Satellite Signal Acquisition, Tracking, and Data Demodulation - Understanding GPS: Principles and Applications - page 220

Signal location versus
search threshold

1/2 chip

1 cell

1 Doppler bin 6

Signal migration direction

V t3

4

V t2

V t1

2

Start of search
(expected
value of
Doppler)

Search direction

1

3

5

7

1,023 chips

Doppler
search
sequence

Figure 5.34 Two-dimensional C/A code search pattern.

nal is acquired. Signal obscuration (trees, buildings, snow or ice on the antenna, and
so forth), RF interference, ionospheric scintillation, and antenna gain roll-off can all
significantly reduce C / N 0 (see Chapter 6).
Referring to Figure 5.34, the search pattern usually follows the range direction
from early to late in order to avoid multipath with Doppler held constant until all
range bins are searched for each Doppler value. The direct arrival of a signal subject
to multipath is always ahead in time of the reflected arrivals. In the Doppler bin
direction, the search pattern typically starts from the mean value of the Doppler
uncertainty (zero Doppler if the actual LOS velocity estimate is unknown) and then
goes symmetrically one Doppler bin at a time on either side of this value until the
3-sigma Doppler uncertainty has been searched. Then the search pattern is repeated,
typically with a reduction in the search threshold scale factor. It is important to rec-
ognize that the C/A code autocorrelation and crosscorrelation sidelobes can cause
false signal detections if these sidelobes are strong enough. The sidelobes tend to
increase as the search dwell time is decreased. To counter this problem, a combina-
tion of both increased dwell time (to minimize sidelobes) and a high detector thresh-
old setting (to reject sidelobes) can be used for the initial search pass. On subsequent
search passes, the dwell time and threshold can be decreased. The penalty for this
scheme is increased search time when the C / N 0 is low.
During the dwell time, T , in each cell, the I and Q signals are integrated and
dumped and the envelope is computed or estimated. Each envelope is

compared to a threshold to determine the presence or absence of the SV signal. The
detection of the signal is a statistical process because each cell either contains noise
with the signal absent or noise with the signal present. Each case has its own proba-
bility density function (pdf). Figure 5.35 illustrates a single trial (binary) decision

I

2

+ Q

2