The dangers of GPS/GNSS

The following discussion for obvious reasons mostly refers to GPS, but the arguments are generally valid for all global navigation satellite-based systems.

Actual performance

Today’s average performance of GPS is used as a starting point for our discussion. In not too distant a future, even better numbers can be expected.

PDOP Availability: Requirement – PDOP of 6 or less, 98% of the time or better; Actual – 99.98%.

Horizontal Service Availability: Requirement – 95% threshold of 36 metres, 99% of the time or better; Actual – 3.7 metres.

Vertical Service Availability: Requirement – 95% threshold of 77 metres, 99% of the time or better; Actual – 5.3 metres.

User Range Error: Requirement – 6 metres or less; Constellation Average Actual – 1.2 metres.

Fig. 1. GPS users in 2006

What’s the problem?

The problem is that nothing works 100 %. GPS is very close, but for some users under some circumstances,“very close” is not good enough. The situation in general is as follows:

• Most GPS users know nothing about GPS vulnerability.

• Most users don’t care.

• Most GPS users can stand some interruptions or performance reduction.

• Most politicians and representatives of authorities in the fi eld of navigation don’t know of GPS vulnerability.

• Back-up systems are being closed down (e.g. LORAN-C), and there is little or no contact between different countries about these matters.

GPS (and all satellite navigation systems, more or less) are vulnerabe because of

• Very low signal power received;

• A few frequencies (in the GPS case today, only one for general use)and a known signal structure;

• Spectrum competition;

• Worldwide military applications drive a GPS disruption industry; • Jamming techniques are well known, devices are available, or can be built easily (fig. 4).

In 2001 (just before the infamous 9/11), the U.S. Department of Transportation’s Volpe National Transportation Systems Center published results from an investigation into the vulnerability of the transportation infrastructure relying on GPS. Conclusions to be drawn from that investigation are:

• Awareness should be created in the navigation and timing communities of the need for back-up systems or operational procedures;

• All transportation modes should be encouraged to pay attention to autonomous integrity monitoring of GPS/GNSS signals;

• All GPS/GNSS receivers in critical applications must provide a timely warning when the signals are degraded or lost;

• Development of certifiable, integrated (multipurpose) receivers should be encouraged;

• A comprehensive analysis of GPS/ GNSS back-up navigation and precise timing options (e.g. LORAN, VOR/DME, ILS, INS) and operating

procedures should be conducted.

Fig. 2. Experienced and expected use of GPS/GNSS in cars

Fig. 3. Experienced and expected use of GPS/GNSS in mobile phones

Fig. 4. This dice is a 10 mW GPS jammer.

Fig. 5. Example of a car navigation problem.

Causes of trouble

There are many possible reasons for degraded performance or service

interruption for users of GNSS:

• Satellite or controlsegment malfunctions.

• Unintentional interference:

• Radio-frequency interference (RFI) from external sources (spectrum congestion, harmonics, high-power signals saturating receiver front

ends);

• Testing at system level;

• Ionospheric infl uence (solar maxima, magnetic storms, scintillations);

• Multipath.

• Intentional interference:

• Jamming;

• Spoofing (false signals into the receiver);

• Meaconing (interception and re-broadcast of navigation signals).

• Human factors:

• User equipment and satellite design errors;

• Over-reliance;

• Lack of knowledge and/or training.

The main technical explanation of GNSS receiver vulnerabilities to external interference can be summarised as very low power received from the satellites. The minimum power level is usually between -150 and -160 dBW. Looking at the details for GPS, we see that

• acquisition requires 6 – 10 dB higher signal-to-noise ratio (SNR) than

tracking,

• loss-of-lock sometimes occurs for interference-to-signal ratios (I/S) below 30 dB,

• receiver detection of loss-of-lock is delayed because of narrowband code-tracking loops,

• some lines in the GPS C/A-code spectrum are more vulnerable than others because of higher power levels (Gold code spectra do not exactly follow a sinc shape, and spectral lines work as local oscillator frequencies for received interference signals),

• modulated interference is generally worse than white noise, and narrowband interference is worse than wide-band.

Recorded examples of GPS troubles

During the long (and ― don’t forget! ― usually very successful) history of GPS, a number of satellite malfunctions and interference problems have been recorded. Taking the record of April – August of the year 2005 for examples of satellite malfunctions, we fi nd the following:

• SVN37 (PRN7): 3 Apr – Load-shed;

• SVN31 (PRN31): 14 Apr – Baseband reset;

• SVN27 (PRN27): 14 May – Rubidium #1 runoff leads to clock swap;

• SVN26 (PRN26): 9 Jun – Rubidium #1 clock jump;

• SVN15 (PRN15): 22 Jun – Comparator Reference Value Change;

• SVN32 (PRN1): 24 Jul – Load-shed;

• SVN26 (PRN26): 21 Aug – Crypto Variable Upload.

Experienced jamming and other records of intentional interference are (for obvious reasons) less available, but unintentional interference examples are abundant. Let us look at just two of them.

An infamous example is Moss Landing in California. From May 2001 and several months onward, no use of commercial GPS receivers was possible in the whole harbour area out to a distance at sea of at least one kilometer. After the first user reports about GPS unavailability, considerable efforts were launched to find the source of the interfering signal(s). Finally, it was discovered that there were in fact three sources, all of them being active UHF/VHF TV antennas with preamplifi ers onboard pleasure boats.

In December the same year, a GPS jammer caused GPS failures within 180 nautical miles of Mesa, Arizona. Boeing was preparing for upcoming tests and accidentally left a jammer on the L1 frequency, radiating just 0.8 mW. The jammer operated continuously for 4.5 days. There were several impacts to ATC operations during the six days of jamming:

• Aircraft lost GPS 45 nm from Phoenix, performed a 35° turn toward

traffic;

• NOTAM was not issued until 2nd day;

• numerous pilots reported loss of GPS;

• There were reports of handheld GPS receivers not working.

Time users

Users of GPS as a time and/or frequency reference are an often forgotten or unknown but very important part of the GPS community. Some applications where GPS time is used:

• Navigation;

• Telecommunications;

• Digital broadcasting;

• Power generation and distribution;

• Metrology;

• Meteorology;

• Radar;

• Tests and measurements

• Time tagging (Internet and transport)

• Time-of-day distribution.

Many users of these applications are crucially dependent on GPS for

their systems to work properly.

Countermeasures

If satellite signals do not meet requirements, the only thing users can do is to acquire information about the malfunction(s) from other sources as quickly as possible. Such information acquisition is called integrity monitoring. It can be receiver autonomous (RAIM) or received from external wide-area augmentation sources (e.g. WAAS, EGNOS, MSAS, GAGAN) or from local monitors (GBAS = Ground- Based Augmentation Systems).

Integration of GNSS receiving equipment with other sensors (e.g. inertial systems, LORAN) can be very useful in case of satellite signal malfunctioning. Such integration can also be efficient against interference and jamming.

Other countermeasures against all kinds of external interference are fi ltering and advanced signal processing, including adaptive antennas and null-steering. A question which is often asked is whether upcoming systems (Galileo, INRSS, etc.) will solve the problem.

The answer is that they will reduce the problem but not solve it completely.

Satellite navigation problems in cars

Drivers using car navigation equipment experience problems fairly often (Fig. 5). Due to lack of knowledge, these problems are often wrongly ascribed to the satellite system instead of to the real cause, the map system.

(In fact, a lot of users call their car navigator “my GPS”). A summary of these problems might be written:

• Satellite-based car-navigation equipment attracts negative attention for doing a job logically… but unintelligently;

• Expectations for such equipment are high – expected to be smarter than the driver (which-in fact-it often is!);

• Road classifi cation is diffi cult – one person’s farm track is another’s handy shortcut.

Conclusions and recommendations

All GNSS users must evaluate and analyse their own situation:

What would be the consequences in case of GNSS problems?

If the answer is ‘no serious consequences’, then “business as usual”. But if there are possible serious consequences, users must prepare for the unexpected !