Tech Insights: LightSquared’s Technical Woes
By now most wireless industry insiders have become aware of the very bumpy road that LightSquared has traveled in attempting to realize its vision for a nationwide “wholesale only” broadband LTE network.
With immediate access to a big chunk of L-band spectrum (1525-1559 MHz) across the U. S., LightSquared hopes to become a major player in the wireless broadband sector by giving a variety of commercial service providers a means to offer broadband services and features without the enormous cost and complexity of building their own networks.
In a nutshell, the LightSquared plan is to build out a network with ubiquitous nationwide coverage using a combination of satellites and terrestrial base stations, then offer capacity on that network to entities such as smaller regional wireless operators, cable service providers, content providers and others. These companies might then offer the service to retail consumers in the form of general broadband Internet access (pretty much like an MVNO) or to support dedicated services like content delivery or features, such as asset tracking, that rely on machine-to-machine communications.
On the surface, the LightSquared business model isn’t radically different from that of Clearwire, which essentially provides wholesale broadband service to Sprint. Clearwire has experienced many financial and technical difficulties, but it does operate functioning networks with a substantial coverage footprint – at least in terms of POPs. But despite deep-pockets financial backing and seemingly bountiful spectrum resources, it’s anybody’s guess as to when and if LightSquared will ever come close to its stated goals.
Broadband via Satellite Limitations
Why is LightSquared facing such difficulties? Some may argue that there are deep flaws in its wholesale business strategy, but I believe that its fundamental technology problems are far more serious. One of these problems, which has received a good deal of publicity, relates to the potential of terrestrial LightSquared base stations interfering with operation of Global Positioning System (GPS) receivers. A more fundamental difficulty is impracticality of providing mobile broadband services via satellite except in very limited circumstances. And as a further complication, even without the interference issues, L-band spectrum is far from ideal for providing broad geographic coverage using terrestrial networks.
Internet service via satellite is not technically unworkable. Hughesnet, for example, offers a service with “up to” downlink speeds in the range of 1-2 Mbps through geostationary satellites. The big difference is that the Hughesnet service, which is intended for rural customers, relies on fixed user terminals equipped with relatively high gain dish antennas. This allows for high modulation efficiency, so that the data-carrying capacity of available spectrum is far greater than can be achieved using mobile technologies such as LTE.
But even with this advantage, individual throughput speeds provided by Hughesnet usually drop dramatically during hours of heavy demand. This is because the “beams” defined by the Hughesnet satellite antennas are so large that each serves a vast number of users. As with any common carrier wireless network, available capacity in each sector – in this case a satellite beam – has to be shared by the users therein.
The LightSquared plan is to use satellites that provide more focused, and thus smaller, “spot beams.” But even these are gargantuan in geographic coverage compared to the largest cells in a terrestrial network. Since the number of users that each spot beam can support with reasonable “broadband” speeds is limited to perhaps a few thousand, satellites are obviously impractical for providing LightSquared service to urban and suburban areas.
Of course, LightSquared could have limited its focus to providing Internet access to the most rural areas – with perhaps 1 percent of the U. S. population – where satellite-based service is arguably practical, but that’s not a particularly lucrative business opportunity. To provide the capacity required to serve the other 99 percent, a vast terrestrial network also will be required.
And because propagation of the 1.5 GHz L-band is much more constrained than some bands used for mobile networks, for example 700 MHz, that terrestrial network will require a staggering number of high-power base stations to provide enough coverage so that the very limited capacity of the satellite service can serve the remaining POPs. The notion that the planned service can rely primarily on satellites with a few terrestrial base stations in and around large cities, as once promoted by LightSquared’s predecessor company, is simply out of tune with practical reality.
Why the GPS Interference?
That brings us to the problem of interference with GPS receivers. In order to understand why this is such a major hurdle for LightSquared, a little background information of GPS technology is helpful.
GPS relies on accurately receiving digital signals that are simultaneously transmitted, at relatively modest power, by multiple satellites from among a constellation in low earth orbit. All GPS satellites transmit on the same frequency, 1,575.42 MHz, but because of Doppler shift (caused by the very high speed of the satellites with some potential contribution from motion of the GPS receiver), the signals as observed by the receiver can be frequency shifted by up to about ± 6 KHz.
Because of this Doppler shift, “unassisted” GPS receivers in a “cold start” have to “hunt for” signals being transmitted by the various satellites in view. This is complicated by the fact that on or near the Earth’s surface, the GPS signals are very weak – on the order of -130 dBm for receivers with the commonly used 0 dB gain antenna. Obviously, GPS receivers must have very high sensitivity, a good portion of which is typically enabled by the assumption that there will be no vastly stronger signals in the frequency bands surrounding the GPS signals. This allows for the use of very low-noise preamplifiers with minimal dynamic range and eliminates the need for narrowband RF prefilters that would incur significant signal losses.
Until LightSquared made plans, and initially received FCC approval, for terrestrial base stations, the assumption of GPS safety from much stronger near-band signals was reasonable because the surrounding L-band was dedicated to various satellite communications applications. Even the most powerful satellite transmissions, by the time they reach the ground, are not so much stronger than GPS signals that they will cause receiver front-end saturation or intermodulation interference, the two potential problems whereby much stronger near-band signals can drastically reduce receiver sensitivity. But putting powerful L-band transmitters on or near the ground changes things dramatically.
As previously noted, without its terrestrial network, LightSquared has a greatly reduced business opportunity. Understandably, the company has tried to demonstrate that with some modifications and constraints, terrestrial L-band base stations will not significantly impact GPS operations. The most important of these is limiting high power transmissions to the portion of LightSquared’s spectrum farthest from the GPS operational band, where the minimal prefiltering in GPS receivers will provide a bit of attenuation.
LightSquared also has suggested that it somewhat reduce maximum base station transmit power. The company now states that with these and other constraints, the problems with GPS receivers have been effectively resolved. The GPS industry strongly disagrees, and as things stand now, GPS likely will prevail.
It’s not just that GPS service enjoys the benefit of incumbency, or that interference from terrestrial LightSquared base stations could impact the reliability of location and navigation capabilities used by tens of millions of consumers in the U.S. alone. GPS operation is also fundamental to many critical U.S. military systems, and it forms the basis of the “next generation” air traffic system currently in development. Already, GPS provides sole guidance for landing aircraft to within 200 feet of the ground – not a good time to experience loss of service due to interference. If nothing else, this illustrates that GPS interests are not the ones you want to be fighting if you are trying to launch a new wireless service offering.
Drucker is president of Drucker Associates. He may be contacted at firstname.lastname@example.org.