The Global Positioning System: A Shared National Asset (1995)
Chapter: Appendix G Increased Bandwidth Performance Analysis
Appendix G
Increased Bandwidth Performance Analysis
To determine more quantitatively the sensitivity of increasing the bandwidth, an analysis was performed using relationships given in the literature for comparing the performance characteristics of the existing C/A-code (narrow band) with that of a wider band signal format.1
The code pseudorange error for a narrow correlator design was then fed into a covariance analysis to determine the smoothed pseudorange errors after further carrier-phase smoothing.2 The following scenarios, which are typical of difficult vehicular applications, were investigated:
- pseudorange accuracy 100 seconds after signal re-acquisition, for zero and high multipath conditions, and
- pseudorange accuracy 10 seconds after signal blockage recovery, for zero multipath and high multipath conditions.
The results are shown in Table G-1. Note that the errors in Table G-1 are pseuodoranges errors (la), noise plus multipath.
Table G-1 Accuracy Recovery Characteristic in Multipath for a Narrow, C/A-Type Code and a Wide-Band, P-Type Signal Format
|
Signal Type |
10 seconds after re-acquisition |
100 seconds after re -acquisition |
|
|
|
|
|
Narrow, C/A-type code |
1.4 meters |
1.3 meters |
|
High multipath |
|
|
|
Wide-band, P-type code |
0.4 meters |
0.3 meters |
|
High multipath |
|
|
|
Narrow, C/A-type code |
0.37 meters |
0.18 meters |
|
No multipath |
|
|
|
Wide-band, P-type code |
0.1 meters |
0.057 meters |
|
No multipath |
|
|
|
a. Strong vehicular multipath-to-direct reflection ratio of 0.2, distributed uniformly over full code chip width. Vehicular multipath at code tracking loop output modeled as zero mean Gauss-Markov with a 10-second correlation time. b. C/A-code receiver with 8-MHz bandwidth and 0.2 chip spacing. c. Wide-band signal receiver with 20 MHz bandwidth and 1 chip spacing. d. In all cases 40 dB-Hz carrier-to-noise ratio. e. In all cases Code loop bandwidth 1 Hz, followed by carrier-smoothed-code filter matched to multipath and ionosphere temporal correlation characteristics. f. With no multipath, carrier-smoothed code accuracy limited by code minus carrier ionospheric drift. g. No cycle slips |
||
Under ideal reception conditions, and given sufficient settling time, the pseudorange errors are at the decimeter level for both signal structures. But in the important case of strong multipath, both 10 and 100 seconds after signal blockage, the wide-band, P-type signal is substantially better in performance.
Finally, the relative performance of narrow, C/A-type code and the wide-band, P-type code signal under conditions of in-band interference was examined. In a large number of important civilian applications, a critical requirement is continuous tracking of carrier phase. Beyond the obvious need to recover satellite ephemeris parameters, continuous phase availability allows for smoothing of code pseudorange noise, as well as precise kinematic positioning, Therefore, the susceptibility of phase tracking to in-band interference was of interest. Assuming a phase-tracking threshold of 30 dB-Hz, the tolerable range from a 1-watt, wide-band jammer was computed. The narrow, C/A-type code loss-of-carrier distance was 40 kilometers; the wideband, P-type signal loss-of-carrier distance was 13 kilometers.
