Microwave Journal
www.microwavejournal.com/articles/2251-gps-in-10-years

GPS in 10 Years

Improvements planned for the Global Positioning System (GPS) over the next decade

February 1, 1998

GPS in 10 Years

What will the Global Positioning System (GPS) look like in 10 years? This article discusses improvements to the overall GPS planned over the next decade and examines their impact on system performance for several applications. The Presidential Decision Directive (PDD) released in March 1996 states that selective availability (SA) will be turned off within 10 years. Efforts have been ongoing during the past year to place a second civilian frequency on the Block IIF satellites. In addition, a program known as the GPS Modernization Effort, or GPS-III, is underway to identify future enhancements to GPS. Finally, the Air Force is in the process of upgrading the Control Segment, which includes the Accuracy Improvement Initiative (AII). These enhancements to GPS, combined with improved user equipment expected to be developed over the next 10 years, will improve its accuracy, integrity and availability significantly. For example, removal of SA not only improves the GPS positioning accuracy, but allows a significant increase in the availability performance of integrity-monitoring algorithms such as receiver autonomous integrity monitoring (RAIM), and fault detection and exclusion (FDE). Upgrades to the Control Segment also will improve the overall integrity of GPS. Will these changed improve GPS enough in 10 years for it to be a stand-alone Global Navigation Satellite System (GNSS)? For many applications, the answer to this question is a definite yes. For other more demanding applications, GPS still will need augmentations, but this article shows these augmentations can be much simpler and less costly than often envisioned.

Karl L. Kovach
ARINC Inc.
El Segundo, CA

Karen L. Van Dyke
US Department of Transportation/Volpe Center
Cambridge, MA

Think back to what GPS was like over 10 years ago...way back to 1987. It was two years since the final Block I satellite had been launched and still two years until the first Block II satellite would be placed in orbit. Generally, there were not enough visible satellites to form a position solution and when four or more satellites were visible to the user, it was only for a short period of time. These visibility windows opened and closed four minutes earlier each day. The old-timers will remember waking up at 2 am to catch the start of the window at 3 am. Of course, one advantage they had was that these satellites did not have SA on them.

There was not much in the way of a user population. A small number of military receivers and some surveying systems were in existence, along with a few commercial manufacturers with grand ideas for the future. GPS was envisioned for oceanic aviation use and possibly nonprecision approach operations, but certainly never for precision approach.

Between 1987 and 1997, GPS blossomed into a system that has been referred to as the next utility. Twenty-six Block II/IIA satellites were launched successfully while many of the Block I satellites significantly outlived their design life of five years, providing the opportunity for users to gain operational experience with the system. In July 1997, the first launch of the Block IIR satellite to replenish the GPS constellation occurred after a IIR was lost to a launch failure earlier in the year.

During this time frame, GPS use has exploded into applications previously unimaginable, ranging from recreational use (hiking, sailing and golfing) to all modes of transportation and more transparent applications such as synchronization of communication networks. Although the military has made great use of GPS (most notably during Operation Desert Storm), civilian use has dominated the market.

The problem of not having four or more visible satellites became a rare occurrence rather than the norm. However, as applications began to push this technology to the limit, new problems arose. Users required more accuracy, integrity, continuity and availability from the system. For example, RAIM requires five visible satellites for fault detection and six satellites for fault detection and exclusion. This requirement leads to holes in the coverage that become significant (lasting an hour or more) for more stringent phases of flight.

In March 1990, SA was turned on, degrading the accuracy of GPS significantly. This decreased accuracy angered civilian users of the system and has been a sore spot with them ever since. As a result, differential GPS (DGPS) systems began sprouting up not only to defeat SA, but to provide even greater accuracy by removing ionospheric errors.

The past few years have seen increased civilian input into GPS operation and its future capabilities. During 1997, the GPS Modernization Program began. The program is a combined Department of Defense (DoD)/Department of Transportation (DoT) effort to define near-, medium- and long-term requirements for GPS. Those providing input on this activity were encouraged to think outside the box and not be restricted to the current system’s operational characteristics.

What will happen from now until 2007? Of course, this forecast is difficult to predict precisely (crystal balls are notoriously prone to give whatever answer is wanted, rather than the truth). Just try to imagine predicting GPS as it exists today, based on its status in 1987.

Basic Paradigms
The PDD, issued in the spring of 1996, is a good framework from which to operate when examining where GPS will be 10 years from now.
1 The basic GPS constellation still will be operated by the DoD and the service will be available free of any direct user charges. The PDD is elaborated on in the 1996 Federal Radionavigation Plan (FRP), which defines the roles for the DoD and Department of State.2

By 2007, there will be a combination of GPS regional satellite-based augmentation systems (SBAS), local area ground-based augmentation systems (GBAS) and special application augmentation systems, as listed in Table 1 . The SBASs will comprise the Federal Aviation Administration (FAA) Wide Area Augmentation System (WAAS), the European Geostationary Overlay Service (EGNOS) and the Japanese Multifunction Satellite Augmentation System (MSAS). The SBASs are designed primarily for aviation use although they may serve other user communities as well.

Table I
Regional/Local/Special Augmentations

Regional (SBAS)

Local (GBAS)

Special Applications

WAAS

LAAS and APLs

Geodynamics

EGNOS

RTCM SC-104

Local Survey Nets

MSAS

Commercial Services

Special Kinematics (< 1 cm)

For aircraft precision approach, local area augmentation systems (LAAS) will be installed, which may include airport pseudolites (APL) that broadcast GPS-like ranging signals from the ground. Also, government-provided DGPS networks will be available to serve the land and marine communities with RTCM SC-104 correction data and commercial services, such as broadcasts using the FM subcarrier. These systems already exist and most likely will be expanded in the future. Other differential networks also are and will be available for special application projects.

In addition, non-GPS-based augmentations will be implemented, such as the Russian Global Navigation Satellite System (GLONASS) and inertial navigation systems. However, complementary US radionavigation systems may not be available. According to the 1996 FRP, Loran-C will be turned off at the end of 2000 and the VHF Omnidirectional Range and Instrument Landing System navigation aids will begin to be phased out starting in 2005. This situation will lead to strong reliance on GPS and essentially placing all of our eggs in the GPS basket. However, given the strong reluctance of the user community to turn off navigation aids as witnessed by the recent user conference on Loran and the demand for backup systems, it is difficult to imagine that these systems will be phased out without more operational experience with GPS for safety-critical functions.

Important Changes Ahead
Some important changes are in the works for GPS that should make GPS better for everyone by 2007. We say, "should make GPS better" rather than "will make GPS better" to acknowledge that our anticipation of these GPS changes is still fundamentally a prediction of the future. No matter how solid the plans for these system changes may seem today, tomorrow is another day and anything can happen.

We can foresee two major possibilities that might derail the anticipated GPS changes. Realistically, we do not think that either one will come to pass, but both situations are possible. Either or both could happen. However, in keeping with our optimistic outlook for the future, we make several assumptions about the two possibilities not happening.

We assume the US Congress will continue to fully fund GPS, which must be operated and maintained. Eventually, GPS satellites will fail and will have to be replaced. Cutting back to only 18 satellites could save taxpayers a lot of money (GP$). This sort of a cutback has happened before. Originally, GPS was designed as a 24-satellite constellation but was cut back to 18 satellites plus three operating spares in the early 1980s for budgetary reasons (recall 18 + 3 was the GPS baseline back in 1987). Although GPS cutbacks might save GP$, the total economic impact on taxpayers surely would be a net loss due to the additional funds that would have to be spent elsewhere (for example, more robust augmentations).

We also assume the world will remain at peace (at least suffer no greater level of conflict than at present). We must not let ourselves forget that GPS is still a military system. The PDD1 reminds us of the fact that GPS was designed as a dual-use system with the primary purpose of enhancing the effectiveness of US and allied military forces. GPS will remain responsive to the US National Command Authorities (NCA). If a hostile force was to use GPS to drop bombs on Washington, DC (or Pearl Harbor), then we would certainly expect that GPS characteristics and signal formats might change per NCA direction such that GPS would cease operating for the hostile force. We hope (and explicitly assume) that these scenarios will remain as nothing more than plots for Hollywood movies. Peacetime GPS operations should be normal GPS operations.

Five Important Changes
If we are lucky and the two assumptions discussed previously hold true (that is, that disruptions will not happen), then our crystal ball leads us to predict that five important changes will take place over the next 10 years for GPS: no more SA (SA set to zero), the establishment of a second civil frequency, the use of 24 space vehicles (SV) plus spares, 30 dB more anti-jam capability for the military and better accuracy for everyone (2.5 m signal in space (SIS)/user range error (URE)).

SA Set to Zero
The SA-set-to-zero change is very important, but not a very bold prediction. It is actually spelled out in the PDD1 as, "It is our intention to discontinue the use of GPS SA within a decade..." Given a 1996 date for the PDD, this means SA will be set to zero no later than 2006. The PDD also calls for an annual determination by the President on the continued use of SA beginning in 2000, thus setting the no-earlier-than date for setting SA to zero.

For civilian GPS users, this change will improve SIS accuracy substantially by instantly eliminating the approximately 24.2 m of pseudorange (PR) error induced by SA (note that 24.2 m is the root-sum-square difference between the reported Standard Positioning Service (SPS) SIS accuracy value of 24.6 m 1 sigma and the Precise Positioning Service (PPS) SIS accuracy value of 4.3 m 1 sigma).3 For military PPS users, this change offers no benefits.

Second Civil Frequency
The addition of a second civil frequency to the next generation of satellites (that is, Block IIF) is another important change whose coming is predicted by official government documents.
4,5 This second civil frequency will carry a clear and acquisition (C/A)-code signal and navigation message data such that it can serve as a full function backup or alternative to the current L1 C/A-code signal. The exact carrier frequency is not yet known, but should be decided upon by March. Until the second civil frequency C/A-code signal is available, civil access to the L2 Y(P)-code signal using codeless techniques has been guaranteed. By 2007, roughly one-half to two-thirds of the on-orbit satellites should be broadcasting this second civil frequency C/A-code signal.

For civilian GPS users, this change will provide the capability to compensate directly for the PR errors caused by ionospheric delays using dual-frequency measurements. Typically, these PR errors are the second largest contributor to the SPS accuracy budget after SA. Thus, the ability to compensate for them directly will improve the received SPS accuracy substantially. For military PPS users, this change provides no improvement since PPS receivers have always been able to track the Y(P)-code signals on L1 and L2 for their dual-frequency ionospheric delay compensation. (However, note that some PPS receivers are limited to L1 operation for cost reasons.)

Twenty-four SVs Plus Spares
The availability of 24 SVs is perhaps more of an important recognition of the way that GPS constellation sustainment strategy has evolved. The Air Force has committed to maintaining a 24-SV GPS constellation.
6 To fulfill this commitment, the sustainment strategy has become launch on expected failure rather than launch on need or launch on schedule. Under this strategy, replacement SVs (satellites) are launched before they actually are needed to prevent holes in coverage from occurring when the expected satellite failure finally happens. Typically, this strategy results in more than 24 satellites operating in the constellation at any one time.

As an example of this launch-on-expected-failure strategy in action, Table 2 lists the nominal probabilities of various numbers of satellites operating in the constellation for a launch-on-need strategy alongside the probabilities of various numbers of satellites based on the current implementation of the launch-on-expected-failure strategy. The probabilities are taken from the actual observed on-orbit performance of the GPS constellation over the 750-day period from the internal declaration of full operational capability on April 27, 1995 until May 15, 1997 (roughly 50 total satellite years). The probabilities only include bona fide operating time where the satellite was operating normally and there was no Notice Advisory to Navstar Users (NANU) indicating any scheduled satellite downtime.

Table II
Nominal vs. Actual Satellite Constellation Number Probabilities

SV's

Nominal Satellite Probability (%)

Actual Satellite Probability (%)

26

0.00

3.94

25

0.00

31.81

24

72.00

53.10

23

17.00

10.79

22

6.40

.36

21

2.60

0.00

20

1.20

0.00

19

0.60

0.00

< or = 18

0.20

0.00

Anticipatory NANUs tend to be conservative, indicating a larger scheduled satellite downtime than actually ends up being used. To maintain this conservativeness, the listed data count all scheduled-down-but-still-up times as downtime. For reference, the results correspond to an unscheduled outage rate of 1.2 events per satellite year with a mean duration of 35 hours per event, plus a scheduled outage rate of 2.2 events per satellite year with a mean duration of 13.8 hours per event. For both civilian and military GPS users, this change improves availability dramatically by reducing the occurrence of SIS coverage holes. The holes are far less frequent and, when they do occur, are much less severe.

30 dB More Anti-jam
Unlike the other important changes expected to occur over the next 10 years, the 30 dB increase in anti-jam change is focused almost exclusively on military GPS users and is subject to much less certainty as to how the change will be implemented. Several avenues are being explored as possible routes for this change, ranging from fancy antennas and advanced signal processing techniques for military GPS receivers to some potentially major modifications to the GPS SIS itself.
7–9 The efforts to bring about this change are a large part of the GPS Modernization Program being pursued by the GPS Joint Program Office.10

For military GPS users, this change will help ensure survival of their GPS satellite navigation (SATNAV) capability on future battlefields. There is nothing magical about the 30 dB value for this change. It is not enough to guarantee survival of a user’s GPS SATNAV capability in the future, but it does go a long way toward helping ensure that survival. For civilian SPS users, this change offers no benefits. At worst, it may instead impose some drawbacks if any SIS changes are not 100 percent backwards compatible. Although no incompatibilities are anticipated, a concept for bandwidth limiting the SPS to ±4 MHz, which may or may not be innocuous, has been described.11

2.5 m SIS URE
The 2.5 m SIS URE change is the result of a number of different efforts associated with the GPS Operational Control System (OCS). These efforts are lumped together under the AII
12 and span a broad spectrum of activities. At one end is a long-term project to incorporate up to 10 National Imagery and Mapping Agency PPS reference stations into the OCS to improve the real-time observability of the on-orbit satellite performance. At the other end are the never-ending OCS attempts to work smarter, not harder. The success of these AII efforts over the long term is demonstrated clearly by the gradual reduction of the PPS SIS URE from a value of 4.3 m 1 sigma, as reported in 1992,3 to a value of 1.9 m 1 sigma, as reported in 1997.13 (Note that the recent 1.9 m SIS URE number is indeed already less than the predicted 2.5 m SIS URE change — this just shows that predictions can come true even as those predictions are being made.)

For military GPS users, the improved accuracy resulting from this 2.5 m SIS URE change is useful and available immediately. For civilian GPS users, this improvement currently is lost in the noise compared to the inaccuracies caused by SA and single-frequency ionospheric delay compensation. However, once these two error sources are omitted (SA set to zero and the second civil frequency), the SPS SIS URE and the PPS SIS URE then will be basically the same and civilian GPS users thus will be able to benefit from the 2.5 m SIS URE change as well.

The Impact of These Five Important Changes
By 2007, the basic GPS (that is, the baseline GPS of 1997 plus the five important changes described previously) ought to provide good service for all users. But will that GPS (PPS/SPS) service be good enough to use without additional augmentation? If the answer is yes, then perhaps the scope of those augmentations — especially the government-provided ones — can be scaled back to save taxpayers some money. If the answer is no, then perhaps GPS should be cut back and the resulting GP$ savings could be spent on more robust augmentations.

To explore this GPS vs. augmentation trade space, a worthy place to start is by considering the requirements for a generic GNSS to support en route aircraft navigation. The GNSS requirements for such an application, expressed in terms of the required navigation performance (RNP), are becoming fairly mature with broad international consensus being reached through the auspices of the International Civil Aviation Organization (ICAO). Although the ICAO concept of RNP is focused on aviation, its rigorous definitions and categorizations of navigation characteristics also are widely applicable to nonaviation applications. Therefore, the specific situation to deal with when exploring this one small section of the total GPS vs. augmentation trade space is to determine whether GPS (baseline + changes) will satisfy the selected RNP characteristics.

RNP Characteristics
Currently, the RNP characteristics for GNSS are provided in draft form by the ICAO.
14 These RNP characteristics come in two forms: one for the GNSS navigation system (the system onboard each aircraft) and one for the GNSS SIS (the signals shared by every aircraft in a region). For simplicity, the example discussed in this article is limited to the GNSS SIS RNP characteristics for the RNP-1 en route phase of flight. For this phase of flight, four RNP characteristics are necessary: accuracy, integrity, continuity and availability. The current draft values for these characteristics are listed in Table 3 .

Table III
GNSS RNP Characteristics (Draft Values)

RNP Characteristics

Example En Route (RNP-1) Values

Accuracy

1.0 nautical miles; 95 percent, lateral

Integrity

1x10-7 /hr; unalerted hazardously misleading information (UHMI)

Continuity

1x10-6 /hr; unanticipated loss of RNP service type

Availability

0.9999; applies to the entire set of RNP characteristics

A Future GPS Performance Examination
The performance of the future GPS (baseline + changes) now can be examined with respect to the four necessary RNP characteristics. Because the RNP characteristics build upon each other, this examination must be performed in a particular order. The summary correlation matrix between the five important changes and the four RNP characteristics they impact is listed in Table 4 .

Table IV
Changes/RNP Characteristics Correlation Matrix

 

Accuracy

Integrity

Continuity

Availability

SA set to zero

major

major

minor

minor

L2 C/A-code

major

major

minor

minor

24+ Satellites

minor

minoro

minoro

major

30 dB AJ

-

-

minor

minor

2.5 m SIS URE

major

major

-

minor

o Impact is primarily through continuity availability

The combination of the improved SIS URE provided by the SA-set-to-zero, L2 C/A-code and 2.5 m SIS URE changes, along with the generally improved horizontal dilution of precision (HDOP) values afforded by the 24+ satellites change, will allow the future GPS (baseline + changes) to satisfy the 1.0 nm, 95 percent RNP accuracy characteristic with an even greater margin than the current GPS (not a difficult value to meet). Of course, there still are times when the HDOP is so poor that the RNP accuracy value cannot be met even with a 2.5 m SIS URE. However, these occasions are rare enough that they can be defined as unavailability events and simply be charged off against the RNP availability number. Therefore, we can equivocally predict that the future GPS will satisfy the 1.0 nm, 95 percent value for RNP accuracy.

Similar to accuracy, GPS integrity (in the RAIM/FDE sense) also reaps major benefits from the improved SIS URE provided by the SA-set-to-zero, L2 C/A-code and 2.5 m SIS URE changes. The SA-set-to-zero and L2 C/A-code changes reduce the background noise level that the SPS receiver RAIM algorithms must contend with to roughly the same level as for the current PPS SIS URE. The performance of RAIM algorithms with the current PPS SIS background noise level has been found acceptable15 whenever there are sufficient satellites (or, optionally, sufficient satellites plus a barometric altimeter) in view with adequate fault-detection geometry. Acceptable SPS RAIM algorithm performance also has been demonstrated with only the SA-set-to-zero change.16 The combined benefit of all SIS-URE-related changes is more than enough to allow GPS receivers with RAIM algorithms (PPS or SPS) to satisfy RNP integrity even under poor fault-detection geometry. Of course, the limiting factor here is how poor the fault-detection geometry can be while still allowing the GPS receiver to satisfy RNP integrity. The limits on the adequacy of the fault-detection geometry are analogous to a cutoff threshold for a binary available/unavailable switch. Fault-detection geometries better than the cutoff threshold mean RNP integrity is available; fault-detection geometries worse than the cutoff threshold mean RNP integrity is unavailable. Since there is a separate RNP characteristic for availability where the accounting for inadequate fault-detection geometries belongs, we can unequivocally predict that the future GPS will satisfy the 1 • 10–7/hr probability of UHMI value for RNP integrity provided RNP availability exists.

Continuity is the RNP characteristic that deals with unexpected (surprise) losses of availability. In common usage, availability can mean either availability of accuracy (that is, availability of navigation) or availability of integrity (that is, availability of fault detection). These two availabilities, and the corresponding continuities, are typically different. Say there are five satellites in view with good HDOP (that is, good accuracy geometry) and good fault-detection geometry. Hence, both accuracy and integrity are available. If one of those five satellites should suffer a surprise hard failure, there would be a surprise loss of integrity (that is, the continuity of fault detection would be impacted). The four remaining satellites could still have a good enough HDOP such that there would be no loss of accuracy (that is, no impact on the continuity of navigation).

Although the distinction between these two types of continuity exists, the limiting case for GPS turns out to be when there are five satellites in view with good fault-detection geometry and one of those satellites suffers a surprise soft failure (that is, when the potential for HMI occurs). In this case, assume the GPS receiver RAIM algorithm detects the failure and issues a do-not-use alert. This condition does not really represent a loss of integrity/fault detection since adequate fault-detection geometry exists and the RNP integrity characteristic is still satisfied. Similarly, the do-not-use alert condition does not really represent a loss of accuracy/navigation since there is still an available combination of four satellites that satisfies the RNP accuracy characteristic. Instead, this limiting case condition constitutes a surprise loss of service that, in turn, impacts the continuity of service (COS), sometimes called continuity of function.

For the future GPS, the predicted COS for the limiting case of five satellites is not really much different from that for the current GPS: roughly 1 • 10–5/hr. Since this value is approximately an order of magnitude less than the RNP continuity characteristic discussed previously, the minimum acceptable case for RNP operation is thus six satellites in view (or five satellites plus barometric altimeter) with the geometry available to support both fault detection and fault exclusion. Following the precedent for RNP integrity by counting inadequate FDE geometries against the RNP characteristic for availability, we can predict that the future GPS will satisfy the 1 • 10–6/hr probability of COS value for RNP continuity provided GPS is available.

Availability is the final capstone RNP characteristic. It is the one that suffers all of the degradations that result from the need to have adequate geometry for accuracy, adequate fault-detection geometry for integrity and adequate FDE geometry for continuity. If GPS is available in the RNP characteristic sense, then the RNP continuity, integrity and accuracy values will be satisfied. If any one of the lower tier RNP characteristics cannot be met, then neither can RNP availability. This situation is where the 24+ satellites change proves its worth. If the 24-satellite constellation is maintained as predicted, then the future GPS can meet the 0.9999 value for RNP SIS availability. However, to accommodate this satisfaction of RNP SIS availability, the GPS receivers that are attempting to meet the RNP characteristics must participate in the solution. Those GPS receivers must be capable of either tracking GPS satellites down close to the horizon when absolutely necessary for FDE or using a barometric altimeter when necessary for FDE. Neither of these requirements is particularly onerous. The first is already recommended in the FAA primary means oceanic approval17 while the second one is in the Technical Standard Order for supplemental GPS navigation.18 Given such GPS receiver participation and assuming the 24+ satellites change comes to full fruition, we can predict that the future GPS will satisfy the 0.9999 value for RNP availability.

Future GPS Performance Results
The results of the future GPS (baseline + changes) performance examination are listed in Table 5 with respect to the RNP-1 characteristics. The accuracy, integrity and continuity RNP characteristics are predicted to be satisfied unequivocally. The availability RNP characteristic is predicted to be satisfied based on two reasonable assumptions: GPS receiver participation through the ability to track low elevation angle satellites and/or the use of barometric altimeters, and the 24+ satellites change.

Table V
Future GPS (Baseline + Changes) Performance Predictions

RNP Characteristics

Draft En Route (RNP-1) Values

Satisfied by Future GPS (Baseline + Changes)

Accuracy

1.0 nautical mile; 95 percent, lateral

Yes

Integrity

1x10-7 /hr; UHMI

Yes

Continuity

1x10-6 /hr; COS

Yes

Availability

0.9999 for accuracy, integrity and continuity

Yeso

o Assuming GPS reciever participation and the 24+ satellites change

Augmentation Benefits
Over the next 10 years, the five important changes described in this article will make the basic GPS a much better system. These improvements also will make the system much easier to augment. For example, DGPS augmentations will have a much easier task keeping up with the naturally occurring error growth over time (for example, a rate of 0.0001 m/s) than they do trying to keep up with the SA error growth over time (that is, a rate of up to 2 m/s).
19 Users will be able to perform their own high accuracy ionospheric delay compensation using dual-frequency measurements, thereby simplifying WAAS implementation. For applications that need extremely high accuracy, such as Category III precision approach, direct access to L1 and L2 might make wide lane ambiguity resolution reliable enough for operational use. Twenty-four+ satellites will improve the availability substantially, perhaps enough such that ranging source augmentations may be no longer needed in many cases. If the SIS URE continues to improve and ends up being much lower than 2.5 m, some DGPS augmentations may be able to shift their focus to monitoring integrity rather than improving accuracy and thereby free up some transmission bandwidth for additional value-added services such a broadcasting weather information or GPS NANUs.

Conclusion
In 10 years, GPS will be extremely improved all by itself. Assuming GPS receivers continue to improve and the predicted changes to the system described in this article occur, GPS will be good enough to use as a stand-alone GNSS for applications such as en route navigation at the RNP-1 or higher levels. However, augmentation systems still will be required for more demanding RNP levels, such as those required for precision approach.

Even though we are strong believers in the current GPS and have what some might call an extremely rosy outlook for the future of GPS and its augmentations, we are still not completely comfortable with the notion of putting all of our navigation eggs in one basket. As described previously, there are at least two major possibilities that might derail the future GPS. We are sure there are many minor possibilities as well. It does not seem a prudent course of action to intentionally rely exclusively on any one system regardless of whether that system is GPS, an augmented GPS or some other single-string GNSS. A number of radionavigation systems2 exist, all of which currently are slated for termination. With the initial investments in these systems already paid for, it may be wise to consider keeping one or two of them as a backup just in case.

Acknowledgment
This paper was originally presented at the Institute of Navigation (ION) GPS-97 conference in Kansas City, MO, September 18, 1997.

References
1. President of the United States of America, Presidential Decision Directive NSTC-6, Subject: US Global Positioning System Policy, March 28, 1996.

2. 1996 Federal Radionavigation Plan, DOT-VNTSC-RSPA-97-2/DOD-4650.5, US DoD and US DoT, July 1997.

3. S. Malys and S. Holm, "PPS and SPS Integrity Monitoring with an Independent Global Tracking Network," Proceedings of the Fifth International Technical Meeting of the Satellite Division of the Institute of Navigation (ION GPS-92), Albuquerque, NM, September 16–18, 1992.

4. White House Commission on Aviation Safety and Security, Final Report to President Clinton, February 12, 1997.

5. DoT and DoD Assure GPS Access for Civil Users, US DoT Press Release 21-97, February 27, 1997.

6. "President Opens Door to Commercial GPS Markets: Move Could Add 100,000 New Jobs to Economy by Year," White House Press Release, March 29, 1996.

7. National Research Council, The Global Positioning System, A Shared National Asset, National Academy Press, 1995.

8. S. Ragahavan, J. Holmes, S. Lazar and M. Bottjer, "Tricode Hexaphase Modulation for GPS," Proceedings of the 10th International Technical Meeting of the Satellite Division (ION GPS-97), Institute of Navigation, Kansas City, MO, September 16–19, 1997.

9. J. Anderson and S. Lazar, "Lm and Lc: How the Military and Civil Users of GPS Can Reuse Our Existing Spectrum," Proceedings of the 10th International Technical Meeting of the Satellite Division (ION GPS-97) Institute of Navigation, Kansas City, MO, September 16–19, 1997.

10. Global Positioning System Joint Program Office, "Special Notice: Global Positioning System, Request for Information on a New GPS Signal in Space, Commerce Business Daily, June 26, 1997.

11. J. Armor, "GPS: A Joint Program Office Perspective," Proceedings of the 53rd Annual Meeting of the Institute of Navigation, Albuquerque, NM, June 30–July 2, 1997.

12. S. Malys, "The GPS Accuracy Improvement Initiative," Proceedings of the 10th International Technical Meeting of the Satellite Division (ION GPS-97), Institute of Navigation, Kansas City, MO, September 16–19, 1997.

13. R. Conley, "Results of the GPS JPO’s GPS Performance Analysis Baseline Analysis: The GOSPAR Project," Proceedings of the 10th International Technical Meeting of the Satellite Division (ION GPS-97), Institute of Navigation, Kansas City, MO, September 16–19, 1997.

14. Standards and Recommended Practices (SARPS) for Global Satellite Navigation System (GNSS), Draft, International Civil Aviation Organization (ICAO), summer 97 version.

15. K. Kovach, H. Maquet and D. Davis, "PPS RAIM Algorithms and Their Performance," NAVIGATION, Journal of the Institute of Navigation, Vol. 42, No. 3, Fall 1995.

16. K. Van Dyke, "Removal of SA: Benefits to GPS Integrity," Proceedings of the Ninth International Technical Meeting of the Satellite Division of the Institute of Navigation (ION GPS-96), Kansas City, MO, September 17–20, 1996.

17. Notice (N) 8110.60, Subject: GPS as a Primary Means of Navigation for Oceanic/Remote Operations, US DoT, FAA, December 4, 1995.

18. Technical Standard Order (TSO) C129, Subject: Airborne Supplemental Navigation Equipment Using the Global Positioning System (GPS), US DoT, FAA, December 10, 1992.

19. Global Positioning System Standard Positioning Service Signal Specification, second edition, US DoD, June 2, 1995.

Karl L. Kovach received his BS degree in mechanical engineering from the University of California at Los Angeles. He has over 19 years of experience in various aspects of the GPS program, including three years as the Air Force officer-in-charge of the GPS control segment when it was at Vandenberg Air Force Base, CA (1983–1986). Currently, Kovach is a senior principal engineer with ARINC Inc. in El Segundo, CA.

Karen L. Van Dyke received her BS and MS degrees in electrical engineering from the University of Massachusetts at Lowell. She is a member of the technical staff with the Center for Navigation at the US DoT/Volpe Center in Cambridge, MA. Currently, Van Dyke is conducting GPS integrity studies for the FAA in support of RTCA SC-159.