thank you.

 

[stealthflanker]

3 Years passed and things are changed.

Here i would like to present the latest iteration of the AESA radar calculator. Given the difficulty of some to download from Mediafire due to firewall restrictions. I have moved the latest version into Google Drive. In following address :


DOWNLOAD-AESA Radar calculator V3.

So what's new ?

Spoiler: Fixes and improvements

1.Fixed the inconsistent unit in the calculation of detectability factors. Despite giving "reasonable" result, this actually may lead to erroneous calculation.
2.Improving the atmospheric factors, now the target altitude are used to calculate the elevation angle which then become input to the atmospheric absorption factor
3.Cutting down of the unnecessary pages, as apparently users are not really interested in additional tools such as RCS conversions (This one is partially integrated in the main sheet now) or tools to estimate limitation of radiated power based on cooling capacity.

Spoiler: Additions and new features

1.Custom dwell time.
The earlier version of the calculator only have limited options in terms of selecting dwell time for the beam. In this version it has been rectified to allow user to specify beam dwell time on his/her own. The user only need to select "custom dwell time" in the respective drop box. Which then the value for the custom dwell time will be used.

2.Custom Detectability Factor
User can now choose to allow the sheet to estimate the detectability factor (SNR) Or manually input the value by him/herself. By default it selected to Yes.

3.Addition of Radar Response Factor.
These factors are based on K.Barton's book "Equations for Modern Radar". The factors are included mainly to take account of several variables which affects detection range, especially in modern radar.

4.Result Page
The sheet now have a "special" page where the user can review the radar parameters and perhaps make a rough comparison with some existing fighter aircraft radars. The sheet is also attempt to roughly estimate the power and cooling requirement for the radar, an important subject which i hope could expand in future development and release.

5.Tracking capacity.
The sheet now take account of "radar resource management" Where for now it is only search and track. Most if not all current modern phased array radar are multimodes, thus it may interleave modes during search. Therefore the radar may not always have its 100% capacity to provide maximum possible detection range, there are also other tasks such as Raid assessement, HOJ, etc which will take resources from the radar. Unfortunately however there are no tracking implementation yet.

Spoiler: Limitations

1.This calculator assumes a FFOV or Full Field Of View array, thus may not be applicable to other form of array e.g Thinned array, it also only applies to active array not the passive one. Passive array may require additional loss evaluation.

2.Ground clutter has not been implemented/not available, thus the calculator basically see target in "infinite skies"

3.Path propagation factor is based on flat earth approximation with "smooth reflective surface" This could be representative for sea water but not so much for land. Better model is underway, thus complimenting the num 2 when activated. The radar is looking at target in a Sky of a Flat, smooth Earth.

4.Can't estimate jamming scenario. This is rather hard and complex to implement especially if the capability to estimate the effect of deception jamming is desired. Deception jamming such as Double Cross, VGPO,RGPO does not affect detection range but rather affect tracking mode. Noise jamming in other hand is possible, although with additional complexity of having to specify the number and specification of deployed jammers and path propagation factors for each of them, this will inevitably make the calculator more complex and goes against the desire of making it as simple as possible. Plus i wonder if it's possible for excel to do so.

5.Accuracy. It is unfortunately hard not only because of secrecy but also that i have no one to fall back on. All i have are some radar related books and papers and some Rosoboroneksport brochures and ausairpower website it does not really specify on what condition of test or a complete overview of radar parameters. Accuracy is therefore will always be source of doubt and at least for now nothing i can do for it. However i believe more data would be disclosed on AESA radar technical parameters which then could allow better benchmark.

Some Example Results :

Su-57 main nose Radar

Zhuk-AE, the version with 652 T/R Modules

The range performance is presented in 3 categories based on the "level of confidence" Which is 50% (R50), 90%(R90) and Instrumented. The 50% probability of detection indicates the presence of a contact with 50% probability it's a "real" contact or ghost. Old NIIP Thikomirov site presented their radar data in this level of confidence. 90% Indicates higher threshold of detection/ high SNR, meaning that the target at that range is "90% a target" At this level of confidence the radar may begin process such as tracking or engagement/locking target with weapon systems. The Instrumented range is the range where the radar is "allowed" to "announce" detection. The value is about 20% of the maximum detection range. In this sheet it is assumed to be 20% of the R50, for increased "confidence" of a detected contact.

The "benchmark" fighter radars presented in the result sheet is based on open sources, APG-63V2, APG-79 and APG-83 however are estimated with ideal assumption (The platform is "matched" with the radar) Thus my overestimate the detection range value. All the range values are for 90% probability of detection. Given that there are no real disclosure on the assumption of detection probability for the radars i decided to make assumption that all the range values from the open source materials are for 50% probability of detection, thus needs to be "normalized" to 90%. To do that i first made estimate of SNR using the simple approximation provided in R.G Curry's "Radar Essentials" on SW1 target, which yield SNR value for respective detection probability (50% and 90%) The detection range can then be simply estimated by 4th Root using SNR as value. The result is estimate of the R90. eg Irbis-E with claimed detection range of 400 Km for 3 sqm target, in 90% it becomes 225 km.



This section is perhaps... well mod can edit this one out if necessary.

Spoiler: an Appeal

This calculator has been under development since 2015. It was at first to mainly satisfy my curiousity toward range estimation for a radar and what factor saffecting it. There is however another purpose, which basically to "escape" from at least some difficulties of my real life. Trying to make my 3D modeling hobby as a living, enjoyable but the income is not steady especially now. That my clients are apparently more concerned on surviving and therefore i received no income. Finding steady job is difficult here in Indonesia. So yeah. If you found this calculator helpful.. Maybe a little bit of donation to help me at least taking off the edge. it could be send through paypal at ffcconcept@yahoo.co.id.

Thank you very much for your attention

The last but not the least is i hope it would be of some good use :3. Feedbacks are appreciated.

 

[Ronny]

Some Example Results :
Su-57 main nose Radar

Zhuk-AE, the version with 652 T/R Modules

Peak power of 100W per element? isn't that a bit too high?

 

[stealthflanker]

Peak power of 100W per element? isn't that a bit too high?

Yeah bit of oversight there, supposed to be 10.

 

[Cordy]

Two questions

US Navy stated that radar sensitivity scales as a cube of of the size of the radar aperture, and while improvements can be made to the T/R modules said this is a linear not cubic relationship and only adds marginal capability on the order of +1 or 2 dB.

BAE marketing the Typhoon to Finland, released details of the Leonardo EW Stealth Master fitted, saying that it enables a/c if detecting radar can stay out of range and if that's not possible to digitally hide its signature, becoming invisible to the radar, or digitally create a complex and confusing picture, noise, denying a clean targeting opportunity to launch missile.

If either above true does that reflect in your calcs/results for range etc.

 

[stealthflanker]

Two questions

US Navy stated that radar sensitivity scales as a cube of of the size of the radar aperture, and while improvements can be made to the T/R modules said this is a linear not cubic relationship and only adds marginal capability on the order of +1 or 2 dB.

I am not very sure about the definition of "sensitivity" there along with the decibel merit and how it connected to detection range. As far as i know the sensitivity or "Minimum detectable signal" Is related to the threshold of the receiver. This can be set low, thus making the radar more sensitive but more prone to false alarm. The hard limit here is the natural background noise and the receiver's naturally occuring noise.

However. The Cube relationship appears from the assumption that same amount of module is used for both receiving and transmitting and the array is a planar array with regularly spaced elements transmitting broadside, which well basically what our current AESA atm (The assumption should be revisited if any other form of array is used, e.g Geodesic or conformal array).

In my calculator the range roughly scales as the cube of number of active TRM (means more TRM=more range) Which yes, drives the size of the radar aperture. The increase in range however still follows the fourth root law of Radar transmission. Which have linear relationship.

Thus if the US Navy's definition on sensitivity is actually relationship between the numbers of elements vs range then yes. My calculator got that covered. If not then unfortunately it may not be usable if one desire to use it to estimate the merit of an AESA based on the USN one.

BAE marketing the Typhoon to Finland, released details of the Leonardo EW Stealth Master fitted, saying that it enables a/c if detecting radar can stay out of range and if that's not possible to digitally hide its signature, becoming invisible to the radar, or digitally create a complex and confusing picture, noise, denying a clean targeting opportunity to launch missile.

If either above true does that reflect in your calcs/results for range etc.

Unfortunately my calculator only deal with range modeling. Not for judging the merit of active cancellation. So no.

The best one can do is specify the RCS which then the range can be calculated.

 

[Cordy]

@stealthflanker , thx for reply

As understand the USN 'sensitivity" refers to the comparison of the old '70s SPY-1 ship radar for BMD to the new SPY-6 AESA GaN radar. Target for SPY-6 was 15dB , a 32 sensitivity improvement in the signal to noise ratio against a given target compared to the SPY-1, seen stated as giving (32)1/4 = 2.3x range increase or able to detect target with half the RCS at twice the distance? The SPY-6 arrays ~75% larger compared to SPY-1.

Raytheon now claim initial tests show SPY-6 100x more sensitive (will be fitted to the new Burke Flight III destroyers with five 300t A/C plants to keep things cool, IOC ~2024.

 

[Ronny]

@stealthflanker :not quite related but can you estimate the frequency limit of AN/ASQ-239 base on antenna length? Can it detect VHF radar?

 

[Dragon029]

You can't determine frequency limits from a diagram like that; even if the length of those boxes (representing aperture locations) were actually scaled to the exact length of each aperture, at least some of the apertures are going to be arrays of multiple smaller antennas, a bit like 1-dimensional AESAs.

By having multiple, smaller antennas, an RWR / ESM suit might lose a bit on max range (if an antenna is pointed directly at some distant radar), but you can get more accurate time-of-flight and phase-shift data (which means more precise geolocating of threats), plus you'll get better detection performance against threats that are perpendicular to your flight path.

 

[Ronny]

Does this help in the estimation:

AN/ALQ-218 system features:

• Broad Radio Frequency Range: Bands 0, 1, 2 and band 3

AN/ALQ-218 RWR/ESM/ELINT Sensor System Datasheet -- Northrop Grumman Corporation | Engineering360

AN/ALQ-218 can cover band 0, 1, 2, 3
AN/ASQ-239 can cover band 2, 3, 4 and potential for band 5

So ALQ-218 can detect lower frequency while ASQ-239 can detect higher frequency

 

[stealthflanker]

Knowing the type of antenna would be helpful.

 

[Ronny]

Knowing the type of antenna would be helpful.

From BAE brochure, I think they are Vivaldi antenna


Vivaldi antenna for reference:


Band 3 and Band 4 antennae on F-22 supposed to cover 0.5-18 GHz

 

[yangke123567]

Well as title said. I tried to brew some sort of excel spreadsheet to calculate detection range of radar. Optimized for use on AESA radar and hopefully work good for fighter aircraft application.

Unlike other calculators, i tried to take account some factors like pulse integration, search sectors, pulsewidth, PRF's etc. Which for some reason not included in others. The most recent iteration i present here can automatically calculate required SNR (Signal to Noise Ratio) required for 90% detection probability of Swerling 1/2 target. Representative for maneuvering fighter aircraft.

The calculator also take account of the Aperture weighting algorithm. As we know ESA radar or even Slotted planar array radar radiator element arrangement does follow specific mathematically defined weighting scheme to shape its radiation pattern. The purpose of such arrangement is to reduce sidelobe which would be beneficial for the electronic warfare and radar signal processing. For now however the effect of aperture weighting only applies on width of the 3dB beamwidth of the radar, aperture efficiency and width of physical aperture the radar might have.

In later iteration i might try to include some Electronic warfare and ESM considerations, like calculating possible detection range of ESM to the radar and its burn through range against commonly used noise jamming. Deception jamming modeling is possible but quite limited to prediction of S/J (Signal to Jammer ratio) as i lack skill to actually model monopulse tracking to predict effect of sophisticated techniques like cross eye.

The following are some screenshots of the calculator :

The variables :

https://orig15.deviantart.net/3253/f/2017/240/c/e/screenshot_1_by_stealthflanker-dblkr0p.png

Drop down menu containing select-able antenna weighting algorithms

https://orig08.deviantart.net/0239/f/2017/240/0/1/aperture_weighting_by_stealthflanker-dblkr0s.png

The weighting algorithm table, for now there are only 8 weighting scheme available.

https://orig12.deviantart.net/c2c7/f/2017/240/8/d/weighting_by_stealthflanker-dblkr0c.png

The result coloumn :

https://orig09.deviantart.net/8b98/f/2017/240/1/d/screenshot_3_by_stealthflanker-dblkr0h.png

As seen there are 2 range predictions. the 50% and 90%.

The detection range at 90% probability states the range where the radar can likely lock-on. Thus engage the target. While at 50% probability is the detection range where the radar have some certainty that the target does exist and may follow up to track and perhaps guide another sensor or wingman to the target.

The spreadsheet can be downloaded through link below :

http://www.mediafire.com/file/7wrkyslc1p4d36r/AESACalcTrial.xlsx

I am open to any correction or feedback to the excel spreadsheet. and of course, happy using. :-*

cant download

 

[stealthflanker]

Well as title said. I tried to brew some sort of excel spreadsheet to calculate detection range of radar. Optimized for use on AESA radar and hopefully work good for fighter aircraft application.

Unlike other calculators, i tried to take account some factors like pulse integration, search sectors, pulsewidth, PRF's etc. Which for some reason not included in others. The most recent iteration i present here can automatically calculate required SNR (Signal to Noise Ratio) required for 90% detection probability of Swerling 1/2 target. Representative for maneuvering fighter aircraft.

The calculator also take account of the Aperture weighting algorithm. As we know ESA radar or even Slotted planar array radar radiator element arrangement does follow specific mathematically defined weighting scheme to shape its radiation pattern. The purpose of such arrangement is to reduce sidelobe which would be beneficial for the electronic warfare and radar signal processing. For now however the effect of aperture weighting only applies on width of the 3dB beamwidth of the radar, aperture efficiency and width of physical aperture the radar might have.

In later iteration i might try to include some Electronic warfare and ESM considerations, like calculating possible detection range of ESM to the radar and its burn through range against commonly used noise jamming. Deception jamming modeling is possible but quite limited to prediction of S/J (Signal to Jammer ratio) as i lack skill to actually model monopulse tracking to predict effect of sophisticated techniques like cross eye.

The following are some screenshots of the calculator :

The variables :

https://orig15.deviantart.net/3253/f/2017/240/c/e/screenshot_1_by_stealthflanker-dblkr0p.png

Drop down menu containing select-able antenna weighting algorithms

https://orig08.deviantart.net/0239/f/2017/240/0/1/aperture_weighting_by_stealthflanker-dblkr0s.png

The weighting algorithm table, for now there are only 8 weighting scheme available.

https://orig12.deviantart.net/c2c7/f/2017/240/8/d/weighting_by_stealthflanker-dblkr0c.png

The result coloumn :

https://orig09.deviantart.net/8b98/f/2017/240/1/d/screenshot_3_by_stealthflanker-dblkr0h.png

As seen there are 2 range predictions. the 50% and 90%.

The detection range at 90% probability states the range where the radar can likely lock-on. Thus engage the target. While at 50% probability is the detection range where the radar have some certainty that the target does exist and may follow up to track and perhaps guide another sensor or wingman to the target.

The spreadsheet can be downloaded through link below :

http://www.mediafire.com/file/7wrkyslc1p4d36r/AESACalcTrial.xlsx

I am open to any correction or feedback to the excel spreadsheet. and of course, happy using. :-*

cant download

The download link for the latest version is here :

https://www.secretprojects.co.uk/threads/an-aesa-radar-range-calculator.29367/#post-385925

 

[kiers]

Hi Stealthfighter,

For a beginning student of phase array Antenna theory (myself), what would you estimate is the 3db Beamwidth of say a F-18 or F-22 AESA radar with 1600 elements (40x40). For example extrapolating from the classical equally excited element case as a baseline would indicate a 40 element array might have 3db BW of the order of 2.5°. However this seems rather high for such sophisticated equipment as a AN/APG-81.

 

[stealthflanker]

what would you estimate is the 3db Beamwidth of say a F-18 or F-22 AESA radar with 1600 elements (40x40)

The quick rule of thumb from 3rd Edition of "Radar Handbook" also suggest 2.5 Degrees. So i would think it would be typical. Probably higher given that AESA can have tapering law algorithm applied e.g Taylor to reduce sidelobe and therefore increasing the beamwidth even further.

However this seems rather high for such sophisticated equipment as a AN/APG-81.

There is no degree of sophistication that can make beamwidth lower. If such are desired, phased array designer will resort to operate in higher frequency, which if half wavelength spacing is retained more TRM can be packed for the same available aperture. Or increase the aperture size by adding more TRM's. Using the same rule of thumb as above, Beamwidth of 1.3 degrees can be achieved by increasing the TRM in a factor of 4 (6400). But whether this is practical is open to question.