Receiver designers should be prepared for a more complicated spectral environment
What is your perspective about the present controversy involving GPS and LightSquared?
The bottom line is that the LightSquared signals interfere with GPS, and LightSquared should not be allowed to proceed until the interference problems are resolved. To understand this perspective, some background is useful. The spectrum allocated to GPS for its L1 signal is from 1559 MHz to 1591 MHZ. This spectrum is shared by other Global Navigation Satellite Systems such as the EU’s Galileo, China’s Compass, and shortly also by Russia’s GLONASS, which uses the 1598 MHz to 1605 MHz band for its present signals. Basically, GNSS signals occupy the 1559 MHz – 1605 MHz band.
The Mobile Satellite Services (MSS) spectrum is from 1525 MHz to 1559 MHz, directly below the GPS L1 band. LightSquared has two 10 MHz bands within this spectrum, called Low 10 (1525 MHz – 1535 MHz) and High 10 (1545 MHz – 1555 MHz). They also have paired 10 MHz bands above the GPS and GLONASS frequencies at 1627 MHz – 1637 MHz and 1646 MHz – 1656 MHz. The two lower 10 MHz bands would be used for base stations (cell towers) and the upper two 10 MHz frequencies for LightSquared handsets or mobile devices.
The MSS band has for a long time been allocated by FCC rules and international agreements to low powered satellite-to-earth services. The GPS signals in the adjacent band are also low powered satellite-to-earth signals, so these bands were compatible, and GPS receivers operated properly without interference from devices in the MSS band. However, when LightSquared got conditional approval from the FCC in January 2011 to use Low 10 and High 10 for high powered terrestrial signals, it became clear that this use would interfere with GPS, and would degrade or deny GPS when a GPS receiver was near a LightSquared base station. The range of the interference depends on the type of GPS receiver, with some being more sensitive than others, but all classes of GPS receivers would be seriously affected, some at quite long ranges (15 miles or more). With a plan to roll out 40,000 base stations across the US, it became clear that LightSquared represented an extremely serious challenge to GPS usage, civilian, industrial, and military.
LightSquared is a complex issue with political, technical, and financial components. To address these one by one:
There is little question that the United States needs more wireless broadband service. Many areas are underserved, and the existing services are limited in speed and capacity. In addition, a very large proportion of these services are provided by only two companies, ATT and Verizon. So the entry of a new company into this marketplace should be welcomed if it can offer nationwide coverage, high speeds, lower costs, and competition to ATT and Verizon.
LightSquared’s desires as a private company are in concert with the National Broadband Plan, established by the Obama Administration, which seeks 500 MHz of spectrum to allocate to broadband over 10 years. However, there is essentially no technically appropriate spectrum that does not have existing uses, so this is not a simple undertaking, and little has been accomplished in the first two years of this plan. LightSquared, should it be allowed to proceed, would represent a small but significant 40 MHz down payment on the 500 MHz.
The political concerns might readily be resolved if they were the only problems. However, the interference with GPS is quite severe. Since GPS is an integral part of the national and global infrastructure, and deeply embedded into many everyday applications, permitting the rollout of a system that interferes with GPS is very problematic. In its grant of conditional approval to LightSquared to proceed, one of the conditions imposed by the FCC was that the interference with GPS be resolved (there is no definition of “resolved”, so it means what the FCC chooses to have it mean). This has resulted in extensive interference testing and studies by many parties, including the DOD, FAA, NASA, GNSS companies, LightSquared, and others. The most significant test was conducted by the GPS Technical Working Group (TWG), which was established pursuant to the FCC’s requirement to test GPS interference. The TWG results were released at the end of June 2011, and showed massive LightSquared interference with essentially all classes of GPS receivers.
The TWG testing showed multiple types of technical problems. The definite problems can be classified as a) overload, b) augmentation, and c) handsets. The potential problems can be classified as d) third order intermodulation and e) handset out-of-band-emissions (OOBE).
– this problem stems from the closeness in frequency of the very high powered LightSquared signals to the GPS band. When a GPS receiver is near a LightSquared base station, it cannot filter out this level of power, and the LightSquared signal overloads the RF front end of the receiver.
– many GPS receivers use augmentation signals to increase the accuracy of their navigation results. These augmentation signals are typically delivered by satellites using frequencies in the MSS band. The presence of high power terrestrial LightSquared signals in this band prevents the reception of the low power augmentation signals, denying the use of these augmentation signals when near a LightSquared base station.
– the signals emitted by LightSquared handsets in their return bands above GLONASS interfere with GPS receivers, just as do those emitted by LightSquared base stations. Because the handset signals are of much lower power than the base station signals, the range of the interference is much less, on the order of a meter or two.
Third Order Intermodulation
– if LightSquared uses both the Low 10 and High 10 bands, they will create interfering signals that fall directly into the GPS L1 band and cannot be filtered out. If only one of the two bands is used, there are no intermodulation effects.
– while we know that LightSquared handsets interfere with GPS due to their communication in the return signal bands above GLONASS, the handsets are permitted under present FCC rules to emit out-of-bandinterference into the GPS L1 band that is at a level high enough to seriously interfere with GPS (if they do in fact emit at the permitted levels).
Evaluating the interference effects has become more confusing as LightSquared, faced with clear evidence of serious interference to GPS, has altered their system design at least twice. The first proposed alteration was to use only Low 10 for now, with High 10 postponed (but only postponed, not abandoned). Since High 10 is closer in frequency to GPS and causes more interference than Low 10, this was a positive but insufficient step, as Low 10 still interferes with GPS to an unacceptable level. The second proposed alteration was to use only Low 10 (again, just for now) and limit aggregate power on the ground to a lower level than would otherwise be the case. This is also a welcome but also insufficient step, as interference remains, just at reduced ranges.
One of the results of studying the LightSquared interference has been to make it clear that we cannot build GPS receivers that can deal with High 10 and still function properly. The High 10 power is simply too close to filter out without severely degrading performance and greatly increasing the size and weight of receivers. Should LightSquared be allowed to eventually proceed, it must be with Low 10 alone.
While it is also clear that many existing receivers are impacted by any of the LightSquared alternatives, particularly high precision receivers used in many high value applications, whether we can build receivers that can deal with Low 10 alone without performance effects remains unsettled. We can adequately filter out Low 10, but whether this filtering will cause undesirable performance results remains to be determined.
LightSquared is largely owned by a hedge fund (Harbinger Capital Partners) that has taken a huge gamble in buying a satellite company and renaming it LightSquared. They have invested heavily in LightSquared, presumably assuming that the spectrum could be converted from satellite to terrestrial spectrum.
However, there is a fairness concern with LightSquared’s approach; ATT and Verizon paid billions of dollars to the US Government at auction for their terrestrial frequencies, while LightSquared, by virtue of converting satellite frequencies to terrestrial frequencies, would avoid those costs and the US Government would receive no compensation for licensing the LightSquared frequencies. The value of the LightSquared frequencies increases enormously if the spectrum is terrestrial rather than satellite spectrum (one estimate is from $2B to $10B).
The other major financial issue concerns who should pay for the remediation required if LightSquared is allowed to proceed. Some existing GPS receivers and/or antennas will have to be replaced with LightSquared resistant models (presuming such can be built without degrading performance), and while the number of replacements strongly depends on how long a transition period is allowed, it is clear that some will have to be replaced.
The cost to replace receivers and/or antennas also depends very strongly on the application in which they are used. Replacing military and aviation equipment would be very expensive. Replacing survey, construction, agriculture, and scientific equipment would also be expensive. The cost to replace other types of receivers would be less. Who should bear this cost? The FCC precedent is that a new user that impacts existing users is responsible for the costs of remediation.
In summary, LightSquared should not be permitted to proceed until we know how GPS interference will be managed. If they are permitted to proceed, it must be under the following conditions:
• Low 10 alone and not High 10 (now or ever)
• Enough time must be allowed for the transition from existing receivers to LightSquared resistant receivers
• LightSquared should pay for the remediation that will be required
How did the whole thing go wrong?
The problem first arose when conditional approval for LightSquared was given without a full understanding of the interference problems that would result.
What the FCC is proposing to do with the MSS spectrum is re-purpose at least part of it for terrestrial use instead of satellite use. The appropriate process for this is called Notice of Proposed Rulemaking. This process would have involved extensive public comment, and the GPS interference problem would have become apparent at an earlier stage in the proceedings.
The second major problem arose because LightSquared either did not understand or ignored the extremely serious effects that their system would have on GPS. In failing to understand this problem, and insisting repeatedly that there would not be interference, or that the interference was minimal, or that it affected only a tiny minority of GPS receivers, LightSquared created a crisis in which they and government and industry and users have had to invest major resources in seeking a resolution. While it might not be consistent with LightSquared commitments to their potential business partners or investors, the process of evaluating the re-purposing of MSS spectrum should have played out over a much longer time frame.
Do you foresee any solution to the existing deadlock?
At some point, the FCC will rule further on the GPS interference issue. There is no required timeframe for a decision, and given that the National Telecommunications Information Administration (NTIA), the regulator of government spectrum, has asked for further testing of Low 10 effects on various classes of GPS receivers, we don’t expect this ruling to occur until possibly sometime in the first quarter of 2012, at the earliest.
We believe the FCC would very much like to give LightSquared Low 10, if there is a way to resolve the GPS interference issue. Note that resolve does not, to the FCC, necessarily mean that there is no interference. We believe the FCC will be inclined to allow some interference to occur – it is a question of whether the results are tolerable or intolerable. The federal agencies, particularly DOD and FAA, may play a big role in the FCC decision.
The bottom line is that it is not clear what the FCC will decide, but it would be prudent for receiver designers, even if LightSquared does not get a go-ahead, to prepare for a more complicated spectral environment in the future.
How do you see some of the suggestions made to address this issue?
A lot of suggestions have been made, some fanciful, some potentially useful. Among the potentially useful ones, from our perspective:
• Low 10 only, never High 10
• LightSquared pays for remediation for impacted existing receivers
• The time frame for receiver replacement or update is 5-10 years
• LightSquared power levels are reduced • Augmentation signals are placed tightly against the GPS band (so that filters that include GPS can include the augmentation signals)
• The FCC OOBE limits for handsets are lowered
If FCC decides contrary to the expectations of GPS community, what is the way forward?
It seems unlikely that the FCC would authorize the use of High 10; the interference to all classes of GPS receivers is massive, and there is no way technically to filter it out without causing severe performance problems. If the FCC authorizes the use of Low 10, the responses of manufacturers and users will be to upgrade and replace receivers, obtaining the best performance that we can in this environment. If this is what the FCC decides, there will still be problems, and we will have to deal with them in the context of the FCC decision. The time frame and the allocation of costs will be the most critical issues to address.
Any lessons for GPS users abroad from this controversy?
GPS and GNSS users here and abroad should be quite concerned about this controversy, for several reasons.
• It affects the use of all GNSS in the US, not just GPS. The benefits of multiple GNSS in the US will be lost along with GPS when receivers are near LightSquared base stations.
• LightSquared would like to have a global system, not just a US system. If they can succeed in the US, it greatly strengthens their position abroad.
• Spectrum is dear; the LightSquared case is just one example of how the pressure for intensified use of the available spectrum may affect navigation. We should expect other instances of pressure on GNSS spectrum.
• We must be careful about the cumulative effects of electronic systems that raise the noise floor. While GNSS may be able to tolerate some systems, it may not be able to tolerate all such systems.
• We should strengthen the filtering and processing in GNSS receivers to prepare for an environment in the future that is not as friendly to GNSS as at present.