08.06.2013 Views

Bernese GPS Software Version 5.0 - Bernese GNSS Software

Bernese GPS Software Version 5.0 - Bernese GNSS Software

Bernese GPS Software Version 5.0 - Bernese GNSS Software

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

6. Data Preprocessing<br />

6.2.4 Code Smoothing<br />

The final step consists of the smoothing of the code observations. The previous program<br />

steps have cleaned both, code and phase observations. This enables us now to smooth the<br />

code observations for the continuous data arcs, using the carrier phase observations. For<br />

code smoothing the code observations in a clean observation arc are actually replace by the<br />

phase observations shifted by the mean difference code minus phase in the arc. Of course<br />

we have to account for the opposite sign of the ionospheric effect for the code and phase<br />

observations. The smoothed code at epoch t may then be written as:<br />

where:<br />

˜P1(t) = φ1(t) + ¯ P1 − ¯ φ1 + 2 ·<br />

˜P2(t) = φ2(t) + ¯ P2 − ¯ φ2 + 2 ·<br />

f 2 2<br />

f 2 1 −f2 2<br />

f 2 1<br />

f 2 1 −f2 2<br />

<br />

· (φ1(t) − ¯ φ1) − (φ2(t) − ¯ <br />

φ2)<br />

<br />

· (φ1(t) − ¯ φ1) − (φ2(t) − ¯ <br />

φ2)<br />

(6.1)<br />

˜PF(t) Smoothed code measurement at epoch t and frequency F.<br />

φF(t) Carrier phase measurement at epoch t and frequency F.<br />

¯PF − ¯ φF Mean difference between all the accepted code and phase measurements in<br />

the current observation arc on frequency F.<br />

Figure 6.3 shows the effect of code smoothing. Shown are the residuals of a point positioning<br />

procedure, estimating only the receiver clock offset for each observation epoch and using<br />

the CODE final orbit and satellite clock estimates. The RMS error of the residuals in Figure<br />

6.3 are 1.53 and 0.17 meters for the raw code residuals and the smoothed code residuals,<br />

respectively. The code smoothing has been quite successful. The smoothed code residuals<br />

show systematic errors of up to one meter. The size of these biases is a function of the noise<br />

of the code observations and the number of observations used in the smoothing interval.<br />

One may consider smoothed code observations as ambiguity-fixed phase observations where<br />

the ambiguities were fixed only approximately.<br />

(a) Normal code residuals (b) Smoothed code residuals<br />

Figure 6.3: Code residuals from point positioning. Data from a receiver installed at USNO<br />

was used for day 133 of 1999.<br />

Page 106 AIUB

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!