PAF J-10CE News, Updates and Discussion

[ghazi52]

Shukar Alhumdulillah For Having The Pakistan Air Force.

 

[Tariq Habib Afridi]

here was a text response i had written up elsewhere discussing why LINK-16, and by extension, link 17 cant be used for CEC/anything more than mid course guidance- for those who may be interested




"so LINK-16 can be used for targetting, but theres some very key differences;

CEC relies on and uses raw data, particularly raw radar data (ToD, RTT, Azimuth, Elev, Doppler shift, signal shift, SNR, Dwell ID)- the PAF's issue is the inability to feed raw data real time, but also, the inability to facilitate fused composite tracks- so these would be real time, Fire Control quality tracks, from various sensors, updated real time, with a track history, velocity vectors and also error elipses. Also, it would take hardware changes and youd have processors on both sides (CEP's), taking dwells from all units, building a unified, synchronized track to give enough info to actually launch a missile off of. Basically, youd have range resolution within meters, also you can do target handoffs in multi domains and then not even need to see the target yourself to actually engage.

Link 16 (and i assume 17) use a series of "J" Codes- its basically just data packets used to transfer information within the network. The issue is, these are pre filtered, slimmed down data packets- for example, air tracks are coded "J3.2" and theyd carry a track ID, latest pos, basic course/spd data, and IFF status. Youd be missing alot of the good stuff required for the resolution needed to be able to hand off a fire control quality track.

Also, LINK-16 and probably 17 uses a TDMA based priority system to reduce interference, so a 24 hour day is broken into 128 time frames per second, which is further divided into 1536 time slots, this basically means each unit is assigned a given Tx slot, but is Rx/Listening the remaining slots, so you're having to wait- again, no good for fire control tracks- Thats why LINK-16 only has periodic, not continuous updates

You basically have multi second refresh cycles, no good for high speed targets (brahmos ;p), but is sufficient for slower threats, due to the limitation of slot allocations. Basically, in current form, you could provide periodic guidance offboard- like a ship for SM6 or maybe BMD as you're tracking probably in its boost phase eventually, so you've got distance and also detection range on your side, but for fighter, shorter, high speed engagements ur cooked.

US CEC uses a CEP(processor) and DDS- data distribution system, the processor takes all the raw data i mentioned and combines into a common track file, then via DDS ( a UHF link) feeds it dwell by dwell (so a very high update cycle?) through the network. Also, EACH node run the same track fusion algos so they all get an identical picture. Link 16 only will distro processed track reports and each node will do its own track fusing, cec's 'net' will refresh on every dwell (imagine how crazy that is with like 1200 trm aesa's, every hit being fed), this then means that cec can hand off tracks and data even if it loses its own lock on the tgt, but link 16 cant, all the cec units recieve high quality tracks and data which are then stored locally on their own 'track' files. Basically in practice, it means youve got continuous fire control level tracks at each unit.

what all this means operationally is tdls are lower refresh, high latency, cec ranges are larger due to the fusing of data/track, MIDS has few hundred kb/s bandiwdth but cec has HUGE levels of bandwidth, no numbers but they say its got tracking accuracy seen never before. The other big thing is with LINK-16, its all decentralized, each asset is its own isolated node, with CEC, each co operating unit forms a battle group, uploading to a single, shared, high quality net, so the groups are smaller. Also, with stuff like Link-16, u can have nodes 'disagreeing' on the data/picture, this doesnt happen with CEC as its all centralised, so there is no disagreement, the track files are local, but the feed is all shared, commonly so all assets have identical pictures, this also means losing assets is no big deal as the picture is not effected. With CEC the integrated fire control pic allows EOR modes (remote eng), so again, i can fire someones missile using somene elses data, cant do this with link 16, just some basic midcourse updates.

This all means u need more hardware too- processors, dds, antennas around the airframes etc, new comms suites. Totally new CMS and software (during tests, the USN found conflicts btween cec and link16 data being fed into AEGIS simultaneously). New ECM kit and all the training jazz.

TDLs legit just lack the fidelity, a j3.2 track lacks covariance, doppler and dwell info, timestamps etc. CEC also shares dwell tme, echo strenght, beam angle, etc, all things needed for missile cueing, cant do this with LINK 16 since its not raw data. Sensor ownership is also local, if JF17 goes down, that link and pic is gone, if with CEC it goes down, the netted sensors will make do and probably not have any effect. Also, LINK-16 is very loosely integrated, its all local terminal based, unlike CEC which is deep hardware and software integration"

Assalam U Alaikum Arslan brother,

First of all what a great insight from you and thank you for the your time and answeing this. I have few questions regarding above. my english writing is bad so i used chatgpt to clearly articulate them.


Question 1: What’s the assumed architecture of Link 17? Does it use TDMA like Link 16, or could it have a different scheme (e.g., CDMA) for higher bandwidth or lower latency?

Question 2: How much bandwidth do you think Link 17 has compared to Link 16’s 115–238 kbps? Could it support raw radar data for missile guidance?

Question 3: What enhancements would Link 17 need to guide PL-15 missiles via Erieye’s radar data? Would 10–50 Mbps bandwidth or sub-millisecond latency suffice?

Question 4: Could Link 17 enable networked targeting (e.g., Erieye guiding J-10’s PL-15) short of full CEC? What technical barriers exist?

Question 5: How feasible is integrating Saab’s Erieye with Chinese J-10 and PL-15 via Link 17 without proprietary source codes? Could Pakistan have reverse-engineered interfaces?

Question 6: What upgrades would Link 17 need to approach CEC-like functionality for an ABC kill chain? Is a hybrid TDL (processed + raw data) possible?

Question 7: How does PAF use Link 17 to integrate Erieye, J-10, JF-17, and SAMs? Is it limited to situational awareness or capable of offboard targeting?

Question 8: Could Link 17 incorporate Chinese tech (e.g., J-10 data links) to work with PL-15? How would this differ from Erieye integration?

Question 9: What refresh rate would Link 17 need to guide PL-15 against high-speed targets? Could dwell-by-dwell updates be feasible?

Question 10: What evidence would convince you Link 17 supports an ABC kill chain? Are there examples of other nations’ TDLs approaching CEC functionality?

And my finat the most important question is
How all this changes when we procure KJ-500?

 

[Oscar]

here was a text response i had written up elsewhere discussing why LINK-16, and by extension, link 17 cant be used for CEC/anything more than mid course guidance- for those who may be interested




"so LINK-16 can be used for targetting, but theres some very key differences;

CEC relies on and uses raw data, particularly raw radar data (ToD, RTT, Azimuth, Elev, Doppler shift, signal shift, SNR, Dwell ID)- the PAF's issue is the inability to feed raw data real time, but also, the inability to facilitate fused composite tracks- so these would be real time, Fire Control quality tracks, from various sensors, updated real time, with a track history, velocity vectors and also error elipses. Also, it would take hardware changes and youd have processors on both sides (CEP's), taking dwells from all units, building a unified, synchronized track to give enough info to actually launch a missile off of. Basically, youd have range resolution within meters, also you can do target handoffs in multi domains and then not even need to see the target yourself to actually engage.

Link 16 (and i assume 17) use a series of "J" Codes- its basically just data packets used to transfer information within the network. The issue is, these are pre filtered, slimmed down data packets- for example, air tracks are coded "J3.2" and theyd carry a track ID, latest pos, basic course/spd data, and IFF status. Youd be missing alot of the good stuff required for the resolution needed to be able to hand off a fire control quality track.

Also, LINK-16 and probably 17 uses a TDMA based priority system to reduce interference, so a 24 hour day is broken into 128 time frames per second, which is further divided into 1536 time slots, this basically means each unit is assigned a given Tx slot, but is Rx/Listening the remaining slots, so you're having to wait- again, no good for fire control tracks- Thats why LINK-16 only has periodic, not continuous updates

You basically have multi second refresh cycles, no good for high speed targets (brahmos ;p), but is sufficient for slower threats, due to the limitation of slot allocations. Basically, in current form, you could provide periodic guidance offboard- like a ship for SM6 or maybe BMD as you're tracking probably in its boost phase eventually, so you've got distance and also detection range on your side, but for fighter, shorter, high speed engagements ur cooked.

US CEC uses a CEP(processor) and DDS- data distribution system, the processor takes all the raw data i mentioned and combines into a common track file, then via DDS ( a UHF link) feeds it dwell by dwell (so a very high update cycle?) through the network. Also, EACH node run the same track fusion algos so they all get an identical picture. Link 16 only will distro processed track reports and each node will do its own track fusing, cec's 'net' will refresh on every dwell (imagine how crazy that is with like 1200 trm aesa's, every hit being fed), this then means that cec can hand off tracks and data even if it loses its own lock on the tgt, but link 16 cant, all the cec units recieve high quality tracks and data which are then stored locally on their own 'track' files. Basically in practice, it means youve got continuous fire control level tracks at each unit.

what all this means operationally is tdls are lower refresh, high latency, cec ranges are larger due to the fusing of data/track, MIDS has few hundred kb/s bandiwdth but cec has HUGE levels of bandwidth, no numbers but they say its got tracking accuracy seen never before. The other big thing is with LINK-16, its all decentralized, each asset is its own isolated node, with CEC, each co operating unit forms a battle group, uploading to a single, shared, high quality net, so the groups are smaller. Also, with stuff like Link-16, u can have nodes 'disagreeing' on the data/picture, this doesnt happen with CEC as its all centralised, so there is no disagreement, the track files are local, but the feed is all shared, commonly so all assets have identical pictures, this also means losing assets is no big deal as the picture is not effected. With CEC the integrated fire control pic allows EOR modes (remote eng), so again, i can fire someones missile using somene elses data, cant do this with link 16, just some basic midcourse updates.

This all means u need more hardware too- processors, dds, antennas around the airframes etc, new comms suites. Totally new CMS and software (during tests, the USN found conflicts btween cec and link16 data being fed into AEGIS simultaneously). New ECM kit and all the training jazz.

TDLs legit just lack the fidelity, a j3.2 track lacks covariance, doppler and dwell info, timestamps etc. CEC also shares dwell tme, echo strenght, beam angle, etc, all things needed for missile cueing, cant do this with LINK 16 since its not raw data. Sensor ownership is also local, if JF17 goes down, that link and pic is gone, if with CEC it goes down, the netted sensors will make do and probably not have any effect. Also, LINK-16 is very loosely integrated, its all local terminal based, unlike CEC which is deep hardware and software integration"

Yeah, I remember when Link-17 first started coming together around 2010. There were definitely references to Link-16 and 22 during early discussions, but it was clear from the start that the aim wasn’t to match CEC-level capability. The focus was on building a sovereign TDL that gave basic networked situational awareness without relying on NATO systems. It borrowed a lot from Link-16 in terms of TDMA structure and J-series-like messaging. There was some talk of layering in mesh networking principles from Link-22, especially to improve beyond line-of-sight communication, but the core stayed tactical.

At the time, bandwidth and processing limitations made anything like real-time dwell sharing out of scope. No one was seriously considering synchronized raw data fusion or unified CEP-like processors across platforms. The priority was building an indigenous link that could handle positional updates, IFF, and basic coordination between fighters and ground stations. So in the context of your post, yeah, Link-17 was never architected to support high-fidelity, continuous fire-control tracks. It’s still a filtered-data system that relies on local sensors for final targeting. That kind of fused, netted engagement capability just wasn’t in the design brief back then.

 

[Oscar]

Assalam U Alaikum Arslan brother,

First of all what a great insight from you and thank you for the your time and answeing this. I have few questions regarding above. my english writing is bad so i used chatgpt to clearly articulate them.


Question 1: What’s the assumed architecture of Link 17? Does it use TDMA like Link 16, or could it have a different scheme (e.g., CDMA) for higher bandwidth or lower latency?

Question 2: How much bandwidth do you think Link 17 has compared to Link 16’s 115–238 kbps? Could it support raw radar data for missile guidance?

Question 3: What enhancements would Link 17 need to guide PL-15 missiles via Erieye’s radar data? Would 10–50 Mbps bandwidth or sub-millisecond latency suffice?

Question 4: Could Link 17 enable networked targeting (e.g., Erieye guiding J-10’s PL-15) short of full CEC? What technical barriers exist?

Question 5: How feasible is integrating Saab’s Erieye with Chinese J-10 and PL-15 via Link 17 without proprietary source codes? Could Pakistan have reverse-engineered interfaces?

Question 6: What upgrades would Link 17 need to approach CEC-like functionality for an ABC kill chain? Is a hybrid TDL (processed + raw data) possible?

Question 7: How does PAF use Link 17 to integrate Erieye, J-10, JF-17, and SAMs? Is it limited to situational awareness or capable of offboard targeting?

Question 8: Could Link 17 incorporate Chinese tech (e.g., J-10 data links) to work with PL-15? How would this differ from Erieye integration?

Question 9: What refresh rate would Link 17 need to guide PL-15 against high-speed targets? Could dwell-by-dwell updates be feasible?

Question 10: What evidence would convince you Link 17 supports an ABC kill chain? Are there examples of other nations’ TDLs approaching CEC functionality?

And my finat the most important question is
How all this changes when we procure KJ-500?

1. Link 17 AFAIK uses TDMA like Link 16 but may incorporate limited mesh or dynamic slot allocation concepts. No indication it uses CDMA.

2. Bandwidth is likely comparable or slightly higher than Link 16, but not enough for full raw radar data. Missile guidance would need much more.

3. To guide PL-15 from Erieye, Link 17 would need at least 10 Mbps and sub-millisecond latency plus synchronized processors. Unlikely in current form.

4. It could enable limited networked targeting with heavily pre-processed tracks. Main barriers are latency, bandwidth, and lack of unified processing.

5. integration is not the problem. bandwidth is.

6. A hybrid model is possible but would still fall short of CEC. Would need new hardware, CEPs, and a real-time DDS-like system.

7. Today it’s likely used for situational awareness, basic targeting, and track sharing. Not yet capable of true offboard fire control.

8. Incorporating Chinese data link tech is more feasible than Swedish. Would be more seamless but would still require deep integration across systems. However, PAF prefers proprietary systems which begs the question - perhaps the PAF has done some upgrades and implemented something new which is not declared @arslank01

9. To handle high-speed targets, Link 17 would need updates every 50–100 ms. Dwell-by-dwell updates would need a completely new architecture.

10. Solid proof would be raw radar sharing, CEP presence, and identical fused track files across platforms. Only US, China and select allies have CEC-like nets today.

 

[Tariq Habib Afridi]

1. Link 17 AFAIK uses TDMA like Link 16 but may incorporate limited mesh or dynamic slot allocation concepts. No indication it uses CDMA.

2. Bandwidth is likely comparable or slightly higher than Link 16, but not enough for full raw radar data. Missile guidance would need much more.

3. To guide PL-15 from Erieye, Link 17 would need at least 10 Mbps and sub-millisecond latency plus synchronized processors. Unlikely in current form.

4. It could enable limited networked targeting with heavily pre-processed tracks. Main barriers are latency, bandwidth, and lack of unified processing.

5. integration is not the problem. bandwidth is.

6. A hybrid model is possible but would still fall short of CEC. Would need new hardware, CEPs, and a real-time DDS-like system.

7. Today it’s likely used for situational awareness, basic targeting, and track sharing. Not yet capable of true offboard fire control.

8. Incorporating Chinese data link tech is more feasible than Swedish. Would be more seamless but would still require deep integration across systems. However, PAF prefers proprietary systems which begs the question - perhaps the PAF has done some upgrades and implemented something new which is not declared @arslank01

9. To handle high-speed targets, Link 17 would need updates every 50–100 ms. Dwell-by-dwell updates would need a completely new architecture.

10. Solid proof would be raw radar sharing, CEP presence, and identical fused track files across platforms. Only US, China and select allies have CEC-like nets today.

Thank you for taking the time to provide detailed and insightful answers to my questions about Link 17 and AWACS integration.
My suggestions would be to now aim for 10–50 Mbps using modern UHF or Ku-band transceivers to support near-raw radar data for missile guidance. Also, implement a hybrid architecture (e.g., TDMA with a dedicated low-latency channel) targeting 50–100 ms refresh rates for high-speed target tracking. We need to develop Cooperative Engagement Processors (CEPs) or equivalent to create unified, fire-control-quality track files across platforms. We will also need to integrate software-defined radios and high-performance processors (e.g., FPGA/GPU) on our aircrafts for real-time data processing ofcourse if china allows as we have mostly chinese systems. lastly, i dont know whether these suggestions are practical or not but we need to upgrade and i think if we upgrade our system it may also help us in efficient response time with our air defence capabilities.

for the rest i think PAF has capable people maybe already working on ugrading the system. this is as per my limited knowledge and you people may have many more practical suggestions and better one then mine.

 

[Ali_Baba]

1. Link 17 AFAIK uses TDMA like Link 16 but may incorporate limited mesh or dynamic slot allocation concepts. No indication it uses CDMA.

2. Bandwidth is likely comparable or slightly higher than Link 16, but not enough for full raw radar data. Missile guidance would need much more.

3. To guide PL-15 from Erieye, Link 17 would need at least 10 Mbps and sub-millisecond latency plus synchronized processors. Unlikely in current form.

4. It could enable limited networked targeting with heavily pre-processed tracks. Main barriers are latency, bandwidth, and lack of unified processing.

5. integration is not the problem. bandwidth is.

6. A hybrid model is possible but would still fall short of CEC. Would need new hardware, CEPs, and a real-time DDS-like system.

7. Today it’s likely used for situational awareness, basic targeting, and track sharing. Not yet capable of true offboard fire control.

8. Incorporating Chinese data link tech is more feasible than Swedish. Would be more seamless but would still require deep integration across systems. However, PAF prefers proprietary systems which begs the question - perhaps the PAF has done some upgrades and implemented something new which is not declared @arslank01

9. To handle high-speed targets, Link 17 would need updates every 50–100 ms. Dwell-by-dwell updates would need a completely new architecture.

10. Solid proof would be raw radar sharing, CEP presence, and identical fused track files across platforms. Only US, China and select allies have CEC-like nets today.

So where did you get such classified information?

 

[Oscar]

So where did you get such classified information?

You mean stale information.

Link-17 began life some three cubicles left and up to the left of me(2ft if the guy was leaning into my cubicle)

Then many more joined the fray and I said goodbye because Pk R&D and General work culture was not my ethical cup of tea

 

[Ak01]

8. Incorporating Chinese data link tech is more feasible than Swedish. Would be more seamless but would still require deep integration across systems. However, PAF prefers proprietary systems which begs the question - perhaps the PAF has done some upgrades and implemented something new which is not declared @arslank01

@Tariq Habib Afridi hopefully that should answer your questions!

in terms of this Oscar, i dont think it would be particularly difficult for the paf to make changes because these TDLs are all surface level, loosely integrated systems. Erieye from day 1 was designed to be able to operate multiple standards with ease. I dont see any potential barriers other than Politics/Oem in terms of direct links with J-10s etc, this is why PAF uses intermediaries to connect all assets via GCS

 

[Ak01]

So where did you get such classified information?

you can deduce quite alot based off of public info etc nothing classified or secretive in what was said

 

[Tariq Habib Afridi]

@Tariq Habib Afridi hopefully that should answer your questions!

in terms of this Oscar, i dont think it would be particularly difficult for the paf to make changes because these TDLs are all surface level, loosely integrated systems. Erieye from day 1 was designed to be able to operate multiple standards with ease. I dont see any potential barriers other than Politics/Oem in terms of direct links with J-10s etc, this is why PAF uses intermediaries to connect all assets via GCS

what you suggest for future course for pakistan? while answering this question briefly please keep in mind the air defence as well against high speed missiles?

 

[Ak01]

what you suggest for future course for pakistan? while answering this question briefly please keep in mind the air defence as well against high speed missiles?

im a very strong advocate for either CAMM or IRIS-T inductions, in larger numbers then using them as an offshoot basis for a LR/MR sam.

India is using its MRSAMs as basis for 40N6E style missiles, so is turkey and so did the euros via aster.

IMO having a 'dart', like CAMM or IRIS T would build uo that vital qrsam layer for shorad/mrsams which would carry the bulk of interception work against brahmos etc. then using them with boosters etc like india or turkey to form a longer ranged network of really agile, really capable sams. IRIST even offers a whole family, diehl is developing a HTK interceptor for upper tier ad with it!

 

[Tariq Habib Afridi]

im a very strong advocate for either CAMM or IRIS-T inductions, in larger numbers then using them as an offshoot basis for a LR/MR sam.

India is using its MRSAMs as basis for 40N6E style missiles, so is turkey and so did the euros via aster.

IMO having a 'dart', like CAMM or IRIS T would build uo that vital qrsam layer for shorad/mrsams which would carry the bulk of interception work against brahmos etc. then using them with boosters etc like india or turkey to form a longer ranged network of really agile, really capable sams. IRIST even offers a whole family, diehl is developing a HTK interceptor for upper tier ad with it!

Thank you. One more question maybe it will be stupid as i dont have a good knowledge of how SAMs work. but can we integrate JY-27A Radar with any of HQ9 missiles for stealth fighters such as that of B-2 or F-22?

 

[Dasboot42]

Hi,

I havent posted in ages, but am forced to do so now seeing how lax some senior members are with classified info. I am a old timer from Pakdef and OG pk related defence forums from the 90s.

But what I have shockingly observed now is the complete disregard for secrecy and the multitude of Indians newbies posing as Pakistanis trying to obtain peculiar sensitive information.

Tariq Habib Afridi is clearly an Indian and shame on Oscar gleefully answering all of his queries.​

Group admins need to assert themselves again with an Iron hand as was done in the old days with Indian posers.

Esteemed senior members of old always ensured that real sensitive info was not relayed and rather fed BS and disinformation to obvious Indians. These types of Indians posers would be banned permentaly in a jiffy after being exposed and insulted with classic gali galooch.

A sad travesty to watch senior members here being petty enough to relay info just to highlight their insights on info and self serve their small ego. PATHETIC!!!!

 

[Ak01]

Thank you. One more question maybe it will be stupid as i dont have a good knowledge of how SAMs work. but can we integrate JY-27A Radar with any of HQ9 missiles for stealth fighters such as that of B-2 or F-22?

tbh im not super familiar on Chinese SAM's, but generally when it comes to VHF radars, you can detect an inbound VLO asset, but the resolution would be very coarse. Their longer wavelength gives them an easier time in detection, but at the cost of tracking accuracy and resolution. In practice, its more of an early warning tool, used to cue other sensors and or effectors.

With the longer wavelength, you have a far broader beam, like just a silly example but a 3m antenna at 100mhz may have a beam width of 8 degrees, vs an X band aperture would have a sub 1 degree beam width- just an example.

You also cant have full broadband stealth right, youll likely make tradeoffs against VHF band by improving X and S band stealth characteristics as most fire control radars will operat ein these bands. So JY27 etc all exploit this, giving u detection at reasonable ranges, however, those large beam widths cant give u the 'tight' lock llike an x band or s band radar can, meaning you can have multiple hundred meter inaccuracies, really dropping ur pK down hard.

 

[Ak01]

Hi,

I havent posted in ages, but am forced to do so now seeing how lax some senior members are with classified info. I am a old timer from Pakdef and OG pk related defence forums from 90s.

But what I have shockingly observed now is the complete disregard for secrecy and the multitude of Indians newbies posing as Pakistanis trying to obtain peculiar sensitive information.

Tariq Habib Afridi is clearly and Indian and shame on Oscar gleefully answering all of his queries.​

Group admins need to assert themselves again with an Iron hand as was done in the old days with Indian posers.

Esteemed senior members of old always ensured that real sensitive info was not relayed and rather fed BS info to obvious Indians.

A sad travesty to watch senior members here being petty enough to relay on just to highlight their insights on info.

there was literally nothing secret about what was said. a blind man could realise link 17 will loosely be based off of 16, with probably the same limitations and equal or worse perf...