|
From: | Ian Buckley |
Subject: | Re: [Discuss-gnuradio] How to use USRP to detect and collect weak satellite signals |
Date: | Wed, 4 Sep 2013 09:33:42 -0700 |
Ruoyu, First off, Singapore is a very noisy place (From a radio perspective), so you are going to have to pair any external low noise amplifiers with appropriate filters for your signals of interest. It is vital that none of the components in the radio signal chain receivers too much power which can lead to non-linearity and ultimately to damage to the radio. Placing a very high gain amplifier such as you have, in conjunction with a wide band antenna is not a safe proposition with a USRP daughter board, other loud and local signals will overwhelm the initial analog stages of the radio. I find even when using a dish pointed skywards in an urban area that introducing a wideband LNA without a filter causes other local cellular (etc) signals to saturate my USRP frontend. When I'm working with satellite signals close in frequency the Iridium signals you are interested in (~1.6GHz) I use a combination of a Minicircuits ZX60-242GLN-S+ LNA and a NBP-1560+ bandpass filter with good results with a variety of USRP's/daughter boards. For weather satellites in the 137MHz band (which is the same band as the Orbcomm downlink) I use a custom LNA+filter from SSB in Germany. In both cases gain is approximately 30dB and the noise figures very low, typical numbers for LNA's ideal for satellite use are 0.4-0.8dB NF. You should rethink your antenna(s) completely, Orbcomm use a right hand circular polarized signal for their downlink and you could thus use any antenna design that was intended for use to receive NOAA's APT weather satellite signal …you will find many references on how to build these if you google "APT antenna"….here's one incredibly simple example of a less than ideal, but simple solution that worked fine: http://websterling.com/tsro/apt/. In fact listening to the APT signal from NOAA-16, NOAA-18 and NOAA-19 may be a good starting point for you to develop your skills. Iridium is also a RHCP signal, and in those bands people typically use a patch or helix antenna for these types of signal, both can be built quite easily and at low cost. You should also familiarize your self with the open source software, "predict" and "gpredict" (There are others but these are recommended), as it is important to know when (and where with a directional antenna) your signal of interest is actually visible in the sky. And lastly GPS L1…..in the last few weeks I happen to have been listing to this with Balint from Ettus whilst we have been testing some antennas and other hardware. A USRP nor any other radio is going to see the raw signal above the noise floor with an omnidirectional antenna, the magic of GPS is all down to spread-specturm processing gain and is an interesting study if you have the time. One quick trick is to use autocorrelation to look for the signals, since the L1 C/A spreading code repeats every 1mS, it's relatively easy to prove it's there even though you can not see it in the FFT. And if your curious about what it looks like when you work a little harder to listen to it, then a USRP (SBX in this case) and a 1 meter dish prove more than sufficient: http://ionconcepts.com/ionconcepts/signals/L1-signal.png -Ian On Sep 4, 2013, at 1:12 AM, Chi <address@hidden> wrote:
|
[Prev in Thread] | Current Thread | [Next in Thread] |