LightSquared's national rollout has shifted frequencies to avoid interfering with GPS kit, taking early advantage of a deal with Inmarsat for 10MHz of bandwidth further down the dial. The wannabe operator, which aspires to cover the USA in an LTE network for wholesale leasing, has also offered to cut its permitted transmission …
Not even bell shaped.
Looks more like trying to scale the Pennines, after you throw harmonics into the mix. Fun all round!
Boy am I glad I'm not a radio engineer having to deal with this.
TX spectrum not the whole story
Even if they had a 'perfect' brick-wall style of transmitter spectrum, it would still be a problem for GPS due to a number of factors:
(1) The effectiveness of the pre-LNA filter and LNA/mixer not to be overloaded by what leaks in from the strong local adjacent channel.
(2) The 'reciprocal mixing' problem with the receiver's local oscillator's phase noise. Here the issue of (1) that is not causing overload of the LNA/mixer still can end up increasing the in-band noise floor by (in effect) shifting the LO's noise there.
(3) The noise floor due to quantization errors in the digitising process (as it *will* be DSP-based after all).
Of course all of the above can be minimised by good design, but everyone wants their GPS to be cheap, small & sensitive for use in poor locations, and those are often mutually exclusive with high adjacent channel protection.
The Moment of Truth
Most wireless engineers could have predicted this outcome. An anlysis by Questiny Group http://www.questinygroup.com/qgi-blog) shows that the new frequency plan of a single carrier of 1526-1536 MHz will significantly reduce the intermodulation products created within the GPS receivers. Howevee, this has a major impact on LS's CAPEX and capacity planing. They will need more towers, and they have to accelerate the move of Inmarsat out of the upper 5 MHz portion of this spectrum.
Lightsquared - a useful technology or just a financial machine?
I'm not exactly convinced that they're actually bringing anything new and practical to the table (with the possible exception of Cell-Sat 'Communicators' ideal for emergency workers, itself A Very Good Thing). But they seem to have swung over to the purely financial purpose of sneaking a new to-be-widely-deployed mobile phone service into the spectrum. The Added Value to humanity vs. the Added Risk to existing services is debatable. Regulators certainly need to keep an Evil Eye on them. I wish them luck, but if they fail then I'm not sure that anyone will notice.
The paragraph that begins, "One generally talks about radio transmissions being square" ranks among the most brilliantly clarifying technical explanations I've ever read. Few readers can appreciate how much effort, skill, and broad intelligence are required to craft sentences like these.