Last fall, the Naval Research Laboratory (NRL) successfully tested a shared-aperture antenna. It employed a separate active-array transmitter and an array of receiving elements. This was probably the first such device, and it has increasing significance for warships. For decades, a problem in warship design has been how Io provide enough antennas for an increasing number of functions. Satellite dishes already compete with radars for the best positions in a ship. The satellite antennas become more important with the development of networked warfare concepts. In effect, whatever is connected to the ship by satellites becomes a sensor for the ship, often the most important sensor. That makes the competition for positions urgent, and the choices difficult.
Placing numerous antennas near each other makes for interference problems. This is not trivial. Look at what happened to HMS Sheffield in the Falklands. The British knew her radar warning (ESM) set would pick up side-lobes from the ship's satellite transmitter. Because the satellite system worked at radar frequency, the ESM system would produce an alarm indicating the ship was being illuminated by radar. The solution the British adopted was to turn off the ESM while transmitting by satellite. It was their bad luck that the Argentine fighter popped up and fired an Exocet at that moment. After that, the solution was to burst satellite transmissions and to blank off the ESM receiver during the bursts. As the number of off-board sensors grows, that choice becomes less acceptable.
This is not futuristic; we already use off-board sensors in the cooperative engagement capability (CEC) system, in which ships share fire-control-quality radar data. How would a ship's commander balance the value of what he gets over the CEC link against that of his own radar? It often would depend on circumstances, because in some cases CEC would be the only way he could see a hidden target. How important would the CEC antenna be, and how important would it be to keep it clear of electronic noise?
Right now, it is difficult to design a ship so her satellite dish can maintain contact with one key satellite. Other dishes, operating at comparable frequencies, receive downlinks from airborne sensors such as those on board unmanned aerial vehicles. It is inevitable that future combatants will have to maintain contact with several satellites and aerial vehicles. The need for such information is likely to be greatest in the littorals, where the fleet will be supporting operations ashore. Yet, there is no reason to imagine that future combatants will be able to spread out their antennas any more than at present.
Indeed, things will be getting worse. Two command ships (LCCs) designed in the late 1960s had hulls based on small aircraft carriers to allow them to spread out simple directional high-frequency antennas. Those antennas are gone now, but the large decks of these ships have provided space for numerous satellite dishes for continuous coverage. These ships are about to be retired, almost certainly without replacement. The fleet still will need much of the data stream they currently handle. It can be argued that future computer systems will reduce the number of people needed to understand that data, but the data still has to come to the fleet.
Even without satellite dishes, interference between antennas can be unfortunate. In the 1970s, Naval Sea Systems Command began to analyze the configurations of U.S. warships in greater detail, to see what features determined the delay between the approach of a threat and its engagement. Subtle conclusions emerged. Once radar picked up a target, the information was passed to fire control. Because the radar beam was broader than that of the fire control set, there was a delay. It turned out that antennas around the search antenna could have significant effects, because like that of other radar, the beam of the search set had back- and side-lobes. Whatever it picked up in those lobes it would confuse with what was in the main beam.
Collapsing multiple functions into a single array is an attractive proposition. The technical requirement is that the array elements operate over a broad frequency spectrum and over a wide power range. In NRL's test, the transmitter and receiver arrays are separated widely so they are unlikely to interfere. That has some interesting consequences. One is that there is no need to blank out the radar receiver when the radar transmits. Similarly, the radar can easily be adapted to stealthy operation, as in the current Dutch SCOUT, in which transmission is continuous rather than pulsed.
Elements of each array operate independently, the controlling computer grouping elements together to form beams. Multiple independent transmitted beams can be created simultaneously. The system might track several targets or execute different radar functions at once. Even with a passive phased array like a SPY-1, the radar creates one beam at a time, so time spent on one function is taken from the other. With a shared aperture, resources can be split up. Because the split is controlled by software, it can be changed. The array can send messages at the same time that it operates as radar.
Separating transmission and reception has other consequences. In conventional radar, the acts of reception and transmission are linked. If the link is by software, then it can be changed to suit circumstances. For example, one might think of a jammer as a kind of reverse radar, the transmitter responding to what the system receives. Alternatively, with the appropriate processing, a system conceived for jamming might also be an effective radar. If this sounds odd, think about the way sonars have worked for decades, with separate receivers and transmitters. Modern sonars are phased arrays. The receiving element is normally wide open, using preformed beams. Such a receiver can pick up submarine signatures, but if it has a broad enough bandwidth, it also picks up enemy sonar. The mechanism that generates the sonar pings originally was disconnected from the receiver so the pinger could scan freely, the system not waiting for a ping to come back before looking in another direction. In theory, however, the pinger could be used to jam enemy sonar. The differences in radar are wavelength/frequency and time scale. It just takes a lot more computer power to make the same thing work in the radar world. The reception pattern can be shaped to null out jammers, an advantage already claimed by British developers of the SAMPSON active-array radar.
In effect, the shared aperture highlights the potential merger of communications and radars. Both radio and radar transmit and receive electromagnetic signals. In the past, the difference was in what happened between the two. In radar, the emphasis has been on switching quickly to reception as soon as the transmission ends. In a radio, the interval is longer because , whatever is received has to be interpreted, and whatever is sent has to be thought through. Thus radios lack the automatic switchover capacity typical of radars. The waveforms used by the two kinds of equipment also are different.
The United States is developing software-based radios (JTRS, the Joint Tactical Radio System). As in the shared aperture, the key to JTRS is wideband digital reception and transmission. Software connects the two functions. Typically, the receiv- " ing end is tuned by software, and whatever is received is digitized and then interpreted. Whatever the operator wants to send is turned into a software-generated waveform that goes out at a soft-, ware-controlled frequency. Because electromagnetics are linear, such a device could handle several signals in parallel. The frequencies do not match, but one might come to see the shared-aperture system as a stack of JTRS radios. Successfully producing shared-aperture systems would be valuable not just for the fleet, but for all branches of the military.
If the future lies in merging radar and communications in common apertures, one might ask whether some deeper merger also is possible. For example, might we want to merge some communications and sensing elements? CEC goes a long way in that direction, because its data stream includes transponder pulses that allow ships to measure distances from each other.
Another question that all active arrays raise is whether future radars and other antennas have to be concentrated in single panels, as at present. Antennas always have been vulnerable because of their location in a ship. A big fixed phased array, like that used by Aegis, does better because it can continue to operate with many elements out. It is still a prominent target. Future active arrays may be even better targets, because they may give off heat.
The cost implications are interesting also. An active-array system is built out of standard electronic components that seem to be in cost free-fall. The main expense is developing the software that drives the system. Once that exists, its cost can be spread over many platforms using the system. If the hardware proves inexpensive, it will pay the Navy to use the system frequently. These platforms would add to the fleet's radar picture. There would be an enormous fleet electronic defense advantage also. The U.S. Navy standardized on the SPS-49 radar years ago hoping Soviet radar-detecting satellites would be unable to distinguish carriers from missile frigates. Now the frigates are going, and we are unlikely to put SPY-1 on minor surface warships. But if all U.S. warships had the same flexible radar transmitters, then all of them would offer similar electronic signatures. It might be that some would have larger arrays offering narrower beams, but that probably would not be obvious to a listener. This, of course, is not to mention the logistical advantages that full standardization would offer.
All of these possibilities are well off in the future. For the present, NRL demonstrated that the shared-aperture concept works effectively, generating multiple beams and rapidly changing functions to adapt to different scenarios. That should be enough to make the concept attractive for several current and projected classes of ship.