Timeline for Is possible to build -168 dBm amplifer?
Current License: CC BY-SA 3.0
12 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 23, 2014 at 1:03 | comment | added | Chris Stratton | The implication of the gps performance claim and the physical noise floor is that the bandwidth (of the correlation) is on the order of a Hz or two. But then, all one is trying to do at that level is phase lock a model to reality over a substantial evaluation time - ie, look for small errors. If the model is more than slightly off, a time-consuming search will ensue. | |
| Oct 23, 2014 at 0:06 | comment | added | George Herold | Fun, If what @Barry is talking about is lockin (like) detection, then I agree it can find signals buried in noise. But it does this (mostly) by a lot of signal averaging. Long time constants wrt the modulation frequency. And I think most of the noise reduction is just a narrowing of the bandwidth. (not that lockin's aren't totally cool.) | |
| Oct 22, 2014 at 23:29 | history | edited | Phil Frost | CC BY-SA 3.0 |
added 18 characters in body
|
| Oct 22, 2014 at 23:28 | comment | added | Phil Frost | @Barry What about 30 dB below? 40 dB below? Surely you don't have an infinite detection sensitivity in the presence of noise, and at some point the signal becomes indistinguishable from the noise, right? Does amplification change that in any way? | |
| Oct 22, 2014 at 22:54 | comment | added | Barry | @Dave Thank you for your comment. My point is that straight amplification does not increase SNR but there are signal processing techniques that utilize the known characteristics of the signal to effectively increase its level with respect to the noise. Cross correlation, with which I am very familiar since my company applied it in a passive sonar application, is one of these techniques. We were able to detect the presence of the noise radiated by a submarine when that noise was more than 20 dB below the level of the ambient background noise. | |
| Oct 22, 2014 at 22:15 | comment | added | Dave Tweed | "... it's that the signal is indistinguishable from atmospheric and electronic noise." That isn't true, and that's the point that @Barry was making. If you know what you're looking for, the signal is very distinguishable from noise, and that processing gain is what allows the usable SNR to be negative. | |
| Oct 22, 2014 at 21:21 | comment | added | Phil Frost | @Barry That's true, but that doesn't mean GPS has some magical property that allows it to be detected at the same SNR if the overall power is higher. For any modulation, no matter how robust, there is a minimum SNR at which it can be detected. Amplification of signal and noise doesn't increase SNR. | |
| Oct 22, 2014 at 20:37 | comment | added | David | "Anti-jamming" usually refers to the reduction of a small number of specific, higher strength signals in the same frequency range as GPS so that they do not affect position accuracy. How that works would be a different question. | |
| Oct 22, 2014 at 20:26 | comment | added | man abduo | u-blox.com/images/downloads/Product_Docs/… "This special jamming-mitigation method requires considerable processing power, which u-blox 5 / 6 supplies with an ARM® processor. Under control of this processor, this proprietary combination of hardware and software can reduce jamming signals by 30 dB compared to conventional products" | |
| Oct 22, 2014 at 20:24 | comment | added | man abduo | but I didnt find any commircal amplifiers has input range of -168 dBm ?! I read about some GPS receivers can distinguish signal at noise levels up to 30 dB , like u-blox . | |
| Oct 22, 2014 at 20:23 | comment | added | Barry | However, the GPS signal has very specific characteristics that permit it to be detected at very low levels, even below the background noise. It can be correlated with a reference signal. This process adds significant processing gain which directly increases the effective signal-to-noise ratio. This is not true for a random low level signal whose waveshape and other characteristics are not known. | |
| Oct 22, 2014 at 20:14 | history | answered | Phil Frost | CC BY-SA 3.0 |