Tuesday, May 31, 2016

Distributed Targeting

A reader, Benjamin Oliver, recently made a very succinct comment regarding the use of the LCS and every other Navy ship as shooters.  To paraphrase slightly, he said that distributed lethality won’t work without distributed targeting.   This is a brilliant summation of the issue.

The LCS will have 100+ nm anti-ship missiles coupled with 20 nm sensors.  Yes, the ship will have a helo and a UAV but the helo will be reserved for ASW and near-ship ASuW.  The UAV will have a very limited sensor field of view and a single UAV will be woefully inadequate for broad are target searches.  Thus, the ship will have to depend on off-board sensing and targeting.  We’ve discussed this many times.

The Navy’s plan to use P-8s and Triton UAVs is unworkable.  Both are large, slow, non-stealthy aircraft that will serve only as target drones for enemy aircraft and missiles.

Submarines simply don’t have the sensor range or speed to cover the large swaths of ocean needed to find enemy ships.  Plus, our submarines will have more important tasks than acting as search platforms for the LCS.

Satellites are not capable of real time targeting, contrary to what many people believe, and they won’t last long in a peer war.

So, where will the distributed lethality get its distributed targeting?  The short, simple, and painful answer is that there is no viable distributed targeting system.  The kill chain is missing a key link and the Navy’s distributed lethality is just another unworkable Navy fantasy without it.

The concept of distributed targeting is, however, viable with the right sensors.  Unfortunately, the Navy does not have the right sensors and, worse, seems to have no grasp of the problem and no intention of getting the right sensors.  The task falls, then, to us to define the right sensors.

The requirement is simple.  The sensor needs to be able to penetrate many hundreds of miles of enemy territory undetected, find targets, and transmit the data back to the shooters.  For the sake of this discussion, we’ll assume the shooters are ships, the LCS specifically, although the shooters could also be aircraft or land bases.

So, the sensor needs great range, long endurance, stealth of some form, a decent size/power radar and/or good optical sensor, and secure communications.  There are also a couple of implied characteristics.

Unless the sensor is a very large AWACS / E-2 Hawkeye / P-8 Orion size platform, the radar it carries (assuming it uses radar) will, of necessity, be small and low powered which means the field of scan will be limited.  This suggests that large numbers of sensors will be needed to make up for the limited individual coverage.

This, in turn, implies the characteristic of affordability.  Large numbers of sensors can only be produced if the individual sensor is cheap. 

By definition, many of these sensors won’t make it back.  This, again, argues for extreme affordability to be able to absorb the losses and costs.

So, having defined the requirements, what form of platform can meet the requirements?

Unmanned underwater vehicles (UUV) are stealthy but too close to the surface to have much sensing range and are too slow to cover much territory.  UUVs, then, would not make good general purpose distributed sensors.  They could, however, be useful for monitoring limited, fixed areas like navigational chokepoints or harbors. 

Unmanned surface vessels (USV), like UUVs, lack the sensor range and speed to cover sufficient territory.  In addition, they lack the inherent stealth of a UUV.

Unmanned aerial vehicles (UAV) potentially offer the range, endurance, and speed to cover larger areas.  Combined with the high altitude they operate at, the sensing area is correspondingly greater.  UAVs also offer the potential to be stealthy through a combination of small size (compared to an AWACS or P-8) and airframe shaping.

Blackjack UAV

UAV’s, then, seem to be the best choice for distributed sensors.  The concept of operation would be to flood a region with many dozens of UAVs at any given moment.  Although an individual UAV would provide limited coverage, the large numbers would ensure adequate area coverage and compensate for the inevitable high attrition rate.

The key question is whether the desired characteristics can be made to fit in an affordable package.  Can we build a UAV with great range, small size, stealth, and decent radar/optics for a low enough price to allow us to build thousands of them?  That’s a difficult challenge but that’s the part of the kill chain the Navy needs to be working on.  An LCS with a hundred or thousand mile anti-ship missile is useless if we can’t provide targeting.


Side note:  An upsized Blackjack UAV might make a good design starting point for a distributed targeting sensor.  The Blackjack has an operational radius of 480 miles and is small enough to be operated from any ship.  It has a degree of stealth by virtue of its size and the airframe could probably be shaped to provide a greater degree of stealth.

Alternatively, a downsized MQ-1 Predator might also make a good starting point.  It has an operational radius of around 1000 miles.



  1. Distributed lethality won’t work without distributed targeting. And distributed targeting won’t work without a way to generate a target position fix, and reliable datalinks to convey the target data from sensor to shooter.

    However, a peer enemy will be jamming GPS and comms. So, seems like two additional stated requirements for the distributed sensors should be:

    - Act as a self-assembling resilient mesh network that can use high-power line-of-sight datalinks to convey target info to a shooter over multiple hops in the presence of enemy jamming.

    - Act as part of a geolocation constellation for GPS-denied environments, consisting of fixed ground stations and moving air, surface, and land vehicles. Each node in the network has a current estimate of its own position in space and a confidence level, and can transmit that position estimate to other nodes over datalinks, including the high-power line-of-sight datalink laid out above.

    Worst-case, a node's own position estimate is from inertial nav or dead-reckoning. Intermediate-case, it's from triangulating from the positions of all the other available nodes. Best-case, it's from grabbing an intermittent GPS fix.

    (Obviously we’d need to take measures to prevent the enemy from spoofing us into revealing the positions of all our nodes.)

    Thoughts? Stating the obvious here?


    1. Obvious (though perhaps not to the Navy!) but timely and well said. The odds that a vast network will work in the face of electronic countermeasures is poor. Further, the reports that do make it back to a central collating computer will be fragmentary and contradictory, at best. That will be a challenge - to assemble a viable composite picture out of all the inaccurate pieces of data.

      The Navy is focused (a little bit) on the weapons and is ignoring the targeting link in the kill chain.

      Very good comment.

  2. May be a swarm of small stealth (CFRP?) UAVs with a passive RF receiver using machine learning artificial intelligence computers being developed by silicon valley for self driving vehicles is the way forward.

    DARPA is funding the probabilistic programming for advanced machine learning so as to understand data, manage results and infer insights from uncertain information with software package that is radically more effective.

    1. On paper, some kind of artificial, adaptive, learning software sounds good. In practice, it's impossible to do with our current level of computer programming. Don't believe me? How about the F-35 ALIS program which, by comparison, is just a simple data collection, preventive maintenance program. Look how that's been in development and how badly it's failed, thus far. We are many decades away from having the kind of learning program you describe. We should pursue it as a developmental effort but recognize that it is ages away from becoming a real world product.

  3. How many Blackjacks would we need for a swarm? And, how many can a ship carry?

    I'd imagine a Burke could squeeze one in and if you offload a helicopter, probably a couple of more. An LCS could probably carry a few too. The amphibious and supply ships could probably carry a few each too. Carriers could probably carry a dozen or so. But, we only have 2, maybe 3, carriers at sea at any given time.

    Unless we can operate enough UAV's from land bases, I don't think we have enough ships to provide adequate coverage for distributed targeting.

    1. A Blackjack UAV is 8 ft long with a wingspan of 16 ft. Folded up and packaged, a ship could carry many dozens of those.

      I've also suggested a small UAV carrier (commercial standards - essentially a cargo ship with a flat "flight" deck - that could carry and operate hundreds.

    2. But, the ship also has to carry the launch and recovery equipment, fuel, spare parts, plus additional crew to operate and maintain the system.

      We have more Burke-class ships at sea than other type, but I can't see them carrying more than a few UAV's in addition to everything else required to operate them.

      I like idea of a UAV carrier. But, if you're going to do that, you might as well give them some missiles and make them shooters too.

    3. The ship only needs one set of launch/recovery equipment. That's required whether operating one UAV or a hundred.

      As far as fuel, have you seen how small the Blackjack is? The max takeoff carry weight is around 55 lbs. Even if that's all fuel, that's only 5-6 gallons. A Burke could fuel thousands of UAVs.

      Spare parts? Fine, dedicate a locker to common spares. If it's a more difficult problem, pack the UAV up and store it away for shore-side maintenance. That's the beauty of operating many dozens on a given ship.

      Additional crew? It takes 2-4 people to launch/recover. It borders on a spare time activity and would be well worth the few extra crew for the surveillance provided.

      Missiles on a UAV carrier? That's the Navy's wacky line of thought. Would you really risk a UAV-carrier, with hundreds of UAVs, close enough to an enemy to be sunk? Remember, if you're in range to launch at them, they're in range to launch at you. A UAV-carrier would be a high-value unit and you don't risk them if you can avoid it.

    4. According to Naval Air Systems Command, the RQ-21 Blackjack has a range of 50 nautical miles. I'm guessing it has something to do with its control system. While it can stay in the air for 16 hours, it's limited, at present, in range.


    5. You got that I suggested the Blackjack UAV as a starting point for a design, right, not that it was the final solution?

    6. Yes. But, in your side note, you indicated the Blackjack has an operational radius of 480 miles.

      In upsizing the Blackjack for range and payload, one has to balance that against the ability of being launched and recovered at sea. Unless it can take off and land on it's own (which would affect range and payload too), what's the largest UAV you could catapult and recover on a ship?

    7. Walter, yes, I simply multiplied the endurance times the speed to get the range. If the range is limited by a comm/radio issue then that would need to be addressed.

    8. Blackjack uses the Scan Eagle type launcher and that is probably near to about as big as you can get without moving to some kind of small, built in catapult.

    9. Autonomy could solve many navigation related problems. But, at long ranges, you generally need a larger (and heavier) communication system. Which can drive the size of the UAV.

      The Blackjack UAV is probably the right size, maybe the trick is how they are operated. Instead of out-and-back flights, they fly point-to-point (e.g., ship-to-ship, ship-to-shore, etc.) missions for wider coverage.

      But, whatever is used, it shouldn't interfere with the primary mission of the ship. You wouldn't want a Burke hampered from launching a helo because you have to take down launch catapult first.

    10. The Blackjack/Scan Eagle launcher is just a little cart. It probably takes about two minutes to wheel it back into its storage. I can't imagine any interference with helo flight ops.

  4. I was wondering if you could clarify the detection range of the LCS. On wikipedia, it shows the LCS uses the following:

    Sensors and
    processing systems:

    - SAAB AN/SPS-77(V)1 Sea GIRAFFE 3D air and surface search radar[7]
    - Sperry Marine BridgeMaster E navigational radar
    - AN/KAX-2 electro-optical sensor with TV and FLIR
    - Northrop Grumman ICMS (Integrated Combat Management System)[7]

    The blog entry mentions the LCS has a radar range of only 20nm. To me, a layman, the radar range is the important one, and Sea Giraffe has an instrumented range of 180km, which is about 100nm.

    Which of the four components listed is the one needed for targeting other ships, or land targets, and why is the Sea Giraffe range not valid for targeting?

    Apologies if this is something your normal readers already understand.



    1. The Detection range of radar is highly confusing to someone looking from the outside looking in, it's fine.

      Although this is a gross simplification, it can be said that there are two main types of RADAR - 'Direct' and 'Over The Horizon' (OTH).

      Most Radar is of the 'Direct' type (which means it 'sees' in a straight line) and therefore can only 'see' the surface of the water as far as the curvature of the Earth will allow them - which is to say the horizon.
      Now, the RADAR will continue to look past the horizon, but only in a straight line from the RADAR unit itself, which is to say quite a bit high in the sky when you get past the horizon. This is highly useful for detecting and targeting enemy aircraft, but it is entirely useless for detecting enemy surface ships.

      OTH RADAR, on the other hand, 'looks' in a curve to match the Earth's curvature and will keep 'seeing' the surface of the water for potentially hundreds of miles. However, even though it's been being used for 50-something years now, the technology is still in its relative infancy and small form arrays capable of being fit on any but the largest of ship types just do not exist in a reliable form, which is why very few Navies have committed to the concept shipside.

      Hope this helped at all.

      - Ray D.

    2. Hi Ray D,

      Thanks for your excellent reply. It was very helpful, and I understand why the US talks about "networking" information between planes/ships so much now.

      Thanks again.


    3. A, to grossly simplify, each radar has two detection ranges: one for surface targets and one for aircraft. The surface target range is, essentially, the horizon which is around 20 miles and depends on the height of the radar. The air target range is on the order of a hundred miles and depends on the power of the radar.

      Imagine yourself standing on the ground and looking up at the peak of a roof on a house that's 40 ft away. You can see everything between you and the roof. Someone could be lying on the back side of the roof, 41 ft away and be unseen by you. At the same time, you could see an airplane in the sky beyond the roof peak, many miles away.

      In this example, the peak of the roof is the curvature of the surface of the earth. Thus, you can see how a given radar has a very short range surface detection and a very long range air detection.

      Hope this helps.

  5. It would be interesting to know if seabed sensors are, or can be, integrated into the picture at a sufficient quality for targetting.

    1. Probably not. By the time the analysis is performed by the controlling land station and the data makes its way up the chain to a transmitting facility, to the relevant Navy command, and back down the chain to an individual ship that can make use of the data, the target has long gone.

      This is similar to satellites which so many people think are real-time targeting but are anything but.

      That's not to say that you couldn't make some kind of one-time effort to link a single ship more directly to a satellite/seabed data collation site but that would be an enormously difficult exception, not the rule.

  6. Dumb Questions:

    A) Aren't 'Radar' and 'real time comms' not congruent with 'stealth'.

    B) can the saturation technique work like the large radio telescope arrays? I.E. a bunch of little radars make like one giant one? Which sounds good, but if I'm China, why wouldn't I flood the spectrum with noise?

    1. Any transmission is inconsistent with stealth. Some compromise stealth only a little (we claim) - others a lot.

      I think you're talking about networking radars. This is already a capability. This is one of the ways to detect stealth aircraft. The problem is that it requires a very sophisticated master computer to assemble the myriad bits of data. That has proved quite challenging. It also leads to a single point of failure. Take out the master computing site and you lose everything.

  7. I have to admit, I’m not sure the Tech exists.

    I very much like the initial comments of BTL at the top. Where I think he hits the crux of the matter.

    If a UAV type asset has a short range sensor it’s reasonable to assume it so “deep” in a jamming field it won’t be able to get the info out.

    If it’s a big Asset with large sensor arrays, it could well communicate, but then it won’t be organic to every ship OR able to form a long range net from a larger vessel. Plus we get into an expense discussion.

    My money is on a Blackjack sized UAV launch able and recoverable from a destroyer, perhaps not in swarms, but an Arleigh-Burke should be able to keep 3 orbiting & 200nm +, 24 hours a day given the UAVs endurance.

    Love the UAV mother ship idea, built to civilian standards.. Certainly sea lift vessels should be deploying these.

    A synthetic aperture scanned array Radar is available in the right size.



    There is a selection of interesting pieces on communications lasers, this will get round some of the jamming issues of a small UAV


    I also had an excellent video on UAV relay drones for a 3G phone company doing laser lock on communications one drone to another. But wouldn’t you know it when I need it …

    The UK is investing is pseudo - satellites. Long endurance very high flying UAV for surveillance and communications. Whether this fits the bill?


    I totally don’t think the “Kill Web” has thought of any of this though. I’m afraid I’m convinced it’s a PR announcement in support of the LCS and the distributed lethality (itself a work in progress).

    The other way to consider this of course is as an extension to CEC?


    1. Ben, all good observations and questions. The Navy has clearly not thought this all through and there has been no testing of data links and comms in an electromagnetically challenged environment. Until we start looking at, and testing, these ideas in a realistic manner, this will all be speculation and very unlikely to work in combat. Russia has shown the first inklings of electronic warfare in actual combat and it's been eye-opening even for the small amount of information that has made its way into the public domain. We have to stop pushing these ridiculous ideas that will only work on paper and start testing in a real world scenario.

      The nice thing about electronic warfare is that you can test under actual war conditions without killing people. So why aren't we doing it? Let's send a UAV or an F-35 a thousand miles away and subject it to the most comprehensive ECM we can and see if it can see anything and data link/comm back to us. I'm betting it can't.

      Related side note: were you aware that the F-35's main comm method, MADL, can't be received by any other platform? Who thought that was a good idea?

    2. RN has been running trials last year networking Type 45 and Merlin ASW helo and F35 sucessfully. All carry very modern AESA. Perharp MALD is software transferable to the right AESA ?

      Range was just around the English channel though. So very close.

      Russia was recently showing off a theater level ECM truck. 100 mile range. So if anything they are much more into their ECM than NATO.

      We do have to start running red flag etc in a high EM environment. I cant believe the Navy doest train ESM. I know all RN frigates carry extencive EW suites. I thought USN did too ?

    3. We do have EW systems. We just don't train for EW to any great extent.

    4. The fact that its taken us this long to update SLQ-32 and our jamming pods suggests to me that, like ASW, we (the USN) hasn't taken EW all that seriously.

    5. There are a number of com links to the F-22/F-35 such as BACN, Talon Tate pods and Freedom 550 gateway as part of the 5th to 4th initiative. The notion of MADL/IFDL links to legacy platforms that are not available and not tested is simply not true.

    6. It is absolutely true that the F-35 MADL cannot be received directly by any other platform. As far as "translation" devices/platforms, I'm not aware that BACN can translate MADL to anything. Give me a reference link if you have one. Freedom 550 is a developmental effort, not a standard unit as far as I know, and is geared at the F-22 not the F-35, as far as I know. Again, give me a reference link to the contrary if you have one. I have no idea what Talon Tate is - presumably another developmental item.

    7. It is important to put things in perspective when discussing com links with F-35/F-22.
      (i)F-35 can receive and transmit using Link 16
      (ii) F-22 can only receive using Link 16
      (iii)There are existing gateways available to receive IFDL waveform using either (a)Talon Hate; (b) Legion Pods; (c)Freedom 550; or (d) BACN
      (iv)In regards to MADL waveform, as the F-35 is still in SDD phase various options are still being considered. For example, there is no technological hurdle to install MADL onto a BACN aircraft to facilitate LPI communications.
      (v)MADL by nature is designed to ensure LPI and so there could be a limitation of distance. Speculation on distance I have seen being offered ranges from 25 nm to 125 nm. There is nothing stopping the F-35 from communicating using Link 16 except that would compromise its stealth.
      (vi)Distributed lethality is still a developing concept that is yet implemented and so it could be argued that some form of com link solution will be provided concurrent to some form of demonstrated need in the future. I think it is narratively false to say that there are no solutions at hand.

    8. It is important to keep additional things in perspective when discussing F-35 comms.

      - The F-35 proved incapable of reliably communicating with the ship during the Marine's operational evaluation.

      - None of the comm gateways are standard deployed equipment. They are developmental efforts that may or may not pan out. If there was one that worked and met all needs, there would only be one. Instead, there are a half a dozen or so.

      -Gateways require a second aircraft for the F-22/35 to function. For example, an F-22 can data link to an F-15 with Talon HATE but now the mission requires two aircraft instead of one. This is a severe penalty.

      -I have yet to see that any gateway specifically addresses the F-35 MADL. One or more may but I have not seen that yet.

      This comm issue is yet another example of the military focusing on the acquisition of the platform rather than the concept of operation. If anyone had done a CONOPS for the F-35 they would have quickly realized that it can't talk to anyone without compromising its stealth - kind of a serious shortcoming!

  8. If we hooked an AESA on this...


    ... then got the LCS up to 40 kts or whatever then it has a radar mast like 200 ft in the air!!!!


    1. There have been many proposals to "raise" a ship's radar via blimps or any other scheme. All would work. All are unwieldy.

      A simple UAV solves the problem and is reasonably easy to launch/recover. The only challenge is balancing the conflicting needs: large radar versus small payload, high speed/endurance versus small body, great stealth versus low cost, and so on. As I said, a Blackjack UAV seems like a decent starting point, design-wise.

  9. It seems your plan is to have the UAV's penetrate enemy air space. If it does have a radar could you just have it loiter above the ship? It won't give you super long range, but even a 20 mile increase would be helpful.

    A fire scout has a paylod of 600 lbs. How much does even just a decent radar weight?

    1. In the context of the Navy's distributed lethality concept (insanity but we'll discuss it), the weapons are 100-200 mile range. That means we'll want to scout out to about 400 miles so we don't wander into the same kind of anti-surface trap that we think our enemies will obligingly wander into. An additional 20 mile sensing range is of little use.

    2. Okay. The Navy and I are thinking in opposite directions then. I'm thinking:

      Given: The LCS program will continue
      Given: The LCS native firepower stinks.
      Given: The LCS will be armed with OTH missiles
      Given: The LCS will not have sensors to support said missiles
      Given: The LCS can support a Fire Scout

      With all those, lets find the simplest, quickest, cheapest way possible to get the LCS to improve its weapons load out.

      The NSM has a range of > 100 miles. If you could get the LCS an ad hoc ability to target out to 50 miles using mostly existing equipment (existing radar, added to a platform we know the LCS can support) you at least get an improved range and real punch to the ship relatively quickly.

      There are a raft of draw backs (You have to deploy the firescout to shoot. It has to stay aloft while you shoot. You are dependant upon an asset that likely has lousy uptime...) but at least you could concievably get something quickly.

      The Navy seems to be depending upon something that doesn't exist in any sense, and even given your blackjack idea, is going to be difficult to make happen and maintain.

      Are they thinking an LCS can get data links from a 'burke?

  10. All od the discussion is good but it reminds of the Bomber bosses before WWII. They were convinced that a Flight of B-17s with their 13 (Model G) machine guns could fly through any defense by itself. How'd that work out for all of the downed aircrews?

    This distributed silicon valley pie in the sky targeting and lethality maybe made to work. But it has to be tested in a wartime environment to be sure. Else we will go to war with a Fleet that has only 5 machine guns (the Early Model B-17s) and we may not have time to have a P-51 come along and save us.

  11. There is one other problem with this idea of distributed targeting:

    Everything must work perfectly together in synchronization. Baring extremely realistic tests, I'm not sure that is a good assumption.

    Murphy's law would like a word with you. If you have a point of failure, then it will not work.

    The other is of course, as others noted, the enemy will realize the weak points and try to jam/take out the "eyes", which are often very vulnerable.

    1. You've been following this blog long enough to know that my view of the Navy's dependence on vast networks of sensors, data, and comms is foolish. You also know that I've called for extensive and realistic testing.

      The distributed targeting concept can only work if the comms and data link issues in the face of actual ECM can be solved.

  12. Thanks to Ray D and CNO for their explanation of how radar and targeting information works.

    If the USN has such restrictions, don't the other navies- Russia, China, India etc, also have such restrictions ? Doesn't that render their 300nm+ range hypersonic anti ship missiles moot, given their sensor networks are likely worse than the US's? (they lack the presence of P8's etc.)


    1. Of course they do! That's why the Chinese "carrier killer" ballistic missile, for example, is not a credible threat. That's also why the Chinese artificial islands represent a threat - they'll provide for the long range targeting that is needed for the ballistic missiles to function. We're allowing them to establish a credible long range threat.

    2. CNO- thanks for your reply. I'd only thought of the islands as staging grounds for hardware- planes, missiles, and plain rudeness- putting an island in the backyard of several countries, not in terms of sensor/targetting information.

      I have a better in depth idea why people say Obama was soft in foreign affairs.


  13. Good point A.

    The only counter I could think of is that some of their stuff, like the Shipwreck, have the ability to share targeting information between missiles in a raid.

    They still need the initial targeting information of course, but if they can get that the lead missile can acquire the target and then share the information amongst the rest of the flight.

    1. The Soviets were always on the leading edge of long range detection and some of their methods are still cutting edge.