Tracking GPS Signals - Fundamentals of Global Positioning System Receivers: A Software Approach - page 174

FIGURE 8.7 An input signal and two sections of the real part of e − j 2 πnk/N .

change due to the navigation data. Since there is noise in the input data, the
phase change will not exactly equal the desired values of 0 to

±

π . For example,
if the phase shift is close to 0 or 2 π , it is considered that there is no phase shift.
If the phase shift is close to π , it is considered that there is a π phase shift.

In general, a threshold can be set at

±

π /2. In Figure 8.8, the thresholds are set
at π /2 and 3 π /2. If the absolute value of the difference angle is within the range
π /2 and 3 π /2, it can be classified as a π phase shift. Otherwise, there is no
phase shift. The π phase shift cannot occur within 20 ms and it occurs only at a
multiple of 20 ms.

±

8.10 ACCURACY OF THE BEGINNING OF C/A CODE MEASUREMENT

The input signal is digitized at 5 MHz, or every data point is separated by 200 ns.
With this time resolution, the corresponding distance resolution is about 60 m
(3 × 10 8 × 200 × 10 − 9 ), which is not accurate enough to solve for a user position.

Since the GPS signal and digitizing clock of the receiver cannot be synchronized,
it is not likely to match a data point with the true beginning of the C/A code.
Under the worst condition, the digitized beginning of the C/A code can be 100 ns
away from the true value, when the true beginning of the C/A code falls at
the middle of two digitizing points. The acquisition program can only measure
the accuracy of the beginning of the C/A code to the digitized resolution. It is
desirable to measure the beginning of the C/A code very accurately.
In the conventional tracking loop discussed in Section 8.7, the locally gen-
erated C/A code is updated every millisecond. The purpose of the updating is
to generate a C/A code to match the C/A code in the input signal and gener-
ate a carrier frequency to match the carrier frequency in the input signal. Only