IGS satellite biases

[plutonheaven]

Adds satellite hardware biases from IGS's BIA files.

A column sat_code_bias_m and sat_carrier_bias_m is added to the prx output for prx-processing level 3.

The BIA file discovery takes care to choose the same analysis center as for the SP3 files. An CLI argument --analysis_center has been added to let the user choose the correction source..

[plutonheaven]

@jtec I finished implementing satellite hardware retrieval from IGS BIA files.

It is quite difficult to test. For now, I computed the Melbourne-Wuebbena combination with and without corrections, and I observe that the coorrected one is closer to an integer value + a common bias accross all satellites (rx hw bias). But not visible for all satellites. I guess the satellite phase wind-up may be left to be corrected.

I've also checked that the iono-free combination of the sat code bias is close to 0, as expected from IGS files (since the sat clock is an iono-free clock containing this iono-free code bias).

If you have some idea on how to test the correct implementation of satellite hardware biases, do not hesitate!