New arenas for defense electronics

Modern electronics offers the military technical solutions to many of the problems soldiers must overcome to fight effectively on tomorrow's battlefields. Consider fratricide (friendly fire): During Operation Desert Storm, allied forces took 613 casualties, with 146 soldiers killed in action. Friendly fire caused 72 of the casualties and 35 of the deaths. No commander wants to kill his own troops. But until recently no means existed to reliably identify friendly forces in the confusion of battle.

And consider the costs involved in turning recruits into soldiers capable of using modern weapons systems. Today, being all that you can be requires lots of time driving real equipment, burning real fuel, and making your mistakes in situations that can be physically dangerous. The two defense electronics systems described in this article will make it possible to reduce the high cost of training and, in most circumstances, prevent fratricide.

Who goes there? A secure, millimeter-wave question/answer system, the Battlefield Combat Identification System (BCIS), designed by engineers at TRW's Space and Electronics Group in Redondo Beach, CA, addresses the problem of friendly fire. Operating in the K_a-band at approximately 38 GHz, BCIS consists primarily of a K_a-band interrogator/ transponder unit, a high-gain narrow-beam interrogation antenna, an omni-directional transponder antenna, and a display/interface unit. Vehicles that carry weapons (shooters) will employ all four parts of the system. Unarmed vehicles (non-shooters) don't require an interrogator antenna.

On the battlefield, a crewman on the shooter platform views the target through a gunsight-mounted laser range finder. Slaved to the main gun on the shooter, the narrow-beam antenna transmits a pinpoint signal aimed at the target. "The potential target receives that signal via an omni-directional antenna," says Andrea Yeiser, TRW BCIS program manager. Next, she explains, the target "verifies that the interrogation is a valid signal, and sends back via the omni-directional antenna an answer that says, essentially, 'Don't shoot.'" Within one second, visual and audible cues tell the crewman if the target is a friend or an unknown.

Engineers use the narrow-beam interrogation antenna to receive the target's response. The antenna's high gain makes it possible to close the communication link at long ranges. In addition, the BCIS on the target vehicle tells its operator that an interrogation signal was received.

"The hardest part of this program for us was the schedule. We had about a year from design to production," says Yeiser. "We began the project in September 1993, and delivered our first production units in December 1994." Field trials of the equipment began in January 1995, and TRW's BCIS demonstrated effective ranges of 6.1 km to 7.8 km in conditions ranging from heavy rain to clear weather. Required range was 3.0 km in rain and 5.5 km in clear conditions.

Onboard smarts for BCIS comes from an 80C186 processor running at 10 MHz. It comes with 256 kBytes of RAM and 768 kBytes of flash ROM. The interrogator/transponder system is designed around GaAs Microwave/Millimeter Wave Monolithic Integrated Circuits (MIMICs). This technology enables the system to operate at high frequencies and high data rates with lower power consumption than silicon circuits.

To help reduce the overall cost of BCIS, Yeiser and her colleagues aimed at using commercial parts throughout the system. "We're using non-Mil-spec parts in about 90% of the hardware," she states. "What we've managed to do is reduce the design-to-production unit cost we had in 1993, when we were awarded the contract, by about 72%."

In addition to simply identifying a target as friend or foe, the BCIS can send, receive, and display information about other units on the battlefield. When the BCIS interrogates a target, the target responds by indicating that it's friendly, and also uses this digital data link (DDL) capability to transmit other information via its omni-directional antenna. These data include GPS position information and a vehicle code. The interrogator captures the data, and, by doing so, builds up a map of BCIS-equipped vehicles in the area. At regular intervals, the shooter broadcasts that information using its omni-directional antenna, which can deliver a signal across a little more than 1 km.

Neighboring vehicles pass the data on to one another. "Within a short time,"says Linnie Haynesworth, advanced systems manager, BCIS, "platoons and companies can share information about who is a friendly platform within a short range." Information other than vehicle ID and position could be transmitted, if the Army so desired.

"Timelines for the DDL are processor time-division multiplexed. So transmitters send information in a specific time window," Haynesworth explains. "Passing of data from one company to another occurs in another time window. So it's clear to the receivers when units will be sharing data, and other units will know where the data comes from."

Within BCIS, software carries out data maintenance and removes obsolete data. At regular intervals, the system refreshes the data as new information arrives.

To date, TRW has delivered 46 BCIS units to the U.S. Army, and the company has contracted to build and deliver another 46. Sales to other services may follow: "Right now our contract is with the U.S. Army, but we also have some activities ongoing that put BCIS onto some of the other services' vehicles, mainly the Marines," says Yeiser. If funding comes through, full-scale production of the BCIS for the Army will start in the late 1990s or early in the next century, with battlefield deployment shortly after that.

War without tears. In a time of reduced defense budgets, simulators can help the military train individuals and teams of soldiers without consuming expensive equipment and limited funds. But until recently, most simulators could train only the individual or crew using them. In the real world, soldiers operate as members of units that must move and fight in a coordinated way.

A joint project carried out by the U.S. Army and prime contractor Loral Corporation, the Close Combat Tactical Trainer (CCTT), employs the IEEE's Distributed Interactive Simulation (DIS) protocol to network virtual environments. Doing so allows soldiers to practice maneuvering and fighting as a unit.

A typical CCTT configuration includes simulators manned by crews and a tactical operations center for command and control. Loral augments the manned simulators with other forces that exist only in software. To do so, engineers at Loral create software entities that follow the tactical doctrines of friendly and opposing forces. On the virtual battlefield, these entities look and behave like the manned simulators.

"We spent a lot of time with the users developing a translation of their doctrine," says Nick Ali, Jr., vice president and general manager simulation programs/products. "We established how vehicles, soldiers, and decision makers would behave when confronted with certain situations. We've essentially gone through the process of documenting that information, specifying it, and then including it in our software."

Loral uses IBM's RISC family to drive its simulators, along with the Evans & Sutherland ESIG 2000 high-density image generator. Even prior to November 1992, when Loral began working on CCTT, Ali's team carried out an exhaustive search of a number of technical areas. Networking technology came in for special attention to determine whether or not commercial systems could provide adequate bandwidth. Today, CCTT uses Ethernet within the simulators and FDDI as the local net configuration.

Current plans call for linking sets of simulators at one site. But the CCTT system permits the U.S. Army to link separate sites and--for example--have soldiers in Korea carry out an exercise with soldiers in the U.S.

Engineers put considerable effort into studying the ability of a commercial operating system to support the real-time characteristics of CCTT. "We did a lot of benchmarking to convince ourselves that the IBM AIX operating system would do the job," says Ali. He reports that his team left the AIX operating system alone. "We have used it as a commercial platform. We're on our fourth processor, and we have in each case ported our code to the next generation processor, not in days, not in months, but literally in hours." Currently Loral uses use the PowerPC as their basic CPU.

Using AIX--IBM's flavor of UNIX--"has verified our belief that we could design today's top technology in a pretty open architecture," says Ali. "The system is almost exclusively Commercial Off the Shelf (COTS) or non-development items. There may be places in the simulators where we use a nut, bolt, gage, or something else off an actual vehicle, but it's almost 100% commercial hardware."

Loral's CCTT development program employed approximately 250 engineers and designers, including programmers, hardware design personnel, and manufacturing people. John A. Sorokowsky, program manager/CCTT, says building the manned simulators proved a bit tricky. "Almost every single M113 (armored personnel carrier) has been customized by the troops who own it," says Sorokowsky. "So you get the drawings and then you say: 'Okay, what is really out there today?' We brought in real vehicles, plus the manufacturers' drawings."

Monitors represented another challenge to simulator development. Loral engineers use COTS PC high-resolution monitors, "but the average COTS monitor has a life-expectancy of probably a year or less," says Sorokowsky. When Loral started working on the project, the engineering team selected 17-inch monitors with RS232 inputs. "You don't need those on a PC," says Sorokowsky, "so they're taking features away from monitors today and putting in cheaper boards, cheaper components. We have to trade off--we're not Mil-standard--to get a good monitor."

Such problems become especially difficult when engineers must form-fit a monitor into an existing simulator. "When they go end-of-life on a particular monitor, there's not necessarily one of identical weight and shape, power, and everything else," says Everett A. Goodwin II, development manager, CCTT IDT. And he points out that the same observation is true of speakers. "Speakers last six months or less as a product and a commodity. Go to the store and start measuring speakers," says Sorokowsky. No "standard" speaker exists. So finding speakers and monitors for simulators, along with other hardware matters that appear trivial--not engineering problems at all--become rather complex issues that engineers must overcome.

Loral sought to achieve high reuse of software across systems and between all the different modules and different functions in the CCTT. "As we went along, we got up to 50% reuse for each build," says Goodwin. Obviously, achieving high software reuse improves productivity and reduces design issues, because the code has previously been tested and used successfully. On the other hand, it's true that the last person in the reuse chain may encounter problems. "Everybody else has a chance to use it and modify it. And the last guy who gets it--it isn't what he thought it was to begin with," says Goodwin.

Software writers typically feel a great deal of ownership for their product, which can produce friction when software gets modified or rewritten. To achieve high reuse of software, Loral set out to tear down internal barriers by imposing a concurrent engineering program. Software engineers from different companies were required to work in the same office and use a common tool set.

The U.S. Army began testing CCTT in February of this year. Assuming that the tests go as expected, at the end of 1996 Ali expects the Army to buy 500 CCTT simulation systems. Other markets will exist for CCTT systems in foreign markets. Two such markets now being explored by Loral are Britain's Ministry of Defense and Saudi Arabia. Many other applications may exist for the basic technology in a range of civilian industries that need to train or educate personnel.