Senators Take Aim at LightSquared's GPS Problem

CellStudent said:
Everything I've read so far indicates that the problem isn't that
LightSquared will be broadcasting in the GPS bands, but that for the last 10
years, GPS makers have banked on the fact that no one was using the L-band
spectrum as an excuse to make really cheap, shoddy GPS radios that
actually interpret L-band signals as legitimate signals, even though they're
not.

Now, because they've deployed 100's of millions of shoddy GPS receivers to the
public, they want the LightSquared L-band edge declared a "guard band" because
they don't want to step up and take accountability for 15 years of building
crappy, cheap GPS antennas!

CellStudent, I respect your knowled...
(continues)

WiWavelength said:


CellStudent, I respect your knowledge, but you are off base on this issue. The crux is that LightSquared intends to use its L band MSS spectrum for both satellite & terrestrial deployment -- emphasis on the terrestrial use. But the L band MSS spectrum was not originally licensed for terrestrial use, and GPS receiver designs took this into consideration, apparently, by using less aggressive filtration that allows for greater receiver sensitivity.

As I am sure you know, an ideal receiver filter rejects completely all out of band
energy and passes unaltered all passband energy. But such an ideal filter does not exist. Rather, real world filter design almost always involves compromise. You can

...
(continues)

Overmann said:
Soundproofing for houses does exist.

Did you really need to lazily quote my entire post just to ignore or miss my point?

AJ

AJ,

The FCC has had experience with Northpoint/MVDDS, UWB, AWS-3/M2Z, and PCS H block. I very curious on how they will handle this issue as well.

The 34 member committee appointed to study the lightsquared/GPS problem, has to report the 15th of every month until June 15th when they are to present their findings. My opinion is that no reasonable compromise will be made and Lightsquared will possibly fade away. There is a fear is that the FCC could force everyone who owns a GPS, to upgrade although I highly doubt this would happen.

But if you don't mind, I would like your knowledgeable input and opinion on this.

Thanks,

John B.

WiWavelength said:
But the L band MSS spectrum was not originally licensed for terrestrial use, and GPS receiver designs took this into consideration, apparently, by using less aggressive filtration that allows for greater receiver sensitivity.

As I am sure you know, an ideal receiver filter rejects completely all out of band
energy and passes unaltered all passband energy. But such an ideal filter does not exist. Rather, real world filter design almost always involves compromise. You can design a receiver filter to block effectively all out of band signal, but it will also attenuate some passband signal, thereby reducing receiver sensitivity. Or you can design a receiver filter to pass all passband signal, there

...
(continues)

CellStudent said:
Okay- clarify this one for me.

In the PCS band (1900 MHz) I believe the necessary guard bands implemented between blocks is less than 2 MHz. Actually, I think it's less than 1 MHz, but I'm not going to take the time to dig up the spec right now.

I apologize that I do not have time right now to address all of your points. So, I may have to do this piece by piece.

First, PCS 1900 MHz regulations do stipulate out of band emissions limits but not specific guard bands. That said, most PCS licensees do implement guard bands of half channel bandwidth at the upper & lower frequency extremes of both uplink & downlink. For example, CDMA 1900 networks typically utilize 625 kHz bandwidth...
(continues)

Thanks for the quick point. I look forward to a more detailed answer when you lifestyle allows it.

I just spent a little quality time in MATLAB (fdatool utility) and designed an IIR filter with unity gain at the GPS frequency (1574 - 1577 MHz) and -100 dB attenuation at the 1555 MHz stopband edge.

It's only a 10th order filter which doesn't seem to hard to implement in practicality in the real world.

I used a Sampling Frequency of 5000 because I wasn't sure what a realistic value for F(s) should be. What do you think would be reasonable for that value?