Abstract
The Global Positioning System (GPS) was declared operational on December 8, 1993. Since that time the system and the receivers that use it have been improved to a level of performance unimagined by its creators. Also, the system has been integrated into the daily life of societies all around the world. It is, generally, agreed that a disruption of the system would deal a heavy blow to many of the world’s users. Failure of systems ranging from aircraft navigation to cell phones and automatic teller machines would occur. Suddenly, after more than 20 years of operation, there is an almost frenzied rush to find alternatives to the system in case of a sudden failure. This paper discusses some of the proposals that have been made and possible alternatives to those proposals. It argues that a rational approach is needed, rather than a rush to judgment.
I. Caveat
This paper contains many opinions and suggestions. They are those of the author, who is reasonably adept at developing and analyzing technologies. First person will be used when expressing the author’s ideas. Comments about political issues are only those of an educated taxpayer and voter, who reads a newspaper every day. I have, also, been involved in the use of the government procurement system and the development of and furthering of government programs for fifty years. A short bio will be attached at the end of this paper.
II. Introduction
Satellite navigation aids coupled to navigation systems (GNSS) have become an integral part of modern life. The satellite systems are not, in themselves, navigation systems. Rather, they enable the determination of the coordinates of a point on, or above, the surface of the earth.
The position information is updated at a rate high enough that it can be used with a digital map to determine direction and velocity of travel, and by employing rather complex algorithms, one can generate guidance instructions to enable efficient travel from one point to another. In addition, the satellites contain highly accurate clocks, which are synchronized to a worldwide standard, and, thus, provide accurate time as well as positioning measurements. Many important systems, such as cell phone networks, the Internet, and, even, automatic teller machines (ATMs) require the timing information to operate properly. Loss of the timing signal would, and has a few times, render these systems inoperable.
The Global Positioning System (GPS) was developed by the military services of the United States (U. S.). It resulted from many science and technology experiments performed, mostly, by the Navy (USN) and the Air Force (USAF). The program office which first built the system was managed by the USAF, and the Air Force, is today responsible for maintaining, controlling, and enhancing the system. The USN has, however, remained extremely active in the fields of developing new applications of and developing suggestions for upgrades to the system. The maintenance and enhancement of the timing function of the GPS is, primarily, a Navy task. The S. Army (USA) is a major user of the system. In my opinion, the GPS is the most important system ever developed by the military forces of the United States. The U. S. military has been conscientious and dedicated to the task of supplying GPS services to the entire world, not just to the U. S. military forces. We shall see, however, that the military connection has been the source of some consternation abroad.
Recently, there has been a considerable interest shown in the fact that the nation has become very dependent on the GPS signals and the capabilities that they enable. There have been many presentations by those who claim that GPS signals are weak enough to provide a significant vulnerability to interference and jamming. Government committees have received briefings on the subject, papers have been presented at professional society meetings, Congressional hearings have been held, and, even, a proposed law has been submitted to the House of Representatives [1].
Most of the presentations have been oriented toward the establishing of an eLoran system in the U. S. to serve as a backup system if the GPS signal were unavailable. All of the attention to the subject has resulted in what I describe as a sky is falling attitude in some groups. In my opinion, eLoran is an inadequate and obsolete technology that is not, by any stretch of the imagination, a substitute for, or a replacement for the GPS. ELoran is not adequate for precision weapon guidance, or for precision airport approach. If we fall back to eLoran because GPS is not available, we lose those valuable capabilities immediately and they cannot be replaced.
My point is that we should not be discussing how to replace one of the most valuable parts of American infrastructure because it has some weaknesses. While a backup may be part of the discussion, most of the debate should be centered on ways of eliminating the problems of GPS. Now is not the time to retreat to a last century technology because of fears that may be groundless.
The questions to be answered are: What to protect? Against what? How?
III. Vulnerabilities
1. System degradation
There have been predictions from the government many times that satellite failure will compromise the system. Fortunately, the lifetime of GPS satellites has exceeded their design lifetimes by a significant amount. The oldest operating satellite is IIA-1, SVN 23, which is currently operating in plane E, slot 5. It is now 24.6 years old. Design life was 7.5 years. The USAF guarantee for the constellation is only 24 satellites, 95% of the time. The current standard is the Expandable 24 configuration, containing 27 satellites and providing full operational performance. There are, currently 31 operating satellites in the constellation. Given a reasonable launch program, total system failure probability over the next 20 years is low. The 2016 defense budget includes, approximately, $1B for GPS, with no funding for GPSIII. Funding at that level is adequate for constellation sustainment.
2. Natural Phenomena
There is some vulnerability to solar storms. The satellites, however, were designed and built to work in a high radiation environment, and the storm vulnerability is, mainly, one associated with the ground equipment. Augmentation systems, such as WAAS use geocentric satellites, but, as yet, these satellites have continued to operate during storms. GPS ground equipment was somewhat affected, at least on the sunlit side of the earth by a solar storm in December, 2006, but a massive storm in March, 2012, caused no interruption of service, although communication systems were affected. The GPS appears to be more resistant to natural disruptions than many standard communication systems.
3. Electromagnetic Pulse
There has been much discussion about the possibility that high altitude nuclear explosions above the U. S. could cause considerable damage to computers, power systems, and electronics in general, by projecting an electromagnetic pulse (EMP) across the nation. [7]. The GPS satellites are located far above the most probable altitude of the detonation and are relatively safe. The EMP would be directed downward from the explosion. Electronics, in general, may be at risk, but there is no reason to believe that a GPS receiver would be more, or less, sensitive to EMP than an eLoran receiver. In fact, the lower frequency device may be more vulnerable. While the GPS satellites would survive, the eLoran equipment, which is ground based, could be greatly harmed.
4. Nuclear explosions in space
It was mentioned above that the GPS satellites are hardened against radiation. In fact, they operate in high radiation regions in space. Of course, a single nearby explosion could destroy a single satellite, but studies have shown that explosions at the GPS orbital altitude are survivable for the entire constellation.
5. Terrorist Attack
ELoran stations are, certainly, vulnerable to individual attacks, as are GPS monitoring and up- loading antennas. Damage to the GPS infrastructure would be quickly repaired. The system is designed to operate for 180 days without external corrections, although some degradation would occur. There are two control centers for the system. The master control station is located at Schriever AFB in Colorado, and the backup station is located at Vandenberg AFB in California. Both are well protected by alert military forces. Significant damage to either, or both, is unlikely. The possibility of attack on eLoran stations should not be minimized. Some notable incidents have occurred recently. On April 16, 2013, attackers in San Jose California, disabled the PG&E Metcalf substation by cutting cables and firing rifle rounds into transformers. Repairs required 27 days.
On September 26, 2014, an arsonist set a fire in the FAA Chicago Center Air Route Traffic Control Center (ARTCC). Thousands of flights were cancelled, and repairs required 18 days.
The disabling of an eLoran station near Los Angeles in a time of bad weather coupled with the strategic placement of GPS jammers in a radius of 10 miles could wreak havoc with airport and harbor navigation.
We should not become too sanguine about such possibilities.
It appears that while there may be a slight chance for short term outages from system failures, nuclear EMP, nuclear explosions in space, or terrorist attack, none of these threats are likely, or severe enough to justify the expense of a backup system, such as eLoran. But intentional jamming must be considered.
IV. Intentional jamming
Intentional jamming can take many forms. The documented cases of GPS jamming in the U.S. and in the U. K. fall, mostly in the category of vandalism. The sentinel project in the U. K. stationed jammer detectors along highways [8]. They detected many instances of jamming, and they estimated that between 50 and 450 incidences of jamming occur in the U. K. daily. Most, if not all, of the cases were the results of attempts by drivers to defeat position-monitoring devices installed in their vehicles by their employers. Most occurred during the workweek. For example, none were recorded on Christmas Day, and few on weekends. It is possible that many of the cases detected were produced by the same vehicle passing several times a week. The most common jammer found was of the type shown in fig, 5. It broadcasts on the L1 band with a power of, approximately, 130 mW, or 21 dBm. A jamming chart prepared by the Defense Science Board is shown in Fig. 6. It is seen that a 100 mW jammer against a GPS receiver with a typical 40 dB jam resistance as described above has a range of, approximately, 10 km. There appears to be an assumption built into the chart, too, that a receiver needs an additional margin above the jammer of, approximately, 10 dB.
The chart also gives us an idea as to how much protection is required for shortening the range of a jammer. If one focuses on the upper left of fig. 6, it is seen that for a receiver to function in the presence of a 1 kW jammer at a range of 1 km, the receiver must reduce the effect of the jammer by 100 dB. That is a factor of 10 billion. It was noted above that a standard GPS receiver is designed with, approximately, 46 dB, a factor of, approximately, 40000 to overcome thermal noise. Thus, another factor of 54 dB, or 250,000 is needed. Can it be achieved? Probably!
I believe that most people would agree that a receiver that can tolerate a 1 kW jammer as close as 1 km deserves the label, unjammable.
V. The eLoran alternative
ELoran is based on the World War II loran technology, and it uses the loran pulse shape. It operates at a carrier frequency of 100 kHz in a band, which ranges from 90 to 110 kHz. The figures and specifications used here were taken from the Ursanav website [2]. The Signal consists of a train of eight pules as shown in fig. 1. An individual pulse is shown in fig, 2. The accuracy of the ranging measurement is proportional to the accuracy with which the time of the zero crossing is determined. No doubt the zero crossing was chosen as the reference, because it can be measured more accurately than, say, the beginning time of the pulse. Still, the measurement of the zero crossing to an accuracy of a few nanoseconds at a frequency of 100 kHz is not a trivial task, and the eLoran team deserves to be commended for their success. Note that a slip of one cycle produces an error of 3 km. It should be noted that it is the sixth zero crossing of the signal that is measured. ELoran is sensitive to atmospheric disturbances, such as lightning, and if a disturbance causes a slip to the seventh zero crossing, an error of more than 1 km would occur.
1. eLoran accuracy
It is important to note that eLoran accuracy is quoted in terms of maritime requirements. It is said to have accuracy appropriate for harbor entrance approaches of 20 meters. This value is also said to be adequate for aviation non-precision approaches. There may, however, be some doubt as to the universal accuracy of this statement [5]. Timing accuracy is quoted as ±50 ns. The stability is said to be consistent with the requirement of stratum-1, which is similar to that of an atomic clock on a chip.
A differential error correction system has been developed to reduce errors in specific regions of interest. A test of the system in the United Kingdom (U. K.) in March 2008, demonstrated an accuracy of 12.6 m 95% of the time. An accuracy of 10 m was predicted in the future [3[. It was stated that this is comparable to single frequency GPS accuracy. However, measurements by the FAA at a number of sites gave a single frequency, 95%, accuracy of 3.351 m [4]. (See fig. 3) The U.S. Coast Guard (USCG) operates a GPS differential system in the vicinity of harbor areas, which has an accuracy of 1-3 m. A good summary of eLoran performance is given in ref [6]. GPS values for the parameters of interest will be described below.
2. Support requirements
So, once a signal is supplied from an adequate number of towers and monitor stations are networked to provide corrections, the system is complete. Wrong!! As with any ranging system, distance is measured by determining the time required for a signal to travel from a known location to a receiver. The transit time depends, of course, on the velocity with which the wave travels. The eLoran signal travels along the surface of the earth, and it is well known that the velocity of the ground wave depends on many things, such as the moisture level in the earth and the presence of mountains or manmade structures along the transmission path. The difference between a calculated transmission and the actual time is called the additional secondary factor (ASF). The ASF is not, necessarily, a constant at a given location. Presumably, some variations can be detected by the monitoring stations and transmitted as part of the data message, but, in any case, a map of all areas where the system is to be used must be made by a survey, and the map must be incorporated into the eLoran receiver data. Further, although there is no altitude component associated with the eLoran position measurement, it has been found, as mentioned above [5], that there is a variation of the ASF with altitude that can be quite large, and it is likely that a receiver must be coupled to a precision barometric, or radar, altimeter to certify it for flying, even, non-precision approaches.
Surveying the entire U. S. to generate a nationwide ASF map is, certainly, an activity that will cost tens of millions of dollars. It is likely, that the survey will have to be repeated on a routine basis.
3. Equipping the nation
Even if the eLoran signal is assumed to be a suitable replacement for the GPS if the GPS signal is denied, and a signal covering the entire U. S is supplied, the task of using the signal has just begun. Receivers must be procured. Approximately 2000 will be required to equip several of the major U.S. airline’s aircraft. Too, Crews must be trained in the use of the new navaid. The cost may be in the vicinity of 20 million dollars per airline. Given the near mania that airlines have for return on investment (ROI), the airlines may choose to simply cancel flights as a cheaper alternative. The government will have to choose whether to pass legislation mandating that the airlines equip their aircraft with eLoran equipment, or the government will have to supply the airlines with sufficient funds to buy the equipment.
If it is determined that the signal is, indeed, accurate and reliable enough for non-precision approaches, those approaches will have to be designed and certified by the FAA. For 1000 approaches at a typical cost of $50,000 per approach, that is a charge to the FAA of 50 million dollars. In total, the charge to the nation is likely to be several billion dollars. The question is: who will pay?
4. The British Experience
In the U. K. eLoran signals are available in the vicinity of major ports. This is a reflection of the fact that shipping supports a significant fraction of the British economy. For this application, the poor accuracy of eLoran is not a factor. ELoran is attractive to Europeans, too, because it is not recognized as an American technology. American dominance in the satellite positioning, navigation, and timing (PNT) field was the primary reason for the approval for the Galileo system by European governments. Although there were some claims that it would be an economic bonanza for Europe, the developers would admit in private that the main reason for Galileo was that GPS is American, and “we do not trust the Americans”. Thus, eLoran is not only a backup for GPS in case of jamming of GPS, it also backs up the unreliability of the American government. It is true that the early GPS signals were contaminated with a clock dither known as selective availability (S/A), which limited accuracy of the standard positioning system (SPS) to 100 m. It was established at the urging of the U. S. military to prevent a signal that they considered too accurate from being available to civilians and possible hostile governments. S/A was discontinued in the year 2000, and the IIF satellite family does not have the capability for S/A. Still, suspicion remains. It is rather amusing and ironic, that during the era of S/A, the USCG differential system and the FAA WAAS system were built with the capability of removing the effect of S/A from the GPS signal. Thus, in critical areas, such as Washington, D. C., a low cost addition to a GPS receiver provided accuracy near 1 m.
VI. Alternatives
The eLoran lobby appears to have distracted attention away from the possibility of alternatives, which still operate from land-based transmitters at frequencies lower than and powers higher than GPS. Just as an example, the Hyperfix system was used in Europe and, even, in the Los Angeles area. The system operates on frequencies at 1.6 and 3.4 MHz. Its range is shorter than that of eLoran, but at the 3.4 MHz frequency, the position error is less than 1 m. Presumably, some modifications would provide a timing advantage over eLoran, too.
It is not my intention to be an advocate for Hyperfix, or any other low frequency, land-based system, but it should be realized that alternatives to eLoran exist.
VII. In Congress
A proposed bill to create an eLoran system in the U. S. has been introduced as House Resolution (H.R.) 1678.
Although I have little knowledge of Congressional rules and procedures, my reading of the resolution gave me the impression that it is a rather curious document. Although is ostensibly about PNT, it gives no required accuracy range for any important PNT parameters. The frequency of the system was, however, specified to be 100 kHz. No price range was mentioned, and there appears to have been no consideration of possible alternatives to eLoran, nor does it indicate that any protective steps for GPS were considered. It also, seems to provide no possibility for competitive bidding. To an old lab worker like me, it appears to be a directive to several government agencies to execute a sole-source purchase request for an eLoran system.
VIII. GPS
The GPS characteristics are better known than those of eLoran, and they will only be summarized here. An FAA survey at 28 locations across the U. S. has shown that the accuracy of the single frequency SPS 95% of the Time, or better, is 3.351 m. The measurements were taken at the high data rate of 1 Hz. GPS, unlike eLoran, has a vertical component, whose accuracy was determined to be 4.684 m. A histogram of the horizontal measurement is shown in fig. 3.
A histogram of the vertical measurement is shown in fig, 4. When differential corrections were made by the Wide Area Augmentation System (WAAS), the error was found to range from 0.58 m to 1.2 m, depending on the site.
To determine timing accuracy, I analyzed a data file generated by the U. S. Naval Observatory (USNO), the nation’s timekeeper. For 5778 measurements taken between May 30, 2015 and June 6, 2015, the mean difference between GPS time and the worldwide Universal Coordinated Time (UTC) was 1.44 ns. The median difference was 0.4 ns. The standard deviation was 4.88 ns. Thus for 95.5% of the time, the error was between -8.3 ns and 11.2 ns.
1. The Future
Although the eLoran system has demonstrated its best performance, the technology is at a dead end, and it is unlikely that there will be further improvements. That is not the case for the GPS.
The GPS satellites now being launched, all transmit signals on the L5 channel at 1176 MHz. L5 is an aeronautical band reserved for signals that can affect safety of life operations. The ranging signal consists of a long code which has a pulse rate of 10 MHz, ten times higher than that of the SPS. Two signals, one identified as being in phase, L5I, and the other in quadrature with L5I, and identified as L5Q, are transmitted. When the constellation is fully populated with these signals, it will be possible to reduce the position error to less than one meter. These signals can, also, be used to reduce the vulnerability to jamming and interference.
IX. Jamming of Satellite Signals
Indeed the signal from a GPS satellite is weak compared to the powers that drive our radios or cell phones. This may get a bit too technical for some readers, but I will try to make the main points clear.
First, how weak is the signal?
All received signals must compete with noise, whether natural or artificial, to provide useful information to the receiver. The important quantity is not signal, nor is it noise. The ratio of signal to noise is the important quantity in the end.
First, how does it start out for GPS? Consider the civil signal on the L1 frequency. It has a bandwidth of 2 MHZ and a power in engineering terms that may be as low as -156 dBW. Themal noise, which affects all electrical circuits, and depends on temperature, is distributed over the entire 2 MHz band. A typical value of the temperature gives a noise power of -141 dbW. The noise is, therefore larger than the signal by 15 db. In linear terms, this is a factor of 316. So, the noise is larger than the signal. Is the receiver jammed? Not at all. The receiver locks onto the code transmitted by the satellite and, also, extracts the navigation message, which gives satellite parameters at a frequency of 50 Hz. In order to reduce the effect of the noise, the receiver integrates the received code for a given time. For the 50 Hz data to be visible, the integration time is set at a maximum of 20 ms, corresponding to a frequency of 50 Hz. Thus, the effect of the noise is reduced by a factor of 40,000, which corresponds to 46 dB. Suddenly, a signal-to-noise ratio (S/N) of -15 dB is transformed into a S/N of 46-15=31 dB. Now, the signal is stronger than the noise by a factor of more than 1000. This phenomenon is known as processing gain. It is the mechanism that makes many communication channels, including some cell phones useful, at all. Of course, there may be some residual noise left in the receiver circuits, but, in general, the system works just fine. The problem arises when the interfering signal increases by a factor that puts it above the received GPS signal even in spite of the processing gain. But, now that it is apparent that the concept of processing works, the question is, How far can we extend the concept?
X. Consequences of GPS Loss For Aviation
The effect of GPS loss on cell phone networks, ATM networks, etc. have been described at great length. Suppose we consider, for a moment, the effect of GPS signal loss on the travel industry, particularly airline service. I will assume that the eLoran backup is in place and working as proposed. Assume that eLoran enables non-precision approaches, as promised. A problem is that the non-precision approach has become progressively less important in recent years. The GPS-based WAAS has taken over the approach guidance role at most airports, with the LPV approach giving precision performance at airports that, earlier, had no approach at all. The non-precision approach is almost a relic of antiquity. The effect on airline bad weather operations has been definite.
There are now 3549 WAAS Localizer Performance with Vertical guidance (LPV) approaches at 1731 airports. There is no Instrument Landing System (ILS) backup at 989 of these airports. 1652 runways that have no ILS approach now have WAAS approaches. The term LPV is used by the FAA to distinguish the approach from one enabled by the ILS. The performance is equivalent. ILS equipment is expensive, and it is no longer being installed. Loss of GPS would be a catastrophe for the aviation industry. ELoran is not an adequate replacement by any definition.
XI. What is Needed?
First, it appears that no government agency has done, even, an analysis as simple as the one above, much less an analysis in the detail that the seriousness of the issue demands. Even if eLoran is adopted as a GPS substitute, it will, no doubt, require 20 years before a robust solution to any of these problems will be adopted by a significant number of users, after the expenditure of billions of dollars. Further, the technology will never be capable of a true replacement, even of today’s capabilities. The weapon delivery problem is an example. The enabling of eLoran appears to have been adopted by the U. S. Department of Homeland Security (DHS), because of the USCG connection. I would assume, though, that the task of DHS should be the protection of the nation’s critical infrastructure, not its replacement.
Some satellite modifications have been adopted to make the signal less susceptible to jamming. Examples are M-code, and the focusing antennas planned for GPS IIIc. These are almost trivial when compared to the 100 dB problem described above. Altogether, the satellite improvements amount to, perhaps, 25 dB. The antenna option, however, will only apply to military operations, and only over a very limited area. But, assume that the total improvement is 30 dB. Solution of the problem requires 100 dB. Therefore, the satellite improvements fall short of solving the jamming problem by 70 dB, a factor of ten million.
Rather than adopting the long term, expensive, and inadequate option of eLoran, there is an approach that is likely to be more effective, and less expensive. I divide the important activities into two categories, short term, and longer term.
1. Short term
Much work has been on the detection and location of GPS jammers. A number of companies have developed devices and systems that are very efficient in performing these tasks. See, for example, ref [9].
The DHS should procure a number of these systems and should locate them in critical areas of the Country. Likely locations are at major airports and harbors, and, perhaps, a few could be placed along the interstate highways. The detectors and locators should be attached to a network that reports incidents to a central monitoring location. The government’s cyber security centers is a good possibility. Detection of a jamming event should be reported to first responders in the vicinity of the jamming. The first responders should be trained in the tasks required to locate and disable the jammers and to arrest the operators if they can be identified. Handheld devices carried by first responders may be needed for the final location. Vehicle-mounted sensors for apprehending mobile jammers may also be useful. Systems, training, and networking for these systems could be in place within two years at a cost of less than 20 million dollars. Possibly, it should be made a component of the government’s cyber security program.
2. Longer term
In the longer term, development activities for the unjammable receiver should be well funded. Possibilities exist for creating the device. That is a subject for another paper. Improvement of low cost MEMS inertial systems and atomic clocks on a chip should be an integral part of this program.
XII. Conclusion
When all is considered, protection of the GPS signal is almost certain to be more effective, quicker, and less expensive than attempting to activate a last century PNT system, that will require one, or two decades, at least, to generate even the poor performance advertised for eLoran.
Ultimately, The research and development required to make the GPS invulnerable will benefit the next generation of PNT and will, no doubt, spin off into other fields.
The time to act is passing.
References
[1] House Resolution 1678, 114th Congress, (2005-2006).
[2] eLoran System Definition and Signal Specification Tutorial, A. Helwig, G. Offermans, C. Stout, C. Schue (UrsaNav).
[3] Research & Radionavigation, General Lighthouse Authority of the United Kingdom and Ireland, March, 2008.
[4] FAA Product Team report number 86, July 31, 2008.
[5] G. Johnson, et al, Loran ASF Variations as a Function of Altitude, U. S. Coast Guard Academy, 2006.
[6] C. Hargreaves, ASF Measurement and Processing Techniques, to allow Harbour Navigation at High Accuracy with eLoran. University of Nottingham, September 2010.
[7] Report of the Commission to Assess the Threat to the United States from Electromagnetic Pulse (EMP) Attack, April 2008.
[8] Sentinel Project-Report 001, Chronos Technology, 4 April 2014.
[9] A. Bown, et al, JAMMER AND INTERFERENCE LOCATION SYSTEM. DESIGN AND INITIAL TEST RESULTS, ION GPS, September 1999.
* Dr. Gene H. McCall is an Affiliate Research Professor, Desert Research Institute, Reno Nevada. He completed an assignment as the Chief Scientist with Air Force Space Command at Peterson Air Force Base, Colo. Dr. McCall’s areas of expertise are: Lasers, laser-matter interactions, non-linear optics, nuclear weapon science and technology, Plasma physics, Z-pinch physics. explosive modeling and applications, positioning and timing systems, satellite navigation, aircraft navigation and landing systems, weapon systems. He has now retired from Los Alamos National Laboratory as a Laboratory Fellow and can be reached at ghm7723@gmail.com