Early test results show that LightSquared's planned terrestrial broadband network will, indeed, interfere with GPS receivers. Since the LightSquared spectrum is adjacent to the GPS system, which is widely deployed and widely used but not in exactly the same band, one of the questions being asked is why there is interference to GPS receivers from this proposed system.
The answer is both simple and complex. For this column, I will try to make the explanation easy to understand. Let's start with the fact that GPS receivers are simply that--receivers. They are designed to receive the very weak signals that are transmitted by the GPS satellites. Today there are at least 24 GPS satellites in operation at any given time as well as in-orbit spares in case of a failure. Each satellite is set up for a 12-hour orbit, which means it orbits the earth once every 12 hours and from the ground they appear to be stationary. They are owned and operated by the U.S. Air Force.
GPS receivers provide more than simply a location fix for GPS receivers. They also provide extremely precise timing signals that are used for many things including coordination of hand-offs between cell sites in many wireless networks, validation and secure financial transactions and more. To receive a usable location reading on a GPS receiver, you must be in range of at least three of the satellites. To receive the more advanced readings such at altitude, speed and direction of travel, and other features, the U.S. Air Force recommends line-of-sight to six satellites.
These satellites transmit their signals using 24 MHz of spectrum centered at 1575.42 MHz. These signals are spread-spectrum and are designed to be usable below what is called the Thermal Noise Floor. This is where the explanation can become confusing for many. Noise is inherent all across the radio spectrum and in order to receive a signal, the receiver must hear a signal that is above the Thermal Noise Floor, that is stronger than the noise, or that the receiver can distinguish from Thermal Noise.
Over the years as we add more and more wireless devices on more portions of the spectrum, the noise floor in any given portion of the spectrum rises, meaning that a stronger signal is needed in order for the intelligence it carries to be distinguished from this noise. You might have already experienced one practical way to understand this. If you installed a Wi-Fi access point in the 2.4 GHz band and it had really good coverage, but over time the coverage seemed to deteriorate, it is because so many other people near you have also added Wi-Fi access points, all of which add to the basic noise on the band, making it more difficult for your access point to hear faint signals it might have been able to hear before.
In tests we ran in Anaheim, Calif., a number of years ago on a muni-Wi-Fi system, there were times when we could actually see the street light access point but could not communicate with it. This was simply because there were so many access points in use in that area that the noise floor was raised and the signals were interfering with each other. So, step one is to understand that GPS receivers are designed to listen for satellite signals at or below the thermal noise floor. However, if the noise floor is raised above a certain level, the receivers become deaf; they can no longer identify the satellite signals because of the noise. Therefore, any increase in the noise floor on the spectrum being monitored by a GPS receiver could cause that receiver not to be able to distinguish GPS signals from the noise.
Courtesy of L and S Bands Spectrum Survey in the San Francisco Bay Area by Juyong Do, Dennis M. Akos, and Per K. Enge
Next it is important to understand that every radio signal, even though it operates in a specific portion of the spectrum, introduces noise into adjacent portions of the spectrum. The Federal Communications Commission (FCC) tightly controls the amount of noise that is permissible to bleed over into adjacent spectrum, but it is not possible to eliminate it completely. The more transmit power used, the higher this noise will be in adjacent portions of the spectrum. Notwithstanding the FCC's rules in this regard, it is not possible to prevent some of this noise from increasing the noise floor in adjacent portions of the spectrum. This is one reason today's cellular network operators transmit from cell sites in one portion of the licensed spectrum and from the devices in another portion of the spectrum. This way, any noise that is generated in spectrum controlled by another operator does not end up causing interference problems to the devices of the first operator.
LightSquared intends to install more than 40,000 cell sites using a portion of spectrum in what is referred to as the L-band. Specifically, LightSquared has been authorized (by FCC waiver) to use 1525 through 1559 MHz and 1626.5 to 1660.5 MHz for its system. Since the GPS band is centered at 1575.42 and extends for 12 MHz on either side, the lower limit of the GPS band is 1563.42 and the upper limit is 1587.42 MHz. This means that the lower limit of the GPS band is less than 5 MHz away from the edge of the LightSquared allocation. Because of this, the amount of noise generated by LightSquared into the L-band, especially close to each cell site, could raise the noise floor within the GPS band causing GPS receivers to not be able to receive and/or recognize the GPS signals the receiver is looking for. Indeed, recent tests have shown the noise floor to be raised.
It should be pointed out here that even if LightSquared meets the FCC's requirements for out-of-band emissions, it will still generate enough noise within the GPS band to render GPS devices useless when they are within some distance (to be determined) from each LightSquared cell site. If this portion of the spectrum were used solely for satellite transmissions there would not be a problem since there would only be a few satellite antennas and they would be pointing out into space. With cell sites, of course, the antennas are designed to provide coverage around each cell site at ground level so the problem exists.
If the GPS and LightSquared spectrum were being used for narrowband, signal filters could be used to further minimize the out-of-band emissions below what is permissible by the FCC. However, since both systems are using their spectrum for broadband services, the radio signal is spread out virtually over the entire portion of the spectrum licensed to them, it is far more difficult to fix the problem with filters. Therefore, LightSquared can meet the FCC out-of-band emission limitations and will still cause interference to nearby GPS receivers. There is no way that these out-of-band emissions can be eliminated completely.
This is not the first time out-of-band emissions have been a problem or a potential problem; this is something we have to live with. The difference here is that GPS has become an important part of our lives from military use through finding our way to and from various locations. GPS signals are also used to coordinate both public safety and commercial radios systems and are used to help guide planes to and from airports safely. It is imperative, therefore, that GPS remain a viable service.
There is a high demand for broadband services in the United States and it is no secret that we need more broadband spectrum. However, we should not reassign spectrum that will have a detrimental effect on other services. Perhaps the lesson for all of us, and for the FCC in this case, is to listen to the engineers and run real-world (not laboratory) tests to ensure that the potential for interference to and from other services is fully understood before waivers or spectrum allocations are granted.
Andrew M. Seybold is an authority on technology and trends shaping the world of wireless mobility. A respected analyst, consultant, commentator, author and active participant in industry trade organizations, his views have influenced strategies and shaped initiatives for telecom, mobile computing and wireless industry leaders worldwide. www.andrewseybold.com.