Optical fiber will have a bright future in military aviation. It offers high bandwidth, light weight and immunity from electromagnetic interference. It was deployed on the AV-8B decades ago, it's used on the F/A-18E/F and other aircraft, and will be fielded more widely in coming years. But before fiber-based coms in severe aviation environments can proliferate, standards are needed in component, testing, training and other areas.
Beyond the immediate horizon, techniques such as wavelength division multiplexing (WDM) could dramatically increase the throughput and reduce the footprint required today for digital and analog data communications.
Fiber optics standardization in the aerospace sector today leaves something to be desired. There are no standards for: Inspection: The examination of a cable installation to verify its performance.
One size won't fit all, however, as different approaches are required for different environments, says Boeing's David Zicka, chairman of SAE's Fiber Optics and Applied Photonics Committee. Different polish standards, for example, would apply to non-contact vs. physical contact connections. But within those areas, "everybody should meet a [termination endface] range for that type of polish." Another important parameter would be the cleanness of the endface.
Test jumpers have endfaces, as well, and their polish requirements need to match what's being installed in the aircraft. Without such a standard, you could end up with test jumpers that don't work on the airplane.
Standards for components such as endfaces are necessary in order to "eliminate subjectivity on the engineering side," Beranek says. It's particularly important, now that aircraft with fiber optic systems have been bought, to establish a baseline for the avionics technicians, to instruct them on what a good connector endface looks like." He stresses the point that, "without knowing what we're getting in the first place, how can we train?"
Work on inspection concerns the magnification and other criteria required of the equipment used to examine the termination endfaces and check for damage. Tiny charged coupled device (CCD) cameras or video microscopes, for example, are mounted on probes. Two-hundred-times magnification seems to be appropriate for the multimode fiber and 400x for single-mode fiber, Zicka says.
Termination components are made by connector suppliers, who also participate in the SAE process. Beranek expects a standard encapsulating JELLI's work to be published in about six months.
An SAE industry/government task group also is looking at how to calculate the link loss power budget for digital avionics, something that's never been done before, Beranek says. Link loss power budgets typically are calculated using "worst-case," statistical and numerical methods. But there is no standard approach to the assumptions engineers should apply with the three methods. Although this type of back-to-basics work is difficult and time consuming, it's important in order to understand exactly how a developer came up with a number and to be able to require levels of performance across platforms. Once the standard is developed, the government, for example, "can say that the link loss power budget will have a certain margin based on SAE xxx," Beranek says.
At a high level the areas of interest of the link loss power budget group include transmitter, receiver and cable plant characteristics. Participants are looking at areas such as transmitter output power, jitter and extinction ratio; the receiver sensitivity, saturation and quality of the input waveform; and connector loss.
The first step is to define how to calculate connector loss, receiver sensitivity and transmitter output power, Beranek says. But it's not straightforward. Is connector loss, for example, defined as mean loss (and standard deviation)? Is the extra loss from vibration included? What figure is incorporated to allow for connector degradation in the fleet?
The SAE group has been working on link loss power budget issues for three years, but expects to have a draft in the next six months.
Another fundamental building block is training. Unless technicians are trained to handle fiber, there is little point in having it. The training task group in the SAE's supportability subcommittee focuses on top-to-bottom awareness among those dealing with optical fiber--including expeditors who move cable from the shipping dock to the manufacturing areas, installers, inspectors and repairers, as well as designers, purchasers and managers. "We're trying to cover the whole fiber optic chain," Zicka says.
The task group's document will be known as Aerospace Recommended Practices (ARP) No. 5602, or "A Guideline for Aerospace Platform Training and Awareness Education." Zicka expects it be published next spring.
The objective is not to specify a training curriculum, he says, but to describe a set of skill levels or examples, a baseline around which curriculums can be built.
One procedure Boeing has used since the 1980s is to include violet thread in the harness overbraid in order to alert aircraft maintainers that fiber is present in the harness.
The Naval Sea Systems Command (NAVSEA), with support from NAVAIR, has established a working group to develop a standard for a next-generation, heavy-duty, multifiber connector applicable to both aircraft and shipboard needs. The goal of the Next-Generation Connector (NGCon) initiative is to develop a more maintainable connector that also takes advantage of commercial sector technologies, Beranek says.
Currently, aviation uses the Mil-Dtl-38999 connector, originally intended for electrical applications but applied to fiber optics, as well. The new NGCon connector document will be a "Mil-Prf," or performance standard. It will incorporate features that are available today, but from just a few vendors, says Douglas Parker, a program manager with Tempo Research Corp., a military and aerospace fiber optic service company in Camarillo, Calif. Anyone who has a proprietary claim to some element of the NGCon connector will make the license available to others for a reasonable fee, he says. The Navy wants to be able to control changes to the documentation, enjoy multiple sources of supply, and assure vendor interoperablity and parts interchangeability.
The NGCon connector also is more maintainable. You can remove the alignment sleeve retainer from the connector and easily clean the terminations on each side of the connector. (The alignment sleeve retainer precisely positions the ferrule tips of the corresponding terminations to allow the transmission of light.) In the Mil-Dtl-38999 connector, by contrast, the sleeve stays in one side of the connection. Thus the endface of the fiber termination "is down in a recess, and dirt and dust and fluids can get down into that recess," Beranek says. It's time consuming to clean.
The NGCon connector will have tighter mechanical tolerances than 38999. This will reduce attenuation and increase interoperability. Interoperability is difficult under the current 38999 standard because of its "dimensional spread."
The Defense Logistics Agency has published an NGCon fiber optic termination specification sheet for industry review. NGCon connector spec sheets are expected to go out this month. Suppliers have built some prototype NGCon connectors, but the parts have not yet been qualified. Actual qualified parts could begin to appear by the second quarter of 2007.
An internal Naval Air Systems Command group also is looking at fiber optics built-in test capability. Right now the commercial sector doesn't have a detailed or capable link diagnostic, Beranek says. "We're trying to figure out ways to monitor the health of an aircraft's links. We want to be able to isolate faults if your connector comes loose, your cable breaks or your transmitter dies. Right now the connectors need to be unhooked from the nodes, which takes time. This initiative has not yet moved to a standards group.
SAE also has begun defining an aviation approach to wavelength division multiplexing. The hope is to develop--in coordination with ARINC for commercial aviation--a WDM local area network standard for aerospace. The wavelength division multiplexing technique, used by telecommunications companies in their submarine cable installations, increases the fiber's data throughput by assigning different data streams to different wavelengths, or colors. These can be transmitted simultaneously over a single cable. While the phone companies have multiplexed four to 10 colors, it is theoretically possible to put 228 discrete wavelengths over one optical fiber.
WDM could yield throughput increases of two or three orders of magnitude, predicts Mike Hackert, optical communications specialist with NAVAIR's core avionics group. "Ultimately, we're talking about 10s of terabits/sec of bandwidth--trillions of bits. Different wavelengths of light are used to create separate and distinct channels." he says. And the signals in the different channels can have different modulations and bandwidths.
Rooms of Equipment
The trouble with the telecom approach is that it entails "rooms full of equipment" at either end of the cable, says Hackert. Tunable lasers are also difficult to develop. Part of what's driving the interest in WDM is electro-optical sensors, he says. "[The services] realize these can generate lots of information, which they can convert into a digital format. But they can't put Cray computers on the wing tip or the pod, or wherever the sensor is located."
The telecom industry during its expansion in the late 1990s developed WDM components, but these are designed for a very benign terrestrial environment with no particular space and weight constraints. Aerospace requirements, however, tend to drive towards the highest possible degree of component integration, something the telecom industry is only now starting to do. "It's like where the transistor industry was when ICs [integrated circuits] were first coming out," Hackert says. The standard will set the direction for what capabilities need to be integrated, he concludes.
WDM can help solve the problem of multiple fibers' being overlaid for different applications. "Instead of having `n' fibers, you can have `n' wavelengths running over the same fiber," Hackert says.
In a WDM-based architecture, each application would be allocated its own wavelength and work independently of the others.
While the hope is that WDM will be retrofittable to existing fiber users, NAVAIR is looking ahead, says Hackert, to design a standard for an optimal solution. The standard needs to provide upgradability and interoperability. In other words, the network must allow new applications to be layered over legacy applications without expensive software modifications.
Interoperability, openness and protocol-independence are demanded, as well. The WDM-based network will have a standard interface definition, so that new sensors implementing the interface will be able to plug into and talk over the network. NAVAIR hopes to have a standard supporting a WDM backbone network that enables high-bandwidth, low-cost, lightweight communications with plug-and-play applications.
ARINC is following WDM developments, too, as a possible underpinning of future commercial air transport data networks. The Airlines Electronic Engineering Committee (AEEC) members voted last month to cooperate with other organizations like NAVAIR, SAE and the Institute of Electrical and Electronics Engineers (IEEE) to develop an international WDM architecture/networking standard appropriate for aerospace, according to Dan Martinec, the AEEC's liaison for WDM activity.
Fiber-based system employing WDM could enhance the efficiency of commercial aircraft avionics. For example, a channel of the avionics full-duplex switched Ethernet (AFDX), to be deployed on the Airbus 380, could be accommodated by one of the WDM wavelengths on a fiber while allowing additional wavelengths on the same fiber to be used for other applications, Martinec says. Other candidate applications include displays, electronic flight bags, and the third-generation cabin network, an architecture planned for in-flight entertainment. Hackert predicts a mature draft within three years and technology ready to implement five years from now.
A related area of interest is radio frequency (RF) photonics, which refers to the conversion of analog RF signals into modulations of the light stream. WDM can be used to allow multiple RF channels to flow through a single fiber. While the telecom industry matured digital communications via fiber optics, "less mature analog [optical] communications can provide the greatest immediate benefit for airborne applications," Hackert says.
Heavy, more lossy coax cable currently is used for internal RF signal transmission in avionics architectures. But designers "can almost find a value proposition for a one-to-one replacement of coax for fiber," Hackert says.
In the RF world a modulator would encode the analog signal from a laser light source onto the light stream. The technology has been developed and flight tested, Hackert says. Candidate applications for RF photonics include radar, communications, navigation and electronic warfare.
Fiber Optic Sensors
Fiber optics also can play a role in detecting strain (stretching) and temperature, as part of an aircraft's integrated health management system.
The U.S. Air Force Research Lab (AFRL) is looking at fiber optic sensors in the context of condition-based maintenance (CBM). Under this concept, a structure's health would be determined, based on real-time assessment of its condition rather than by means of scheduled maintenance. "We're trying to put sensors on board, so the component tells you it's OK or not OK, or needs to be repaired," says Mark Derriso, integrated systems health management lead at AFRL's Air Vehicles Directorate.
From in-house experiments, AFRL determined that piezoelectric sensors were more effective than fiber optic sensors at monitoring the status of a bonded structural repair. And piezoelectric sensors have the advantage of being dual-purpose: they can function as a sensor, as well as an actuator. With piezoelectric you can both excite the structure and measure the dynamic response.
But fiber optic sensors perform well in monitoring usage, such as measuring strain. "We could put fiber optics on a vehicle and measure the loads," Derriso says. This information can be included as part of a system's diagnostic algorithm or life-prediction algorithm, he says. Still fiber-based sensors have pros and cons. Multiple sensors can be put in one fiber, but if the fiber breaks in one location, the chain no longer functions.
A total health monitoring solution would include both fiber optic and piezoelectric sensors, Derriso asserts. Such a hybrid approach would provide both life prediction and current state information. Fiber optic sensors also will help in understanding the temperature regime an aircraft has flown through, he says.
Health monitoring will be at the core of futuristic systems such as morphing unmanned air vehicles, Derriso says. In a recent experiment the Air Vehicles Directorate installed fiber optic sensors on a wing-like structure. The sensors are used to determine the changing shape of the structure when a force is applied. Morphing aircraft, still in their infancy, will fatigue much faster than aircraft in normal flight. When the wing of such a vehicle is commanded to change shape, for example, an embedded fiber optic system could verify that such a movement was carried out, Derriso says.
Requirements of WDM LAN Standard