Naval missile guidance thread - SAM systems

[Tam]

What exactly is so difficult about using C-band for missile guidance? Doesn't it roughly just boil down to allowable antenna size? Given how large the HQ-9 missile is, fitting a C-band antenna is not an issue. I think the main reason why we see so many X-band guidance systems is that it facilitates reuse of technology.

The real question is that even if you are able to use C-band for guidance, what makes you think its better to use than X-band? The case is really why use C-band when X-band is superior for this purpose. The HQ-9 seeker is still fairly small and X-band will still give you better angular accuracy. It should not be any bigger than the S-300 missiles like the 48N6 and 5V55 that it is very similar to, and it should be slightly smaller. Yet the 48N6 and 5V55 are both guided on X-band.

If the HT-233 is anything like the Raytheon AN/MPQ-53 then it also has search, in addition to track and illuminate capability. Scale that up a bit, modify the waveguides for naval installation and what you have is a very capable PESA system on 052C. Wild speculation, of course.

AN/MPQ-53 is intended to be able to operate as a self contained all in one unit. Backed by an all search and acquisition radar is an option, although that option is always used. That's probably why it uses C-band it needs to do search. The MPQ-53 has over 5,000 elements while the HT-233 only has over 4,000. Its likely the MPQ-53 is either bigger or the array has a more densely arranged elements due to the use of a higher frequency in the C-band that allows it to be used for fire control.

HT-233 may have search, but it would be secondary search as it is not meant to operate alone. The array itself doesn't have enough range and its why the HT-233 has to work as part of a battery that includes a number of dedicated of search and acquisition radars such as the Type 305A, Type 305B, and Type 120 radars. Type 305A is a 3D AESA. What HT-233 has in relation to these radars the command guidance system and datalink to the HQ-9, which the acquisition radars lack.




This is the radar that you should be comparing the Type 346 with. Not only is it an AESA, but like Type 346, its an S-band AESA.

Now here is something that makes the HT233 difficult for naval use. HT233 happens to be a Spaced PESA.

Please, Log in or Register to view URLs content!

That round thing way in the back of the array and sitting on the truck happens to the array's feed. If you don't get how it works, the feed projects the signal optically with a lens to the back of the array, and the array itself acts like a second lens that narrows the beam and directs it. This type of PESA isn't an uncommon one, it's typical of PESA supporting SAM, including MPQ-53, Flaplid, and Tombstone. The radars that support Shtil and HQ-16 are also of this nature.

This is different from the kind of PESA like SPY-1 or BARS used on the Su-30M. That kind of PESA has connecting cables from a central amp to each transmitting element in the array. This kind of array is very quality intensive, because the signal needs to reach the elements at the exact same time. A difference in latency or time delay among the elements will cause a frequency shift in the elements, which results in frequency steer. That affects the efficiency of the array. Quite frankly what it costs to build such arrays is probably one of the reasons the Chinese steered directly into AESA in the first place, because it is much easier to calibrate the array electronically if there are any time delays among the elements.

With a spaced phased array, you need to bring the big feed along, and putting it within the superstructure takes a large amount of space. This is why you don't see this type of PESA used in ships unless it is placed outside of the superstructure.



Studying the 051C, and ships that have similar radars like the Kirov and Slava class, tells you another thing. These radars are insufficient for search, and the ship in question has to bring another set of radars used for search and acquisition. You can see it on the Type 051C here --- the FREGAT MAE Top Plate radar.

With the Type 052C/D, you don't see anything used like such search radars. If it is accompanied by a Type 382 Top Plate, then you have a stronger reason to suggest the arrays would be on C-band.

The 052C/D does have the Type 517 metric wave radar, but the resolution of that isn't very good, and in the ships that use them, like the Jiangwei frigates, there is still an S-band Type 360S search radar.

The array on the 052C when first seen gave thought it might be a PESA, but this photo suggests otherwise, as it points to liquid and directed air cooling towards the array itself, which points to an AESA.

 

[Tam]

It is very easy to calculate the beam diameter of a phased array antenna.

Each magnitude increase of radiation elements will add ten to the decibel of directional gain.
So, a 1000 element antenna will have 30 db directional gain.

It means that from 100 km the beam diameter will be 6 km.
With 4000 elements it will be 3 km at 100 km.

OF course it is just a thumb rule, the shape of the main lobe is complex than this, but the final result will be in this range.


Now, the phase sifters of the SPY-1B is 7 bit (A was 4 bit).

It means the resolution of the antenna beam steering electronics is 600 m at 100km .

See, the steering is five time more precise than the beam diameter.

So, the spy-1 can pinpoint a target into a not smaller than 600m spot from 100km, with monopulse.


Without that the area will be bigger. (hard to tell how big, it depends on the capability of radar to deferential between the different frequency/polarisation angular differences )

I can write down the calculation, but it is just basic high school math.

A 4 meter square array at 3Ghz would have gotten you a beamwidth of 1.75 degrees and a beam diameter of 3055 meters at 100km. That's in line with what you calculated, and a 4 meter array at 3Ghz would have over 4,000 elements.

A decrease of elements over the same size of arrays points to being a longer frequency and an increase of elements over the same size of array points to a higher frequency as the spacing between the elements is determined by the frequency used by the array. Too closely packed for the given array leads to mutual coupling and too great a space leads to grating lobes. So if you are going to put more elements given a specific frequency, your array size will grow bigger as a result, or your frequency has to be higher if your array is fixed to a size. So again, Brumby isn't incorrect even as you are correct.

 

[Anlsvrthng]

A 4 meter square array at 3Ghz would have gotten you a beamwidth of 1.75 degrees and a beam diameter of 3055 meters at 100km. That's in line with what you calculated, and a 4 meter array at 3Ghz would have over 4,000 elements.

A decrease of elements over the same size of arrays points to being a longer frequency and an increase of elements over the same size of array points to a higher frequency as the spacing between the elements is determined by the frequency used by the array. Too closely packed for the given array leads to mutual coupling and too great a space leads to grating lobes. So if you are going to put more elements given a specific frequency, your array size will grow bigger as a result, or your frequency has to be higher if your array is fixed to a size. So again, Brumby isn't incorrect even as you are correct.

The main way to get weapon quality track is to use monopulse , the beam diameter is too big for any practical purpose.


Of course the monopulse easier to jam than the pure beam, but there is not so much choice, unless you pay billions for 100 000 - 1 million elements.


And the most important part is the size of main lobe depends ONLY on the number of elements and the directional gain of them (YAGI), the frequency has no practical effect on it.

And of course the distance between the elements defined by the wavelength.

The NEBO reach good gain by use YAGI, and at the same time restricting the maximum steering angle of the main lobe .

 

[Tam]

The main way to get weapon quality track is to use monopulse , the beam diameter is too big for any practical purpose.


Of course the monopulse easier to jam than the pure beam, but there is not so much choice, unless you pay billions for 100 000 - 1 million elements.

Should be the reverse. Monopulse is much harder to jam than any previous tracking method, like conical scan. Previously conical scan is heavily used, until jamming forced the shift to using monopulse.



All these are just methods to ascertain direction. You still need to speed determination, which is through doppler or pulse doppler; range through PRF, as well as rate of update.

Do I get the feeling that you think monopulse works only with phase arrays?

 

[Max Demian]

Now here is something that makes the HT233 difficult for naval use. HT233 happens to be a Spaced PESA.

Please, Log in or Register to view URLs content!

That round thing way in the back of the array and sitting on the truck happens to the array's feed. If you don't get how it works, the feed projects the signal optically with a lens to the back of the array, and the array itself acts like a second lens that narrows the beam and directs it. This type of PESA isn't an uncommon one, it's typical of PESA supporting SAM, including MPQ-53, Flaplid, and Tombstone. The radars that support Shtil and HQ-16 are also of this nature.
...
With a spaced phased array, you need to bring the big feed along, and putting it within the superstructure takes a large amount of space. This is why you don't see this type of PESA used in ships unless it is placed outside of the superstructure.

Hmm. If space is lacking, could they not reduce the distance from the feed to the phased array to make it fit within the superstructure?

Another question. Is the phased array in any way different from that used in constrained parallel feeding, like on SPY-1? From the diagrams on this page, it would appear that only the backend differs:

Please, Log in or Register to view URLs content!

Systems.en.html

If not, could they not then conceivably switch from space feed to constrained feed?

 

[Brumby]

The inherent capability of SPY-6 or SPY-1 are not directly relevant because it is the way that the systems provide guidance to their respective missiles which matter.

For example, while it is obvious that the SPY-6 as an S band AESA with GaN is more capable than S band AESAs with GaA (like SAMPSON or MF STAR), the greater capability afforded to it by GaN on an equivalently sized array is not something which is relevant for the discussion about the guidance concept of 052C or 055.

The reason I disagree with your reasoning is because I have a broader interpretation on the scope of the subject of guidance and that is it is a function of kill chain which in turn is dictated by technology. In other words what we are seeing as a trend is the sensor shooter range dynamics continue to increase with technology and in particular due to the introduction of GaN TRs. For example, historically missiles are SARH but technology in the miniaturization of form factor is allowing increasing missiles built with dual and tri mode seekers. Additionally, missiles range have also extended due increasing sensor ranges. All these developments will dictate the configuration of future platforms. In other words, the future is not bound by history but by technological developments.

In other words -- for the purposes of this discussion, radars like SPY-6 and SAMPSON and MF STAR can essentially be grouped together, because even though the SPY-6 is a more advanced radar primarily as a result of its GaN arrays, the fact that all three radars are S band MFR AESAs is much more consequential.

From there, it becomes a matter of teasing out how those three S band AESAs guide their missiles on the respective ships they are based on -- and when you integrate that information with how a variety of other ships guide their missiles (like SPY-1 equipped Burkes, or Sachsen/De Zeven, or Horizon, or Akizuki, or Grigorovich), can help provide the context needed to think about how 052C's guidance system may slot in and what that means for successive generations such as 052D and 055.

I generally agree with your comments regarding current state.

As I had stated in my earlier post, there are primarily two type of frequency radars in play on board those platforms. They are generally S band for long range volume search and x band for horizon search. There may be examples of L band or even C band in the former but those are design choices.

I particularly mentioned that S band as incorporated in the ABs long range radar have sufficient resolution to guide the missiles to the terminal phase for active homing to take over. I have reservations whether L band is sufficient for the task especially in a BMD role.

It should be noted that capabilities requirement is seldom disclosed for obvious reasons and just because certain design choices are made it doesn’t mean that all requirements are met. For example, we know publicly that the SPY-6 configuration as adopted for Flight III does not meet the full requirements of the classified Hull study. Under the most stressed scenarios, the radar needs to be 1000 times more sensitive than the SPY-1(V)D. The present configuration can only deliver up to 100 times sensitivity.

Do you understand the difference between SARH and ARH guided missiles?

Do you understand which radars operating in which bands are necessary for the different engagement phases of a SARH vs ARH guided missile?

I ask these questions not to be a jerk, but because the choice of words you're using to describe the difference between ESSM Block I and Block II suggests you may not have a good grasp of it.

To expand on that:

Both ESSM Block I and Block II can receive targeting information and initial/midcourse guidance from S band radars, regardless of if it's a PESA (like SPY-1) or an AESA (like SPY-6).

However, ESSM Block I, being a SARH guided missile, needs offboard illumination in the terminal phase, and on ships with SPY-1, that illumination comes in the form of SPG-62 illuminators operating in the X band.

On ESSM Block II, being an ARH guided missile, it doesn't need offboard illumination in the terminal phase because it has its own onboard active radar seeker that can guide it in the terminal phase.

In other words, regardless of whether you are a SPY-6 or SPY-1 equipped Flight III or pre-Flight II Burke, your ship will still need SPG-62 illuminators to guide ESSM Block I missiles, however you won't need SPY-62 illuminators to guide ESSM Block II missiles.

My main point was illuminators are a product of old technology. It is single task guidance and line of sight (LOC) dependent. These are the reasons for the demand for Block 2 ESSM. In time, I believe they would be phased out.

My point is that the specific capability of the SPY-6 has no bearing on the discussion because the primary advancement in capability offered by SPY-6 compared to existing S band AESAs is its addition of GaN technology.

GaN changes the landscape because the extended sensor range allow for even greater missile range. It makes ARH missiles even more predominant and illuminators simply cannot keep up in range.

The other important consideration as illustrated by the chart below showing the difference in sensor area coverage between SPY-1(V)D vs SPY-6. It is a poor representation because it is 2-D but if you imagine a 3-D bubble, the search volume is significantly greater. It takes time and processing power to search a much greater area. The choice of whether it is X, S or L band becomes important in conducting searches. By my rough calculation, it would be 3 times faster to conduct a complete 360 degree sweep using S band and 7 times faster using L band as opposed to X band.

As for Type 45; correct, it is not BMD capable -- for a variety of reasons which is worth a discussion of its own -- but it isn't because the SPY-6 is somehow a superlatively unique and capable radar.

The Type 45's SAMPSON operates in S band and it is paired with the SMART-L, yes.

I personally agree and I think such a configuration offers overlapping capabilities -- X band and L band would be more sensible imo, which is what ships like Sachsen and De Zeven, but here my point is that comparing the SPY-6 with the SMART-L/S1850M is not logical given how they operate on different bands.

It would be like comparing SPY-6 with Thales APAR.

Again, this is where a cursory background search might be useful.

Yes, the Type 45 isn't BMD capable, but please note I was talking about the capability of SMART-L for BMD applications, not saying that Type 45 is BMD capable at present.

Specifically, the newest SMART-L MM/N radar is BMD capable and has been installed aboard its first ship last year as well (notably being one of the first ships in the world to field a large GaN equipped AESA in service)

Please, Log in or Register to view URLs content!

BMD is much more demanding because of the RCS of ballistic missiles and the need to be able to discriminate against decoys. It is not surprising that it requires a GaN based SMART-L for that. That said, I have my reservations whether L band can provide sufficient resolution because even GaN can’t bend radar science.

 

[Tam]

Hmm. If space is lacking, could they not reduce the distance from the feed to the phased array to make it fit within the superstructure?

Not only the feed, but you have to bring the entire box below the array as it holds the transmitter and other things.

Another question. Is the phased array in any way different from that used in constrained parallel feeding, like on SPY-1? From the diagrams on this page, it would appear that only the backend differs:

Please, Log in or Register to view URLs content!

Systems.en.html

When constructed wise parallel feeding is much more complex and quality control intensive. All the feeds have to be of the same length and material quality, so the signal is not delayed at different rates traveling through the medium of the feed.

If not, could they not then conceivably switch from space feed to constrained feed?

You have to ask the designers of the HT-233, but this final product is still going to be outperformed by Institute 14's S-band AESA in terms of range. In fact, using their C-band array will only prove inferior in range to the Fregat MAE that was imported and the Type 382 that was mostly copied from it.

The line feed is going to have insertion losses between the amp and the emitting element. Not going to happen on an AESA. There is no point in developing a more expensive PESA that isn't going to perform as well as an AESA after all.

Some examples of phase arrays.

Big Bird radar for the S-300 complex is a reflective space feed.



While Fregat 'Top Plate' used on many ships is a perfect example of a serial line feed. The PLA loves this type for their surveillance radars due to its reliability and low cost.




China never placed a parallel line feed phase array in service, not just in one service, but in any service in any of the branches. But at one point they may have tried to develop it. But for all the expense of trying to develop one, maybe the parallel project using active phase array managed to reach the goal sooner. This may explain why there is no PESA of this type that appeared in service with the PLAAF. There was no point in developing a PESA of this principle when all the other types combined has satisfied PLAN, PLAAF and PLA requirements.

Frequently common among land based Chinese search radars fall into these categories:

Please, Log in or Register to view URLs content!

Frequency scanned, serial line feed
Linear Array
Planar Array

 

[Anlsvrthng]

Should be the reverse. Monopulse is much harder to jam than any previous tracking method, like conical scan. Previously conical scan is heavily used, until jamming forced the shift to using monopulse.


All these are just methods to ascertain direction. You still need to speed determination, which is through doppler or pulse doppler; range through PRF, as well as rate of update.

Do I get the feeling that you think monopulse works only with phase arrays?

Monopulse easier with non phased array antennas , the first was a dish type antenna .


By splitting the power AND aperture to half, and increase the frequency width you automatically makes the jamming easier.

Using dish antenna or dual polarised antenna , like in the case of few early warning design mitigate the aperture problem, but the power still stay half, and the vectors for possible jam double.

But the mechanically steered antenna, like a dish or conical scanning will decrease the required energy to jam with good timing.

 

[Blitzo]

The reason I disagree with your reasoning is because I have a broader interpretation on the scope of the subject of guidance and that is it is a function of kill chain which in turn is dictated by technology. In other words what we are seeing as a trend is the sensor shooter range dynamics continue to increase with technology and in particular due to the introduction of GaN TRs. For example, historically missiles are SARH but technology in the miniaturization of form factor is allowing increasing missiles built with dual and tri mode seekers. Additionally, missiles range have also extended due increasing sensor ranges. All these developments will dictate the configuration of future platforms. In other words, the future is not bound by history but by technological developments.

As dual mode and tri mode seekers are introduced, they can be considered as future guidance categories.

But at present, the introduction of SPY-6 does not offer a new guidance category which didn't already exist before.
The SPY-6 is a radar to mention in the conversation but it doesn't offer anything new in terms of providing new guidance categories or mechanisms which didn't already exist aboard other ships.

I generally agree with your comments regarding current state.

As I had stated in my earlier post, there are primarily two type of frequency radars in play on board those platforms. They are generally S band for long range volume search and x band for horizon search. There may be examples of L band or even C band in the former but those are design choices.

I particularly mentioned that S band as incorporated in the ABs long range radar have sufficient resolution to guide the missiles to the terminal phase for active homing to take over. I have reservations whether L band is sufficient for the task especially in a BMD role.

It should be noted that capabilities requirement is seldom disclosed for obvious reasons and just because certain design choices are made it doesn’t mean that all requirements are met. For example, we know publicly that the SPY-6 configuration as adopted for Flight III does not meet the full requirements of the classified Hull study. Under the most stressed scenarios, the radar needs to be 1000 times more sensitive than the SPY-1(V)D. The present configuration can only deliver up to 100 times sensitivity.

I'm repeating myself in this post, but none of this has any bearing as to the SPY-6 being worthy of any particular mention in the context of overall international existing naval missile guidance categories/mechanisms.

My main point was illuminators are a product of old technology. It is single task guidance and line of sight (LOC) dependent. These are the reasons for the demand for Block 2 ESSM. In time, I believe they would be phased out.

GaN changes the landscape because the extended sensor range allow for even greater missile range. It makes ARH missiles even more predominant and illuminators simply cannot keep up in range.

The other important consideration as illustrated by the chart below showing the difference in sensor area coverage between SPY-1(V)D vs SPY-6. It is a poor representation because it is 2-D but if you imagine a 3-D bubble, the search volume is significantly greater. It takes time and processing power to search a much greater area. The choice of whether it is X, S or L band becomes important in conducting searches. By my rough calculation, it would be 3 times faster to conduct a complete 360 degree sweep using S band and 7 times faster using L band as opposed to X band.

View attachment 57608

Again, none of this changes anything in terms of the discussion about guidance category/mechanism.

AESAs onboard ships using primarily ARH guided long range SAMs will have been in service for well over a decade before the first Flight III Burke enters service.

Yes, SPY-6 will be a very capable radar, but it hardly worth mentioning as a unique system in the context of naval missile guidance categories/mechanisms.

BMD is much more demanding because of the RCS of ballistic missiles and the need to be able to discriminate against decoys. It is not surprising that it requires a GaN based SMART-L for that. That said, I have my reservations whether L band can provide sufficient resolution because even GaN can’t bend radar science.

I have no particular horse in this fight, the only reason I bring it up is to demonstrate the context that there is a much bigger world out there of radars and guidance concepts beyond SPY-6.

 

[Blitzo]

What are we able to confidently state about Type 346?

Last year when I discussed with Tam the hypothesis of Type 346 being exclusively C-band, the arguments against that were:

1. The radar also needs to be able to do volume search. That would mean that Type 346 AESA would need to have ICWI capability. That seemed a bit of a stretch to expect from China in the early 2000s.

2. If in C-band, given the size of the arrays, the number of T/Rs elements would approach 20000 per face to maintain 60° FOV. That would make the radar very expensive, but also very capable.

3. There were some concerns about range, due to attenuation/scattering loss. We never quantified the disadvantage of C-band at 6GHz compared to S-band at 3.4GHz in the 300km operating range.

Some arguments for C-band:
1. Missile guidance. HQ-9 relies on a C-band engagement radar. This assumes that HHQ-9 is also SARH/TVM.
2. According to some western analysts, there were doubts that China was able to produce S-band modules of sufficient power in the early 2000s.

Arguments for dual S-band/C-band:
1. Wiki article: large S-band array sandwiched between two C-band FCR arrays. Looking at the shape of the C-band arrays, it appears unlikely these would serve as illuminators. Instead, they are more likely used for IFF and high data rate communication with the missiles.
2. Evolution of Type 346A/B into even larger arrays. This would be unlikely if they are C-band, but seems consistent with lower frequencies. However, I am not confident that we can completely reject the possibility that these may be fundamentaly different radars, ie C-band on 052c vs S-band in 052d and Type 55.

It is still under some question as to whether the Type 346 is a dual S band/C band array and what exactly that may entail.

However, IMO there is no question as to whether Type 346 has an S band array.


There is no basis to believe that Type 346 is only a C band array because the original proponents of that particular argument were based on the idea that Type 346 was based off the Ukrainian AESA which happened to be C band.

But the overwhelming consensus from insiders, big shrimps, and every other individual with any kind of respected reputation in PLA watching, all say Type 346 definitely has a primary S band radar.
That is to say, the question -- does Type 346 only have a primary S band array, OR does Type 346 also have a secondary C band array?


But there is no one right now asking "does Type 346 have a primary C band array".

If tomorrow pop3 came out and said "actually type 346 is actually just a big C band array" then that would change things.
Unfortunately, the places which originally suggested type 346 was a C band array are from a few western sources back in the mid 2000s operating off dubious assumptions and information, and with the consistent and additional information we have gotten from much better and more trusted sources saying Type 346 has a primary S band, the idea that type 346 is a primary C band is not considered seriously as a theory.