At last fall’s Euronaval, Israel Aircraft Industries (IAI) displayed an important inherent feature of an active-array radar it had developed. Passive phased arrays such as the U.S. Navy’s SPY-1 (the Aegis radar) use antennas consisting of phase shifters connected to radiating elements, all fed by a single power tube. Together, the individual radiators create a beam. Because they are under computer control, and because they can act quickly, the phase shifters can steer that beam electronically. In an active array, there is no single power tube. Instead, each element of the array radiates a signal with a computer-controlled phase. A passive array creates a single beam, because it feeds from a single power tube.
Advocates of active arrays point out that such arrays can create several beams simultaneously, simply by grouping emitters in different parts of its antenna. The alternative, for a passive array, is to switch between different radar roles, sacrificing overall effort in any one role. An active array is also better adapted to countering jammers by nulling them out (i.e., by ignoring signals from some selected direction). On the other hand, splitting up the transmitters on the face of the array results in a series of beams, each of which is not as sharp as one created by the whole face of the array. There are probably also issues associated with making sure that each individual radiator produces the same output (otherwise the beam will not be shaped or steered as desired). In this sense a passive array is far simpler. It may also be difficult to squeeze sufficient power from individual radiators, and active arrays require that each element be cooled (in a passive array, only the power tube, deep in the ship, requires cooling).
Active arrays are now becoming common in fighters (where they are called active electronically scanned antennas, or AESAs), and are also at sea in several navies. It is easier to produce the necessary modules in high than in lower frequencies, which is why AESAs are more common on board fighters than on board, say, frigates. Most naval active arrays operate at X-band (SPY-1 operates at S-band). The exceptions seem to be the British Sampson (S-band) and the French Herakles; the latter uses a combination of an active array and an electronically maneuvered lens.
IAI displayed its MF-STAR active array, the electronically scanned successor to the target-acquisition radar associated with the company’s Barak anti-missile weapon. MF-STAR is to equip modernized Eilat-class corvettes, and there is at least one export customer. Each face of the array carries 168 transmit/receive groups, each of which consists of 16 transmit/receive modules. Signals are converted to or from digital form in the face of the array, so the connection to the face of the radar is fiber-optic cabling (plus power). The radar operates at C-band (5 cm), which has long been a useful compromise between X-band (10 cm) and the better weather performance of S-band (3 cm); the U.S. Navy uses C-band for illumination in terminal missile homing.
What is significant is a new way to use a T/R module. IAI wanted a counter to the small anti-tank missiles that were being fired at inshore patrol boats. There was no question of shooting back; the time line was far too short. However, it was clear that a boat could evade attack, if it could receive timely warning. IAI therefore proposed a system it called Naval Guard: a single MF-STAR module (16 transmit-receive elements) on each of four sides, supplemented by a short-wave infrared detector. A boat equipped with the new system can, at least in theory, pull away in time to avoid a hit, while firing multi-spectral decoys to confuse the incoming missile. The company expects to add a laser warning capability during 2011.
Naval Guard was not costly to develop because an active array is inherently scalable. The modules already existed, and it was not too difficult to adapt the software developed for the larger MF-STAR. By way of contrast, a passive phased array is not easily scalable, although considerable work has been done to adapt SPY-1 to smaller ships such as frigates and even corvettes (in an abortive SPY-1K version). The use of the single power tube places a lower limit on overall radar-array power. Making a really small SPY-1 would require development of a much less powerful tube. Conversely, the radar may also be difficult to scale up, because making it larger requires a more powerful tube. At the least, a radically different radar function, such as ballistic-missile defense, requires a power tube with different characteristics than one adapted to, say, fleet air defense, which is a major reason why Aegis ships have to be adapted to one role or the other.
It is not clear whether an active array would offer a simpler solution. Of course, if each of the transmit/receive elements would have to be replaced for higher overall power, then the active array would be far more (rather than less) awkward to adapt to the new mission. Even so, the speed and low cost of the Naval Guard adaptation of MF-STAR technology is stunning.
Too Much Information?
Recent events in Afghanistan have highlighted an important and pervasive problem: information overload. In this particular case, U.S. forces attacked Afghan civilians and killed 21 of them, including children, because the concentration of Afghans was interpreted as an emerging threat to the American troops. Investigation revealed that U.S. intelligence was aware that there were children in the crowd, hence that they were almost certainly harmless civilians. What went wrong illustrates that this account is far too simplistic.
Saying that U.S. intelligence was aware of something means merely that someone within the intelligence community had the key information, not that anyone in the chain of command was aware of it. It turned out that the air controllers who ordered the strike were flooded with information. They focused on what seemed most important. The information about the women and children in the crowd was available but went unread (or worse, read but forgotten) in the crucial minutes leading up to the strike.
The tide of information has been rising, both in the military and in society at large. It’s always assumed that more information is better; it’s better to be watching multiple screens, with multiple feeds, than to be confined to a single screen or, worse, to a single bit of text. Distractions are continuous. Discussions of information focus on bandwidth: can we provide enough bandwidth to carry more? Certainly bandwidth is an important bottleneck. However, a human operator also has limitations. The brain has an inherent cycle. Training can bring an operator closer to maximum performance, but the point is that there is a maximum. Information received more quickly than the operator cycle can handle it is either missed altogether or eliminates concentration so that other information is missed.
To some extent automation can help, if incoming information can either be edited down automatically or turned into more easily assimilated form, which usually means graphics equivalent to text. Sometimes the distinction is made between information and knowledge, the latter being information processed to concentrate on what matters. Unfortunately, it is not clear that any automatic system can decide what matters in a rapidly changing tactical situation.
In the civilian world, the effect of information overload seems to be to wipe out any kind of concentration. Those presented with too much information often become fixated on whatever has just arrived, rather than on what matters. If this sounds too abstract, think of how you deal with e-mail. It seems to demand instant attention, and whatever came in most recently often gets a lot more attention than the possibly more vital message that came in an hour ago. A few months ago The New York Times reported a case of a computer entrepreneur who became fixated on a news report of a murder in another state—and forgot the e-mail offering to buy his company. He had trained himself, he thought, to multitask, but he was unaware of what that meant.
Now go back to the operator cycle. There is only so much time available to process information. The more separate items the brain tries to process at any one time, the less effort it can spare for any one of them. There must be a minimum acceptable level of effort. The brain solves the problem by simply dropping what it decides are less important items. Oddly enough, this is the same thing that happens when a passive radar array tries to multitask. It can certainly switch from one role to another, but it has only so many resources to spare. Spending half its time, say, on horizon search is like having half as good a radar doing that full-time. It is not like having both a horizon-search radar and a volume-search radar operating more or less simultaneously by interleaving what they do. At some point the radar does neither job acceptably, and an intelligent operator just cancels one of them. The brain is no less intelligent but its choices may be unfortunate, and it is not clear how an operator can be trained to make better ones.
It is also not clear to the operator when multitasking creates mediocre, rather than obviously unacceptable, performance. There is considerable anecdotal evidence that multitasking dramatically reduces the depth of analysis; operators (and others) find it less and less possible to take time to think things through. This logically leads to disasters—like killing innocent people or not connecting the dots to forestall terrorists.
We may have to accept that a lot more individuals are needed to process information. That may mean that single-seat aircraft that collect vast amounts of information in the expectation that pilots can do everything (i.e., the Joint Strike Fighter) may not be the best choice. It may also mean that the multiple screens that have sprouted everywhere have not been such a good idea.