Leica GeosysTelephone +1

Worldwide Beacon DGPSStatus and Operational Issues

Gunnar Mangs, Leica Geosystems Inc.

Satish K. Mittal, Leica Geosystems Inc.

Thomas A. Stansell, Jr., Stansell Consulting

Leica Geosystems, Torrance

May, 1999

tems Inc., 23868 Hawthorne Blvd, Torrance CA 90505, USA 310 791 5300, Fax +1 310 791 6108

1

Worldwide Beacon DGPSStatus and Operational Issues

Gunnar Mangs, Leica Geosystems Inc.Satish K. Mittal, Leica Geosystems Inc.

Thomas A. Stansell, Jr., Stansell Consulting

1. BiographiesGunnar Mangs is Director of the worldwide Marineand Reference Station Business for Leica GeosystemsInc. in Torrance, California. He received his M.Sc. inPhysics in 1970 from the Stockholm TechnicalUniversity. From 1987 to 1997, Gunnar wasPresident and CEO of C A Clase in Sweden, acompany which distributes navigation products andsystems to the commercial and military marinemarkets throughout Scandinavia and the UnitedKingdom. In partnership with Magnavox, C A Clasein 1991 installed the first Beacon DGPS Systems inthe Baltic Sea. Gunnar also was the first chairman ofthe Scandinavian GNSS Industry Council.

Satish Mittal is the Product Manager for DifferentialGPS Systems, including DGPS Beacon Systems, atLeica Geosystems Inc. in Torrance, California.Satish received his BEE degree in 1974 and MEE in1976 from Banaras Hindu University in India. Satishreceived an additional MEE in 1986 from theUniversity of Toronto in Canada. He has more than15 years experience in various disciplines of marineelectronics, including the development ofsophisticated, large scale, integrated navigationsystems and DGPS systems.

Tom Stansell now heads Stansell Consulting.Previously he was a Vice President of LeicaGeosystems in Torrance, California, where he wasinvolved in technology development and strategicrelationships. Tom received his BEE degree in 1957and his MEE degree in 1964, both from theUniversity of Virginia. At the Johns HopkinsApplied Physics Laboratory he participated indevelopment of the Transit Satellite NavigationSystem. At Magnavox, he led the development ofmany Transit and GPS products and their underlyingtechnologies, as well as initial development of theMagnavox Beacon DGPS systems and user receiversfirst installed in Sweden and Finland during 1991.He is the author of many technical papers, receivedthe ION Weems Award in 1996 for continuingcontributions to the art and science of navigation, andhe was elected a Fellow of the ION in 1999.

2. IntroductionThis paper was written at the request of Bill Adams,president of the RTCM. Bill asked Leica to look atBeacon Differential GPS (Beacon DGPS) from anoverall and integrated perspective, rather than themore typical approach of dealing with a specificinfrastructure design, a particular user equipment, orone nation’s deployment status. We agreed becauseLeica has the perspective gained from being asignificant supplier of both the infrastructureequipment, which provides and monitors the DGPSsignals, and the user equipment for a wide range ofmarine applications. In addition, Leica has animportant historical perspective, having been the firstcommercial company to develop and sell BeaconDGPS infrastructure equipment and among the firstto develop and sell beacon DGPS receivers andnavigation equipment.

The paper begins with three sections which provideimportant background material and overallperspective. The first of these shows how the BeaconDGPS service evolved, not as a single system but as aconfederation of many systems. The reasons behindcommon signal standards are explained and praised,but the confusion resulting from differentdevelopment paths, such as inconsistent definitionsand terminology, are noted. The second of thebackground sections describes the extensive presentand evolving coverage provided by Beacon DGPStransmitters. The third of these sections presents thekey results from an international Leica survey ofcommercial users, knowledgeable sales agents, andexperienced service representatives. Some of theresults are quite surprising and already haveinfluenced Leica’s thinking about user equipment.Other insights should influence continueddevelopment of the worldwide Beacon DGPSservice.

Following the three explanatory sections is a sectionwhich focuses on eight specific problems andsuggests possible solutions for each. The titles of theeight subsections speak for themselves and are:

2

• Improve International Coordination andUser Communication

• Provide Simple and Meaningful StatusIndications

• Employ Higher Quality User Equipment• Return to More Intelligent Beacon Signal

Selection• Improve Bridge Integration• Reduce Inter-Station Interference• Defend Against Natural and Manmade

Interference• Encourage and Prepare for an SA-Free

Environment

The paper closes with a brief summary of theinformation presented and a review of theconclusions reached.

3. The Beacon DGPS DevelopmentPathIt is informative to compare development of theBeacon DGPS service with development of theGlobal Positioning System. In the case of GPS, asingle organization, the Joint Program Office (JPO),was responsible for defining, developing, and testingevery element of the initial GPS. This overallresponsibility included the satellites, the groundcontrol system, and many types of military userequipment. The development cost and the ongoingoperational costs are paid by one nation. Sincedeployment, the U.S. Coast Guard Navigation Center(www.navcen.uscg.mil) has been designated as thepoint of contact and the official source of informationfor all non-military GPS users.

The Beacon DGPS development path was remarkablydifferent. Rather than starting with a grand plandirected by one organization, it has evolved andgrown with cooperation and input from many diverseorganizations. It began as an experiment by the U.S.Coast Guard (USCG). At about the same time,several organizations realized that SelectiveAvailability (SA) made differential GPS correctionsnecessary and that a standardized DGPS messagewould be desirable. As a result, the RTCMsponsored Special Committee 104 for this purpose.Participation from both government and industry wasenergetic and widespread. As a result, GPS receiversfrom almost every manufacturer are interoperabletoday due to the SC-104 standards. Because theUSCG was a major participant in the work of SC-104, the Beacon DGPS data format both followedand helped drive the standard. The first operationaldeployment of Beacon DGPS transmitters in the U.S.

was to reduce the cost of buoy tending and improveplacement accuracy, not as an aid to navigation.

Other nations also began to realize the value of GPSand the need for DGPS, not only to improve accuracybut also to maintain control of their own coastalnavigation signals. (It is troubling for a nation’s localtraffic to depend on another country’s militarynavigation system, especially when accuracy isintentionally degraded and could be made worse atany time.) The Beacon DGPS concept wasparticularly attractive because Radio Direction Finder(RDF) beacon stations were well established, havingbeen first deployed shortly after World War I(~1920). The land, the facilities, and, moreimportant, the approved frequency band were allavailable. Furthermore, organizations were in placeto maintain the beacons, decades of use haddemonstrated the excellent coverage they provided,and their current utility was almost nonexistentbecause of newer navigation technology. Leica (thenMagnavox) was the first company to develop and sella complete suite of equipment which enabled existingRDF beacons to transmit RTCM DGPS messages.These first commercial systems were installed inSweden and Finland in 1991.

Since then, the number of Beacon DGPS stations hasgrown tremendously. According to data from theWEB site of Communication Systems InternationalInc. (CSI of Calgary, Canada, www.csi-dgps.com), atthis time 34 countries have installed Beacon DGPSstations, with 189 stations listed as on-line, 37 on-linewithout integrity monitoring, and 10 on-line in a testmode, totaling 236 on the air. Another 76 are listedas planned or under construction, giving a potentialtotal of 312.

Each participating country follows its own path inestablishing a Beacon DGPS service. The USCGinternally developed its own system and software,while using commercially available components suchas transmitters and GPS receivers. Other nationshave solicited proposals and purchased a completesystem from one of several commercial suppliers. Avariation of this is for a nation to prepare a detailedspecification which must be met by one of thecommercial suppliers. Australia is a recent exampleof this approach, as described in Reference [1]. Atleast two nations initially chose to encrypt the beaconsignals in order to collect revenue for their use.Fortunately this decision has been reversed and thesignals are now freely available, as with all otherparticipating nations. Nowadays the manufacturersare required to comply with the RTCM RSIMstandard (issued in 1996) for the communication

3

between the various elements within the beaconreference station. This helps to give current systemsequal operational characteristics. Some of the earliersystems may however not be compliant with theRSIM standard.

Integrity control has taken at least two paths. TheU.S. system is based on command and control. Anyproblem sensed at a reference station iscommunicated to one of two control centers wheredecisions are made and control messages are sentback to the reference station. In contrast, theAustralian reference station software is designed notonly to report to the control center but also to reactautonomously to a problem in order to maintainstation integrity even if communication with thecontrol center is not available.

The preceding paragraphs highlight the differencesbetween a system such as GPS, the development ofwhich was centrally managed from signal source touser equipment, with development, operation, andmaintenance paid for by a single government,compared with the Beacon DGPS service which hasspread around the world in a far less formal andstructured way. The cost and the responsibility hasbeen shared, and the fact that the same set ofnavigation equipment can be used anywhere in theworld that signals are available is a testament toeffective international cooperation. The termConfederated System comes to mind when describingthe Beacon DGPS service

There are, of course, some difficulties with having aConfederated System rather than a centralizedsystem. Some of these will be noted in the section onProblems and Potential Solutions, below.

4. Beacon CoverageBeacon DGPS coverage has grown dramaticallyduring this decade. Figure 1 shows the coverage ofstations declared by their national authorities to befully operational, interim operational, or simplyon-line. The difference between full and interimoperational is driven primarily by on-air availabilitystatistics, not by accuracy or integrity issues. Someof the on-line stations are unmonitored. However,the stations are considered trustworthy and are indaily use. (Note that these three figures have beensupplied courtesy of Communication SystemsInternational Inc., which in turn obtained stationlocations and status information from the nationalauthorities.)

We have not shown those few stations which transmitencrypted signals in order to receive revenue for theiruse. In fact, we regret that any nation permitsencryption of signals in this frequency band forrevenue purposes. The band has been allocated andused for open radionavigation, without tolls, formany decades.

In addition, a few stations in this band have beeninstalled by private companies for their own use, andthe signals are not encrypted. Scanning beaconreceivers will find and use these signals. Our concernis whether they are sanctioned by a nation member ofthe confederation and whether the reference surveycoordinates and the signal standards meet accuracy,safety, and integrity requirements. This internationalradionavigation band should not be used for privatecommercial interests, at least without beingsanctioned and required to provide safe public access.

Figure 2 adds stations which are on the air but are ina test mode. These also are being used by marinersand should be declared operational in the near future.Adding the stations in test mode increases coverageprimarily in Egypt and around India. Figure 3 addsthe pending stations, which are being installed but arenot transmitting. As you can see, the entire U.S. landmass, including Alaska, will be covered. Growthalso occurs in Spain, including the Canary Islands,Russia, Bangladesh, China, and Australia. Todistinguish between pending and operating stations,the pending stations are shown with a black centerdot and the operating stations with a white center dot.This is not the end of growth. Other national systemsare being planned and will be installed in the comingmonths.

In a relatively short time, the world has obtained animportant new infrastructure which promotes safety,protects the environment, supports commerce, andeven aids pleasure activities on every continent. Thechallenge ahead is to consolidate and expand thisgrowth, to improve system performance, and to adaptto a major change on the horizon.

4

Figure 1 – May 1999 Coverage of Stations Declared Operational, Interim Operational, or On-Line(Courtesy of Communication Systems International Inc.)

Figure 2 – May 1999 Coverage of On-Air Stations, Including Those in a Test Mode(Courtesy of Communication Systems International Inc.)

5

Figure 3 – Anticipated Coverage of Stations Whether Operating or Pending Operation(Courtesy of Communication Systems International Inc.)

5. User ExperienceLeica conducted a survey among commercial users ofthe Beacon DGPS service in various parts of theworld. The most experienced agents and servicerepresentatives also were interviewed. The primepurpose was to see if users would identify any majoraspects of the service that should be modified inorder to extend and improve this use of GPS fornavigation. Below is a summary of the surveyresults.

5.1 General View of the Service ProvidedMost commercial users are satisfied with thereliability and accuracy of the Beacon DGPS service.Over 90% of the people interviewed expressed theirsatisfaction with the service and use it whenever theycan receive the signal. It is our opinion that manyusers in the countries that first applied thistechnology regard it as a commodity. There is anexpressed requirement for additional coverage innations that more recently started to transmit BeaconDGPS signals.

The main advantages seem to be:• Stabilization of the ECDIS display. This is

essential, especially when navigating inrestricted waters.

• Accurate speed input to the autopilot. In pilotedwaters this is in many cases the prime (or eventhe only) speed sensor used. Particularly foricebreakers, this is the only reliable speed sensor.

• Stable speed and heading, which racing yachtsfind the essential.

• Higher position accuracy, which is particularlyvaluable for fishing because it permits the fishinggear to be placed closer to the desiredobstructions.

Those who were skeptical about the value of DGPSexpressed concerns regarding:• The value of an accurate position solution is

undermined by the poor accuracy of ElectronicCharts.

• There are inadequate accuracy or qualityindications regarding the DGPS solution.

6

• Apparently there are problems in safelyintegrating DGPS into an automatic navigationsystem.

• It is a problem that control of the system is inforeign (i.e., American) hands and that accuracyis denied.

5.2 Transmission QualityThe move to RTCM Type 9 messages and in somecountries to a 200 Baud data rate make the systemquite robust. The range of the stations is consideredsufficient in most cases.

5.3 InterferenceThe influence of “bad weather” as well as man-madenoise, including own ship’s noise, are the maincauses for interruption of the signal. The realexplanation for this “bad weather effect” is somewhatunclear. Precipitation static, which used to causeproblems in high latitudes, is now considered to havebeen cured by the use of H-Field antennas.

5.4 System IntegrityThere is limited understanding of the systemoperational status. Normally the user looks for theDGPS symbol on his GPS navigator and then judgesthe “stability” of his position display to evaluateaccuracy.

Knowledge about and use of the Type 16 message islimited. In fact, most user equipment does notdisplay this information. Even worse, many receiversdo not display the Station Health Status (healthy andmonitored, unmonitored, or unhealthy) provided ineach message. Some authorities send out frequentwarning messages if the system has not been declaredoperational, and most users do not know how torelate to this information (if their equipment iscapable of displaying it). Most use the correctionsignal anyway and disregard the warning messages.

An alternative DGPS system is used in some areasfor backup. This can be either an RDS FM sidebandservice (e.g. the Swedish EPOS) which is operated bya government authority, a satellite differentialservice, or a private reference station owned by theshipping company itself. Switching to these systemsis done manually and is executed if the beaconsystem fails to deliver usable corrections. Theseusers do recognize that these back-up systems areunmonitored and in some cases have an uncontrolledlatency. The additional cost for such back-upsystems is perceived as a major problem.

5.5 Switching between beacon stationsThe common method used for station selection iseither manual tuning or automatic scanning of thefrequency band and selecting the strongest signal. Itis a well known problem among experienced usersthat the beacon receivers with automatic scanningand switching can lock to a sky wave from a distantstation and ignore a more suitable station thatproduces a weaker but fully usable signal. This mayforce the user to manually select the station andthereby make the operation more cumbersome.

5.6 Shipboard EquipmentThe operators today view DGPS equipment as acommodity. Few understand the operational detailsand quality of the various components. There is anexpressed need for some “position quality indication”and for relevant alarms.

Some users have raised aspects of the integration ofthe DGPS sensor to the ship’s navigation system.Many use the DGPS speed as the only speed input tothe ECDIS and navigation systems and set the GPSreceiver to DGPS mode (forced differential). Theloss of GPS corrections has in some cases caused theship to completely lose automatic steering, since noautomatic backup was provided.

6. Problems and Potential Solutions

6.1 Improve International Coordination andUser CommunicationAs described in the Development Path section above,the worldwide Beacon DGPS service is aconfederation of independent national systems ratherthan one centrally controlled system. Each countryseparately acquires, installs, and operates its ownsystem. A key indication of this is just how difficultit is to determine the status of all existing or plannedBeacon DGPS stations in all 34 countries. Somenations provide a WEB site for their own stations,e.g., www.navcen.uscg.mil for the U.S.,www.amsa.gov.au for Australia,www.trinityhouse.co.uk/dgps.htm for Great Britain,www.fma.fi/radionavigation for Finland, orwww.ccg-gcc.gc.ca/tosd-dsto/awtj/dgps/table.htm forCanada. (A brief WEB search on Yahoo for BeaconDGPS found companies which supply userequipment, but none of these government resources.)It is significant that the best overall source is theWEB site of a commercial organization (CSI ofCalgary, Canada, www.csi-dgps.com).

7

An additional problem with this confederated systemis that there are many different ways to describe thestatus of a beacon station, which makes it verydifficult for a mariner to know what the status reallymeans. For example, the CSI status descriptionsinclude: On-line, On-line no Auto IntegrityMonitoring; On-line Test Mode; No Service, and NoIntegrity Monitoring. The following information wasrecently obtained from the national WEB sites. “OnMarch 15, 1999, the USCG announced that the U.S.system had achieved Full Operational Capability(FOC)”. Prior to that date, the status was InterimOperational Capability (IOC). The status of moststations is simply listed as Operational. There is noindication as to how that does or should affect themariner. The status of the Canadian stations for themost part is Initial Operational Service (IOS). TheAustralian status is Partially Operational, withindividual stations listed as on-air or on-air-testing.The Finnish WEB site simply lists stations which aremaintained by the Finnish Maritime Administration,with no other status indication.

There is a clear need for: (a) standardization of statusterminology, (b) a clear definition of how the marinershould treat the status information (what operationaldifference should it make?), and (c) for a combined,coordinated, standardized, and frequently updated listof all Beacon DGPS stations and their status.

It would seem that the logical organization to fulfillthis role is the International Association ofLighthouse Authorities (IALA, which also is knownas Association Internationale de SignalisationMaritime, or AISM.). Quoting from their new WEBsite (www.iala-aism.org), “Today IALA comprises200 members, 80 of which are national authoritiesand 60 are commercial firms.” IALA publishesBeacon DGPS standards, and IALA meetings (thenext one of which is during 9 – 19 July, 1999, inHamburg, Germany) are used to discuss andcoordinate Beacon DGPS issues. We urge IALA tovolunteer for this new task, which would involve: (a)reaching international agreement on a standardizedstatus reporting format, (b) collecting statusinformation very frequently from national authorities,and (c) making the status information available in atimely way on its WEB site. All participatingnational authorities would be urged to mirror or linkto this information and to make it readily available tomariners by standard communication methods. Allparticipating commercial firms would be urged toinform their customers about how to obtain thisinformation.

Equally important, we urge IALA to create adocument which informs mariners about the meaningof status messages, including the Type 16 messagesbroadcast by the beacons. The main purpose wouldbe to suggest (not mandate) changes in navigationalprocedure based on the status information. Thisdocument should be disseminated by as manymethods as possible. We believe an importantsecondary objective of this task will be to causenational authorities to collaborate in re-thinking theoperational implications of status and warningmessages from the mariner’s perspective, whichcould result in an improved message strategy.

Finally, we believe it would be useful to increase therate and the amount of communication between theconfederation of signal providers and their worldwideusers. Establishing a point of contact, such as theUSCG Navigation Center is for GPS and for the U.S.Loran-C and Beacon DGPS services, should beconsidered. Part of the objective would be todocument user inputs, suggestions, and complaintsand to make them available to all users and signalproviders. Service status changes would be sharedwith all. National organizations presumably wouldlink their users to this central service and forwarduser inputs. The Confederation also could use aprivate part of this medium to exchange insights andeven carry on debates about standards. Continuousand rapid communication should strengthen theConfederation, help raise standards, improve service,and thus benefit all users.

6.2 Provide Simple and Meaningful StatusIndicationsThe current “health status” of each station is providedin its transmitted RTCM messages. There are threerelevant states: (a) monitored, (b) unmonitored, and(c) not working or out of specification. States (a) and(c) are unambiguous. The signals are known to bereliable and accurate or they are known to be bad.Unfortunately, ambiguity is inherent in state (b); thesignals probably are good but there is no independentinformation from an integrity monitor to verify thecondition. In spite of the ambiguity, even worse isthe fact that many DGPS receivers do not display thehealth status conditions at all.

One of the most significant findings from our surveyof ship captains and other well informed mariners isthat no one knows what to do about most, if not all,warning messages. What should he do knowing thata station is unmonitored? What should he do if theGPS receiver begins to warn that the station is nolonger monitored? Some Type 16 messagescontinually repeat a warning, such as being in test

8

status, with no problem being evident to the mariner.Therefore, the common response is simply to ignoreall warnings! As a result, the entire purpose ofwarning the mariner, other than to limit the liabilityof the provider, has been lost. Today’s warningprocedures are less than ideal, and they could bedangerous by promoting complacency.

Leica has realized that equipment suppliers can domore to aid the mariner. The new techniques may becontroversial at first, and they must be subjected toextensive review, field testing, and verification, butsome of the directions are clear. Our user surveyrevealed that most mariners now consider BeaconDGPS equipment to be a commodity that eitherworks properly or else it must be broken. Thesubtleties of propagation issues, natural interference,warning messages, and the like, are not appreciatedand therefore are ignored. We believe the marinersare right and that equipment suppliers should be ableto reduce these variables to a simple Green, Amber,or Red status indication.

We believe it is possible to provide shipboardintegrity checking. Even if the received DGPSmessages are not being monitored for integrity, theshipboard navigation solution with a high qualityGPS receiver can determine if the DGPS correctionsare precise relative to the defined reference stationantenna location. In effect, the shipboard equipmentcan be its own integrity monitor. Even with warningsof an unmonitored station, the navigation equipmentitself could give the mariner a green light. (Theliability problem inherent in such a clear andmeaningful display must be resolved.)

Following this approach, the only problem whichcould lead to a navigation error is for the DGPSmessage to report the wrong position of the referencestation antenna. There are at least three ways for theshipboard equipment to detect such an error. First,with over 200 stations on the air, growing soon toover 300, the opportunity to receive multiple BeaconDGPS signals is substantial, especially because ofskywave propagation. Even if the second signal issubject to many dropouts, a few good messages fromanother location can be used to verify that theprimary message source does not have a substantialantenna position error. The second technique is thesame idea, but using other sources of DGPScorrections, such as a Vessel Traffic System (VTS) orDGPS from an FM sideband system. The thirdmethod is to compare the DGPS position with othermeans of navigation, such as a radar map display.(The DGPS navigation equipment could ask for thisconfirmation before giving a green light.)

In addition to the above, shipboard DGPS equipmentshould be designed to use more than one source ofDGPS corrections. For example, if two or morebeacon signals are continuously processed, the loss ofone signal will not cause even a momentarynavigation outage. Also, by using two or moremarginal DGPS signals, excellent navigation resultsstill could be obtained. Other sources of DGPScorrections also can be used, as mentioned above.Today this is done by manual intervention, switchingfrom one source to another. The mariner has a rightto expect this switching and/or combining of DGPSsources to be performed automatically. He shouldbecome involved only if there is a problem.

We believe DGPS navigation equipment can bedesigned to provide a universal and unambiguousstatus display. This will simplify operation, but moreimportantly it will avoid the general tendency ofmariners to discount and thus ignore the warningsthey receive today.

A key problem, however, is the tendency of shipowners to purchase the least expensive equipmentwhich satisfies minimum requirements. This is forthe sake of economy and because most ship ownersare not mariners. The only solution is to raise theregulatory standards, which is a far more difficulttask than designing better equipment.

6.3 Employ Higher Quality User EquipmentTests have demonstrated that there is a wide range ofuser equipment performance. Reference [1] notes anAustralian Beacon DGPS test in which oneequipment gave satisfactory results while theperformance of another, at the same location, wasunacceptable. Unfortunately, both of these probablymet minimum carriage requirements.

In addition, although we noted above that marinersare ignoring the warning messages, that does notexcuse the fact that some DGPS navigationequipment will not even display the broadcast Type16 messages. The use of unsophisticated navigationequipment aboard large vessels is the result ofinadequate regulation and the poor level ofunderstanding, or caring, by ship owners.

6.4 Return to More Intelligent Beacon SignalSelectionLeica (then Magnavox) originally set the standard forBeacon DGPS receivers. It gave the responsibilityfor tuning the beacon receiver to the GPS receiver.The idea was to take full advantage of the computerin the GPS receiver, as well as its knowledge of

9

position, to determine and then select the best beaconsignal for that area. This concept is why the RTCMSC-104 format includes message Type 7, which is analmanac giving the location and frequency of othernearby beacon signals. By storing almanac messagesfrom many stations, the GPS receiver can build aninternal almanac of all stations it has been near.During a voyage, the GPS receiver, knowing itsposition, commands the beacon receiver to tune to thenearest and/or the best signal in that area.

Unfortunately, the original idea was nearly sweptaside by the availability of less expensive one andtwo channel beacon receivers which, on their own,simply find the strongest beacon signal. Thisrelieved the GPS receiver of having to monitor andcontrol these decisions. The result was cheaper, andit seemed to work well.

However, as the number of Beacon DGPStransmitters has increased and as operationalexperience has grown, a problem has emerged. Thestrongest signal may be a skywave from a moredistant transmitter. Or, something might be wrongwith the strongest signal when a weaker but entirelysatisfactory signal is available.

Although the less sophisticated approach may beadequate for low cost consumer products, we believelarge ships should be equipped with a moresophisticated signal selection process. Too oftentoday’s mariner feels that he must manually select thebest beacon signal because automatic selection bysignal strength alone is not reliable. This task isbetter done by the GPS computer for several reasons:(a) it will not forget to make the next selection, (b) itcan be programmed to recognize signal problems andmake a better selection or selections, and (c) not allmariners are adept at or willing to make appropriatemanual selections.

A related issue is that some beacons do not transmitthe Type 7 almanac message. With a return to moreintelligent signal selection, we strongly recommendthat every beacon should provide Type 7 messages.

6.5 Improve Bridge IntegrationA key issue from the user survey derives from theextent to which DGPS has become the primarysource of speed (velocity) input to the autopilot, theradar plotter, and the ECDIS or ECS. This is anatural evolution because DGPS is the most accuratesource of velocity as well as of position. However,many integrated bridge systems apparently do notautomatically switch to a backup speed sensor whenthe DGPS output is interrupted. Perhaps worse, in

many cases the GPS receiver is set to a DifferentialOnly mode, meaning that it will provide a navigationsolution only if the GPS measurements are beingcorrected. If local or manmade noise or otherproblems interrupt the beacon signal, the DGPSspeed measurement to vital systems such as theautopilot suddenly drops to zero. We understand thatat least one accident was caused or abetted by such anevent.

A subtler version of this issue is the question of howGPS receivers derive velocity. Particular issues arehow much filtering is applied and how long thereceiver will continue to extrapolate position andvelocity if the GPS signals are interrupted. One GPSreceiver model was infamous for continuing to showa straight path well after the ship had begun its turn.

On one hand, the GPS receiver should be tunedproperly for the application. Equally important, thoseresponsible for integrating navigation sensors withvital control systems must understand the individualsensor characteristics and interactions and makeprovision for automatic backup if something fails,even briefly. Regarding the speed or velocity input,an integrated bridge system should make the best useof DGPS navigation when it is available but switchimmediately to autonomous GPS or to a speed logwhen it isn’t. Tuning the system to make use of aless accurate or biased input and warning the marinerof the change should occur without hesitation.

For the sake of safety, companies supplying sensorsand companies combining them into systems shouldcommunicate more effectively and more thoroughlywith each other. Regulatory organizations shouldunderstand these subtle interactions and put pressureon companies to supply and on ship owners to buybetter and safer integrated bridge systems.

6.6 Reduce Inter-Station InterferenceWe think that in the early days of Beacon DGPSthere was little concern about inter-stationinterference. Because of the limited number offrequency slots and because of the long distance askywave signal can travel, this problem is becomingmuch more common.

There are at least three things which can be done.IALA and perhaps other organizations are engaged inpredicting these interactions and beginning to allocateand even change frequency assignments to minimizethe problem. We strongly endorse and encouragethese activities.

10

The second step was mentioned above. By buildingsmarter DGPS navigation equipment, especially withmore beacon receiver channels, the GPS receiver canselect only the best corrections or combinecorrections from multiple sources to provide reliablenavigation results. Use of beacon receivers whichautomatically select the strongest signal is noteffective.

We are not experts in this area, but we suspect that abetter job can be done of employing the availablebandwidth. For example, many stations continue toprovide separate signals for DGPS corrections andfor direction finding, separated by 500 Hz. Thepractical need for marine radio direction finder(RDF) beacons in this age of GPS, Loran-C, andother navigation aids does not exist. The RDFsignals should be discontinued and the transmitterpower for these signals should be used to increase theDGPS signal power. By opening the RDFfrequencies for DGPS signals, the inter-stationinterference problem would be reduced.

Other nations, such as the U.S., space the beaconsignals a full kilohertz apart. By spacing the signalsat 500 Hz intervals, the number of available channelswould be doubled. It may be true that 500 Hzspacing is not desirable when using the preferred200 Hz data modulation, although it does work.However, we must anticipate the removal ofSelective Availability (SA), in which case a muchlower data rate will support today’s accuracy. Somedream of then providing two (or three) frequencyphase corrections to support decimeter or centimeternavigation with the 200 Hz data rate capacity, whichis a valid objective. Another approach would be tomove to 100 Hz, or even 50 Hz, data rates in order tomake space for many more non-interfering channels.It is past time to begin evaluating and debating theoptions possible when SA is removed.

6.7 Defend Against Natural and ManmadeInterferenceBecause the beacon band is at such a low frequency,it is quite subject to natural as well as to manmadeinterference. Much progress against precipitationstatic has been made in the transition from E-Field toH-Field antennas. Even so, lightning storms andshipboard machinery can disrupt the service. It is atleast ironic that accurate navigation may be most vitalduring stormy conditions.

There are only two defenses against lightning storminterference. One is to increase transmitter outputpower. For stations which transmit both an RDF anda DGPS signal, this could be done by discontinuing

the RDF signal and diverting its power to the DGPSsignal. The U.S. Government intends to cover all ofthe U.S. land mass with overlapping Beacon DGPScoverage. This is termed NDGPS, for NationwideDGPS. The Air Force GWEN (Ground WaveEmergency Network) transmitters and antennas arebeing converted for this purpose. Because theantenna height and the ground plane radius is abouttwice that of the antenna installations used for coastalcoverage, the efficiency is much better and thereforethe radiated power from these sites is substantiallyhigher. A bigger antenna is one way to transmit morepower.

The second defense against storm interference isextremely effective, costs nothing, and provides otherimportant benefits. This important safetyenhancement is to discontinue Selective Availability,as discussed more fully below.

Manmade interference comes largely from arcing orsparks caused by electrical machinery or combustionengines. One difficult aspect of this type ofinterference aboard ship is that it can change throughaging of existing equipment or by installation of newequipment. Fortunately, much of this equipment maynot be in use when the ship is underway. However,this type of source can be located with simple orimprovised test equipment (such as a portable AMradio), and there are ways to eliminate or to shieldthese sources. Mariners would benefit from a betterunderstanding of noise sources, how to find them,and how to minimize or eliminate them.

6.8 Encourage and Prepare for an SA-FreeEnvironmentAs you know, Selective Availability (SA) is theprocess whereby the U.S. Department of Defense(DoD) reduces the accuracy of autonomous (i.e.,non-differential) GPS navigation for all but friendlymilitary forces to about 50 meters RMS (100 meters2dRMS). This is the dominant error source forcivilian navigation, and it has been the primarymotivator for differential GPS services.

The U.S. Government has announced that SA will bediscontinued by at least 2006 and possibly as early as2000. You may be surprised to learn that althoughthe DoD will not have its replacement for SA (whichis denial of GPS access to the enemy in a local areaof conflict) in full operation for many years, there islittle DoD support for continuing SA. Therefore,many are optimistic that SA will be discontinued asearly as 2000.

11

Now that DGPS services are so prevalent, thenavigation community realizes that DGPS willcontinue to have a significant role to play even afterSA is discontinued. Without SA, DGPS accuracywill improve and always be better than autonomousGPS accuracy. Equally important, however, is theextra integrity provided by DGPS, which assures thata GPS satellite failure cannot cause a subtle butdangerous navigation error.

Now is the time to plan for changes in DGPS systemswhen SA is discontinued. This is especially true forthe confederated Beacon DGPS service, whichpreferably will continue to employ common signalstandards. Much time will be needed to agree on andto coordinate the changes.

The implications of no SA are profound. Thefundamental change will be to have a very stablesatellite clock rather than one which intentionallywanders randomly. The rate of SA wander hasforced DGPS signals to provide fresh correctionsevery few seconds, otherwise the wander quicklygenerates meters of navigation error. With verystable satellite signals, accuracy can be maintainedwith infrequent DGPS updates. For example, thedrift of a stable satellite signal will be measured incentimeters per minute, which means that correctionsevery few minutes will be more than adequate tomaintain today’s DGPS accuracy.

Many opportunities will be created when SA isdiscontinued. Without a message change (althoughchanges could improve these results), it is clear thatthe range of beacon coverage will increasesubstantially and there will be much greaterresistance to interference. These improvements occurnaturally because complete corrections are neededonly every few minutes. Excellent performance willcontinue even if most messages are lost due to weaksignals or to interference. (Low rate or intermittentcorrections might seem to violate integrityrequirements. However, with improved accuracy,Receiver Autonomous Integrity Monitoring (RAIM)will be extremely robust and should more thancompensate for low rate messages.)

With no SA, it will be possible to modify BeaconDGPS signals to achieve decimeter (if notcentimeter) accuracy, and in three dimensions.Accurate knowledge of keel depth relative tounderwater topography has enormous economicvalue. On the other hand, reducing the data ratewould allow many more non-interfering channelswithin the same band.

The purpose of this discussion is not to offersolutions. It is to say that eliminating SA will have amajor impact on all DGPS services, and this new eracould begin in a few months. An SA-freeenvironment will, of itself, improve performance, butit will invite change in order to take better advantageof the new environment. We should begin work nowto understand and to accelerate the change process.We have three recommendations:

1. This community should actively call for SA to bediscontinued as soon as possible, preferably in2000.

2. Studies, experiments, and discussions of howbest to use the SA-free environment should beginnow.

3. The level of communication among BeaconDGPS confederation members, suppliers, andusers should increase quickly in order to preparefor a coordinated transition when the inevitablechanges occur.

7. Summary and ConclusionsWe are pleased that Bill Adams, president of theRTCM, asked Leica to prepare this paper. Itprompted us to conduct a survey of commercialusers, knowledgeable sales agents, and experiencedservice representatives to determine how they viewedthe Beacon DGPS service. It also caused us to thinkmore deeply about overall system issues. Some ofthe results already have influenced Leica’s thinkingabout user equipment, and other results probablyshould influence further development of theworldwide Beacon DGPS service.

Three sections of background material and overallperspective were presented. The first showed howthe Beacon DGPS service evolved, not as a singlesystem but as a confederation of many systems. Thereasons behind common signal standards wereexplained and praised, but the confusion ofinconsistent definitions and terminology were noted.The second of the background sections described thepresent and evolving signal coverage. The third ofthese sections presented the key results from theLeica survey.

After the three explanatory sections, we focused oneight specific problems and suggested possiblesolutions for each. We called for a more centralizedway to handle and to speed the communicationbetween confederation members and the worldwideusers. In particular we suggested that terminology becoordinated and that worldwide status information be

12

made available from a common source. Ourrecommendation is for IALA to lead these efforts.

It was clear that mariners are confused by theambiguity of the “unmonitored” status and by what todo differently when various Type 16 messagesappear. The result is the potentially dangerousbehavior of ignoring most if not all warningmessages. We recommended that IALA work withthe confederation to standardize meanings andpublish these results. We also raised the possibilityof providing the mariner with a simple but effectivegreen, amber, or red indication, which would includeon-board integrity monitoring.

We pointed out that some navigation equipmentworks much better than others, and we noted thatsuch poor equipment is allowed on-board byinadequate regulations coupled with the desire of shipowners to buy the cheapest equipment that meetsminimum requirements. Unfortunately we did notoffer a simple solution to this problem.

The trend in beacon receivers has been towardautonomous tuning based on signal strength.Because of inter-station interference and otherfactors, we recommended a return to the earlier ideaof having the GPS receiver tune the beacon receiver.This will lead to more dependable performance andreduce the amount of manual intervention nowrequired. We also recommend that all beaconstations should transmit the Type 7 almanac message.

It was troubling to learn from our survey that somebridge integrated systems can allow the autopilotspeed input to drop to zero if DGPS corrections stopfor any reason. We recommended that companieswhich do the integration and companies which supplythe DGPS equipment must increase communicationin order to design interfaces and backups which avoidsuch dangerous conditions. This also calls for betterunderstanding by regulatory agencies.

Inter-Station interference has become worse becauseof the increased number of DGPS beacon signals.Ways to reduce the effect of interference by smartsignal tuning and use were suggested. Also, ways toadd more channels were offered.

One of the biggest problems for Beacon DGPS isboth natural and manmade interference. Suggestionsfor increasing signal strength and for smarter use ofthe existing signals were made. It was recommendedthat mariners be taught how to find and reduce oreliminate sources of own-ship interference usingsimple test equipment. To discontinue SA also was

shown to be one of the most effective ways to defendagainst interference.

Finally, realizing that SA will disappear, perhaps justmonths from now, gave rise to a range ofpossibilities. A natural consequence is that DGPSperformance will immediately become more accurateand more robust. The bandwidth released could beused to increase the number of non-interferingchannels, or it could be used to provide multiplefrequency phase corrections, leading to decimeter orcentimeter accuracy. The desire to maintain commonsignal standards across the Beacon DGPSconfederation was expressed, coupled with a call tobegin analysis and discussions immediately abouthow best to serve navigators in an SA-freeenvironment. We also urged this community toactively call for the elimination of SA as soon aspossible, preferably in the year 2000. This would bea great Millennium gift to all navigators.

8. References[1] R. Franko, S. Mittal, T. Stansell, R. Harris, E.D’Amico, and S. Cannon, “A New Generation ofDGPS Broadcasting Stations”, Proceedings of IONGPS-98, Nashville, Tennessee, September 1998.

Leica Geosystems Inc., 23868 Hawthorne Blvd., Torrance CA 90505, USA

www.leica-geosystems.com

Leica Geosystems Torrance USA 1999. Paper presented at RTCM in Orlando (FL), May 99

1023-Letter-5/99

Leica Geosystems, TorranceMay, 1999BBiographiesIntroductionThe Beacon DGPS Development PathBeacon CoverageUUser ExperienceGeneral View of the Service ProvidedTransmission QualityInterferenceSystem IntegritySwitching between beacon stationsShipboard Equipment

Problems and Potential SolutionsImprove International Coordination and User CommunicationProvide Simple and Meaningful Status IndicationsEmploy Higher Quality User EquipmentReturn to More Intelligent Beacon Signal SelectionImprove Bridge IntegrationReduce Inter-Station InterferenceDefend Against Natural and Manmade InterferenceEncourage and Prepare for an SA-Free Environment

Summary and ConclusionsReferences