Boeing plans to evaluate GPS-based landing systems in parallel with the FAA.

Graham Warwick/ATLANTA

Boeing is leading an industry programme to validate use of the global-positioning system (GPS) for Category III automatic landings. While the US Federal Aviation Administration intends to demonstrate Cat III GPS, Boeing's global landing-system (GLS) evaluation programme is seen by US industry as setting the pace for the worldwide introduction of satellite-based landing systems.

Unsuccessful in its bid for the FAA programme, Boeing decided to proceed with its own GLS evaluation. While the FAA programme focuses on demonstrating the accuracy and integrity required for Cat III automatic landings, the Boeing programme will go further, developing requirements for a certifiable system.

The Boeing programme "...is more than just a trial", says one contractor. The evaluation is designed to determine the feasibility of GPS-based precision approaches and automatic landings, and to develop requirements for Cat III GLS performance, architecture, installation and certification.

 

FLIGHT TESTING

Airborne and ground equipment supplied by four industry teams will undergo flight and simulation testing. Flight testing will be conducted using NASA's Boeing 757, and will provide the manufacturer with experience in interfacing a GLS with an existing autopilot. Simulations will test GLS responses to faults.

A programme ground-rule is that systems must be installed without any change to the existing automatic-landing system. All four GLSs will be installed and tested simultaneously, although only one will be coupled to the 757 autopilot for each approach, with the other three linked to data recorders. The plan calls for 200-250 landings - 50-75 for each system.

As flight tests will use real GPS signals, hardware-in-loop simulations will look at GLS performance under extreme conditions, including worst-case satellite availability, multi-path and atmospheric conditions and selective-availability effects. Simulations will also be used to examine GLS failure modes involving satellites, the ground-control segment, receiver, ground station and datalink.

The teams involved are: Honeywell/ Pelorus Navigation Systems; Interstate Electronics/Airport Systems International; Litton/Wilcox and Rockwell-Collins/ Daimler-Benz Aerospace (DASA). Contractors will deliver prototype equipment to Boeing in June for installation in the 757. Initial tests will be conducted at Glasgow, Montana, and the evaluation will get under way at NASA's Wallops Island, Virginia, airfield in August, where a laser tracker will be used to monitor approaches.

Teams are supplying airborne and ground equipment at their own cost, with Boeing providing aircraft integration and operations, and NASA furnishing the aircraft, airfield and laser tracker. Boeing will also perform the month-long simulation testing of each system, beginning in January 1996.

The Boeing programme will not be the first Cat III global landing-system evaluation, but promises to be the most comprehensive so far. Previous tests concentrated on demonstrating the accuracy required for Cat III.

In October 1994, 110 automatic landings were conducted with a United Airlines Boeing 737-300 at NASA's Crows Landing airfield in California. These trials used two ground-based pseudo-satellite (pseudolite) integrity beacons developed by Stanford University, and a Trimble Navigation airborne GPS-receiver and DGPS ground-station.

Also in October 1994, 50 automatic landings were conducted with a United Parcel Service (UPS) Boeing 757PF at the FAA's Atlantic City, New Jersey, Technical Center. These trials used Novatel Communications "narrow-correlator" GPS receivers in the aircraft and in the DGPS ground-station, which was developed by Ohio University at Athens.

In the United trial kinematic carrier-phase-tracking GPS (see box) were evaluated, with the pseudolites being used to resolve carrier-phase ambiguity. The UPS trials involved conventional code-tracking GPS (see box), with carrier-phase data being used to reduce multi-path interference.

The FAA's Cat III technology-demonstration programme is a more-formal evaluation, encompassing both accuracy and integrity, and is intended to lead to selection of a GLS architecture by the end of 1995. Under the programme, prime contractors E-Systems and Wilcox are demonstrating different approaches to achieving a Cat III GLS capability.

Wilcox has used a Federal Express Boeing 727-200 for flight-tests at the FAA's Technical Center using Novatel 12-channel GPS receiver modules in the Litton airborne avionics and Wilcox ground-station. E-Systems is using an Israel Aircraft Industries Westwind business-jet for flight tests with Ashtech 12-channel GPS receivers in the air and on the ground.

Wilcox has chosen to demonstrate the total-system error (sensor plus pilot/autopilot) required for Cat III, while E-Systems has elected to demonstrate the GPS positioning-accuracy required. The Wilcox approach is based on the concept of a tunnel in the sky, the boundaries of which are determined by the maximum-allowable navigation errors.

E-Systems is using a kinematic carrier-phase-tracking technique in which both GPS signals, L1 and L2, are used to resolve carrier-phase ambiguity. Wilcox is using a code/phase technique in which L1 carrier-phase data is used to reduce errors in code tracking caused by multi-path signals and receiver noise.

 

Basil techniques

The same basic GLS techniques featured in the FAA demonstration programme will be represented in the Boeing evaluation effort. Three of the teams have chosen to supply code-tracking GLS equipment, while one will provide a carrier-phase-tracking solution.

Interstate Electronics has chosen to supply a carrier-phase GLS, although its work on American Airlines' Project DFW has involved a Boeing 757 equipped with its code-tracking IEC 9001 GPS navigation and landing system (GNLS). The Anaheim, California-based company has flight-tested kinematic carrier-phase-tracking GPS and will incorporate the capability as a software change, says director of GPS marketing Keith Howington.

The system supplied for the Boeing programme will have the same architecture as equipment used by American for its automatic-landing trials: a 12-channel GPS receiver outputting an instrument-landing-system (ILS) emulation direct to the autopilot. The flight-critical approach database - of waypoints defining "pseudo-ILS" glidepaths - will be stored in the GLS itself.

The American trials have used an E-Systems/ARINC DGPS ground-station transmitting differential corrections to the aircraft via the ACARS airline-datalink. Under its joint venture with Interstate to develop GPS landing-systems, Airport Systems will supply a ground station for the Boeing programme. This will use the VHF datalink specified for Special Category 1 (SCAT 1) local-area DGPS installations.

The Collins/DASA team will use the US manufacturer's CAGE GPS engine, under development for its AVSAT series of satellite-based avionics. The CAGE features narrow-correlation technology and carrier code- smoothing to reduce receiver noise and multi-path effects. A 15-channel receiver will be used in the Collins airborne unit and a 20-channel version in the DASA ground station, says manager of advanced systems, Thomas Foster.

For the Boeing trial, the DGPS datalink will be provided by Collins VHF-900 voice/data radios operating at a 31.5kbit/sec data rate on a VOR frequency-band with the D8PSK modulation specified for SCAT 1 DGPS. The aircraft's existing VOR antenna will be used.

In the Collins prototype, typical of the GLSs to be evaluated, the GPS receiver, navigation computer and ILS emulation are in one box, with the datalink receiver and data recorder in separate boxes. The navigation computer is linked to the aircraft's inertial-reference system, because basic satellite availability is not adequate and inertial data is needed to provide continuity of service equivalent to ILS, Foster says.

 

Systems modified

Honeywell is modifying its existing global-navigation-system sensor unit (GNSSU) for the Boeing trial. The GNSSU uses a 12-channel GPS-receiver module supplied by Canadian Marconi. The same receiver will be used in the local-area DGPS ground-station being developed jointly by Honeywell and Calgary, Canada-based Pelorus.

The Honeywell GLS uses smooth code/phase differential GPS, says GPS programme manager Randy Hartman. Receiver software will be modified to add DGPS capability for the Boeing trial, but the hardware will be unchanged, although the company has identified the changes needed for a production Cat III system. Honeywell "...expects some carrier-phase tracking to be involved in the final system", he says.

E-Systems will provide a VHF datalink for the Honeywell demonstration as the latter does not have a suitable unit available or under development, Hartman says. The ground station will be a prototype of the local-area DGPS under development by Honeywell and Pelorus for SCAT 1 certification by the end of 1995.

Litton is supplying a prototype of its latest GPS receiver, the LTN-2001 Mk3, for the Boeing trial. While the LTN-2001 Mk1 and Mk2 now in service in Airbus and Boeing aircraft use an eight-channel receiver, the Mk3 features a 24-channel receiver using Novatel's narrow-correlator code-tracking technology.

The receiver design allows more than one channel to be dedicated to each satellite to remove multi-path errors, says director of advanced programmes Vic Strachan. Narrow-correlator technology efficiently rejects multi-path signals and quietens receiver noise, resulting in an accuracy "on the order of 1m", he says. The advantage of code-only GPS, Strachan says, is that it avoids the need for additional satellites in view or ground-based pseudolites to resolve carrier-phase ambiguities.

The same Novatel receiver technology will be used in the Wilcox-supplied DGPS ground-station. Litton and Wilcox form the only team to be involved in both the Boeing and FAA Cat III GLS programmes. Novatel's receiver was used by Wilcox in the NASA trials which first demonstrated Cat III GPS accuracy in 1993.

 

Emerging market

The emerging GLS market represents an opportunity for new companies. Honeywell and Litton dominate the airline GPS-receiver market, with Collins now making a belated entry. Interstate is an established military supplier working to transfer its GPS technology into the commercial sector.

Wilcox, a subsidiary of France's Thomson-CSF, is an established landing-systems supplier, but ground-station tie-ups mean new-business prospects for DASA, Pelorus and Airport Systems. Other contenders are likely to emerge. France's Sextant Avionique is involved in Airbus GLS trials, and E-Systems, Harris and Raytheon are expected to complete for satellite landing-system business. Honeywell/Pelorus already has contracts for local-area DGPS systems in Australia, Canada and the USA. Interstate and Airport Systems are involved in a DGPS trial with Swiss regional airline Crossair. The move to global landing systems has already begun.

 

USA 's three routes to CAT III

While the route to Category I and II landing capability using GPS is now well-established, questions remain over how to achieve Cat III, with its increased demands for accuracy, availability and integrity. Three approaches to Cat III GLS are being evaluated by the FAA: conventional code-tracked GPS; highly accurate carrier-phase-tracked GPS; and GPS augmented by ground-based pseudo-satellites (pseudolites).

Conventional GPS generally does not have the accuracy required for Cat III, but manufacturers are working on techniques to improve the performance of code-tracking receivers. These include using phase data from the GPS L1 carrier-frequency to improve code tracking by reducing receiver noise and multi-path effects, the major barriers to achieving Cat III accuracy.

Because of the limitations of code-only GPS, some manufacturers favour more accurate, but more complex, kinematic carrier-phase tracking, where position is determined by measuring the number of GPS-signal wavelengths, or cycles, between the satellite and the receiver.

The 1.5GHz signal carrying the GPS code has a wavelength of 19cm, but a carrier-phase receiver can, in ideal conditions, determine its position to within a faction of a wavelength, providing even greater accuracy. The drawback is that, to select the correct cycle within the signal, the receiver needs several satellites in view, with good viewing geometry.

Carrier-phase-tracking GPS is routinely used for surveying, where the consequences of selecting the wrong cycle are trivial. If an airborne receiver picks the wrong wavelength the result could be disastrous. The difficult task of resolving carrier-phase ambiguity "on the fly" can be aided by using the DGPS ground station to transmit phase information along with differential corrections.

Integrity is an issue with carrier-phase GPS because of its sensitivity to satellite geometry and the time required to resolve the wavelength ambiguity. If contact is lost with a satellite during an approach, the time taken to re-acquire the signal and resolve the phase ambiguity could be outside that allowed for continued Cat III operation, opponents argue.

The third GLS approach is an effort to increase the integrity of carrier-phase GPS. In the pseudolite concept, low-power GPS "marker beacons" are located under the approach path. As the aircraft flies over these transmitters, "satellite" geometry is improved and Doppler shift in the ground-based GPS signal is used to resolve carrier-phase ambiguities.

Proponents of pseudolites argue that these low-cost, battery-powered, devices improve GPS accuracy to centimetre-level while eliminating concerns about satellite failure during the approach. The drawback is that the technique is more complex than code-only or carrier-phase GPS.

 

 

Airbus promotes GPS use

Airbus Industrie is aggressively promoting use of GPS on its aircraft, beginning with the first deliveries of GPS-equipped A330s and A340s in late 1993. The GPS is approved as a supplemental means of navigation on the A320/A321 and A330/A340. The European consortium plans worldwide primary-means certification on its entire range of aircraft by mid-1996.

Primary-means approval will allow non-precision approaches to a 200ft (60m) decision height using the GPS. Certification is planned on the A300-600/A310 and A330/A340 by the end of 1995, and on the A319/A320/A321 by mid-1996. It will cover both Honeywell and Litton GPS receivers.

The keys to primary-means approval are Honeywell's receiver-autonomous integrity monitoring (RAIM) and Litton's autonomous integrity-monitored extrapolation (AIME). The Honeywell-developed RAIM is widely used by GPS manufacturers to enable receivers to detect and reject faulty satellites, but AIME is unique to Litton.

AIME involves the close integration of the Litton LTN-2001 GPS receiver and LTN-101 inertial-reference system (IRS). GPS data are used to trim the IRS to the point where only one satellite is needed to provide the integrity required for a non-precision approach, explains Litton Aero Products' director of advanced programmes, Vic Strachan.

RAIM limitations result in GPS not being usable for non-precision approaches "a lot of the time", he argues, adding that Litton is the only company close to providing 24h, worldwide, NPA capability with "now" GPS - in order words without any form of augmentation.

Airbus says that AIME provides 99.9% availability, compared with about 90% for RAIM. Litton says the integrity-protection limit provided by RAIM varies with satellite geometry (which, in turn, varies with time and location), while that provided by AIME is relatively constant and matches RAIM at its best.

Airbus, meanwhile, has begun DGPS precision-approach work, conducting automatic-landing tests in October 1994 using an A340 equipped with a Sextant Avionique GPS receiver. Accuracy achieved was equivalent to that of a Cat II instrument landing-system (ILS), the manufacturer says.

Testing continues with Sextant and Litton airborne receivers and Thomson-CSF and Wilcox ground-stations. The manufacturer's objective is to achieve Cat I certification of a GLS as part of a multi-mode receiver (MMR). The new MMR will combine an ILS with optional global and microwave landing-systems. Certification is planned for late 1997/early 1998.

ARINC specifications for the MMR are expected to be finalised later this year, and Airbus plans to flight test an experimental receiver in 1996 on an A320 or A340. The manufacturer expects the initial MMR to combine a Cat III ILS with a Cat I GLS, with later growth to a Cat III GLS providing ILS-like outputs to the aircraft's autopilot.

 

Source: Flight International