Experiment:
Analysis of RF Interference on GPS from a Commercial Television Station Broadcasting Antenna
Back in 1994 a presentation was made to the Institute Of Navigation (ION) GPS Test Standards Working Group that detailed a real work example of interference caused by transmitting television station antenna. The antenna, located near Hudson, MA, transmitted CH 66 with an effective radiated power of 2,000 kW. The goal of this experiment is to verify the aforementioned experiment using a different television transmitter.
For this experiment, an antenna located near San Antonio, TX, transmitting CH 23 at an ERP of 1480 kW will be used. The FCC allocates 6 MHz of bandwidth for each television station, with the frequency allocation of channel 23 being 524 to 530 MHz. While this is well outside the GPS L1 C/A carrier frequency of 1575.42 MHz, the 3rd harmonic produced by the station will be between the frequencies of 1572 to 1590 MHz. Under FCC guidelines, these spurious emissions must be attenuated by at least 60 dB so as not to interfere with other signals. Therefore, the maximum harmonic emissions are reduced by at least a factor of one million, but there can still be up to 1.48 watts of 3rd harmonic power on the L1 band. Figure 1 shows the diagram of the link budget for a television station antenna, and the following equations are used to determine the jamming radius around the antenna where a loss of lock will occur.
Figure 1 Link Budget Diagram
The link budget formula is
(dBw) (1)
where
ERPj = Effective radiated power of the jamming antenna
= Jt +Gt
Jt = Transmitter antenna input power (dBw)
= 10log(jt)
jt = Power of the harmonic frequency (watts)
Gt = Transmitter antenna gain (dBic)
Jr = Received jammer power (dBw)
Gj = Receiver antenna gain (dBic)
Lp = Atmospheric propagation loss (dB)
=
d = Distance to transmitting antenna (meters)
lj = Wavelength of jammer frequency (meters)
Lf = Power loss due to receiver front end filtering (dB)
We will start by assuming that the EPRj of the transmitting antenna is 1.48W (1.7dBw). There is no way to know what the actual number is so this will give a maximum value for the jamming distance. The next part is to determine what is the threshold level of the receiver. This part is an estimate since the manufactures do not give out the actual specifications of their tracking loops. What is used is an approximation of a very good tracking loop, where 44.6 is the performance (Jamming to Signal) of a third-order PLL with a 2Hz noise bandwidth and a 20msec predetection integration time. The minimum guaranteed received GPS signal is SR = -159.6 dBw, so the jamming power needed at the receiver threshold is
dBw
Once that is known, we can then find the line of sight distance to the transmitting television antenna where the power level will be at the jamming threshold. Since the jamming frequency is assumed to be in-band interference, the power loss due to front end filtering (Lf) is assumed to be zero. Also, the receiver antenna gain (Gj) is assumed to be –3.4 dBic which includes the loss due to differences in the right hand circularly polarized receiver antenna, and the horizontally polarized signal . Using Equation (1) and rearranging we have
dB
Solving for the distance from the receiver to the jammer to attenuate the signal to the threshold limit
D = 7.0 km
= 4.3 mi
Test Results and Analysis
An experiment was conducted to verify these results. This was done by taking a handheld GPS unit and driving on the roads around the transmitting antenna. The receiver that was used was a Garmin GPS III handheld (Figure 2). It was chosen because since it is an inexpensive receiver, hopefully the tracking loop would not be as good as the one used in the estimate equations, and there would be a better
Figure 2 The Garmin GPS III
possibility of loosing lock on the satellites. However this was not the case. Not only was there no loss of lock anywhere around the antenna, but also there was no appreciable loss of signal strength that could be attributed to the transmitting antenna. This happened even though there was a straight road that came to within 0.5 miles of the antenna, with an unobstructed view of the antenna at all times. There was some drop in the signal strength of the satellites once on the access road, but that could be contributed to the increase in the foliage cover over the road.
So, what happened? There should have been some noticeable interference effects in the region surrounding the antenna, but nothing happened. There was something wrong with the assumptions of the receiver performance, or the transmitting antenna harmonics. It seems more likely that the problem lies with the uncertainties in the television station antenna than in the performance of the handheld receiver. According to Kaplan (4), having a J/S ratio of 44.6 dB is an optimistic assessment of the performance of a GPS receiver. Given that, the transmitting antenna needs to be looked at more carefully in order to solve this problem.
First off, the maximum power level of the 3rd harmonic given a specific level of attenuation will not be equal to the EPR of the antenna times the attenuation level. Not all of the energy that goes into the fundamental frequency would have made it into the higher order harmonics. The 2nd harmonic will have more energy than the 3rd, which, in turn, will have more than the 4th. In addition, after the presentation of the original findings from the channel 66 antenna in Massachusetts, both the Department of Transportation and the National Telecommunications and Information Administration conducted tests to verify that television stations were attenuating their higher order harmonics. The findings were that in all of the cases, the attenuation levels were below –100dB. That would mean that since the ERP of the station is 148 kW the 3rd harmonic would have at most 0.000148 watts of power.
Since there was no appreciable degradation of performance around the receiver within this area, there has to be another explanation. By looking at the transmitting antenna pattern, we can possibly explain this. In this case, the transmitting antenna is not an isotropic one. Since the antenna is located outside of a metropolitan area, a directional antenna is used to maximize the antenna gain towards the populated areas. A map with the relative field values of this antenna is shown in Figure 3. This is the relative field effect of the antenna that does not include the effects of the tower. A plot of the actual field values with the effects of the tower was unavailable, but would look similar to the one in Figure 3 but with some local variances due to the tower effects.
Figure 3 Relative Field Strength Values for the CH 23 Transmitter Located Northwest of San Antonio, TX
In reality, the transmitting antenna is actually made up of multiple antennas stacked one on top of the other. The RF feed system to each individual element, sometimes referred to as a bay, ensures that the proper amplitude and phase is supplied so as to form the beam in the required direction. Since these bays are usually located thousands of feet above the ground, the beam is usually “tilted” so that the centerline of the radiation pattern points towards the ground. Figure 4 shows two antennas, one with beam tilt and one without.
Figure 4 Comparison of the Effect of Beam Tilt on the Radiation Pattern of a Transmitting Television Station Antenna
Because the antenna is made up of a number of transmitting elements, their phases will add constructively in one direction, while in another direction they will add destructively causing poor to no signal strength in those directions. The null spots are created at elevation angles lower than the center of radiation angle. This mean that in the areas immediately surrounding the antenna, the field strength is not known.