Performance of Stand-Alone GPS - Understanding GPS: Principles and Applications - page 321

codes. Since GPS receiver front ends, in general, do not have constant group delay
throughout the passband, very small intersatellite biases can be observed upon
C/A code pseudoranges. These biases are typically on the order of a few millime-
ters [30].
Hardware biases between spectrally different signals on one frequency, or
among signals on different carrier frequencies, are larger in magnitude. In [31], dif-
ferential group delay biases between L1 GPS C/A and GALILEO Open Service sig-
nals were analyzed within a representative receiver to be on the order of several
nanoseconds (~1m in range). These biases are not common to all measurements and
thus would influence positioning performance if not calibrated or estimated.
Within dual-frequency receivers, a portion of electrical paths followed by the
L1 and L2 signals may be physically different, resulting in sizeable differential range
errors. For the positioning user, the L1-L2 bias may often be ignored since it results
in a common error for every ionospheric-free pseudorange (see Section 7.2.4.1),
which will drop out in the estimated receiver clock bias.
Another error that can be attributed to the receiver hardware is hardware-
induced multipath [32]. This error is caused by reflections of the GPS signal that
occur within the receiver hardware due to the presence of an impedance mismatch
between RF components. This error can be removed or reduced by careful design of
the receiver front end.
7.2.8 Pseudorange Error Budgets
Based on the earlier discussion regarding error constituents, we can develop
pseudorange error budgets to aid our understanding of stand-alone GPS accuracy.
These budgets are intended to serve as guidelines for position error analyses. As
indicated in (7.1), position error is a function of both the pseudorange error (UERE)
and user/satellite geometry (DOP). The geometry factor will be discussed in Section
7.3.1.
The total system UERE comprises components from each system segment: the
space segment, the control segment, and the user segment. This budget can be made
based on either the use of single-frequency measurements or the use of dual-fre-
quency measurements to determine the ionospheric delay. The error components
are root-sum-squared to form the total system UERE, which is assumed to be
Gaussian distributed. The use of RSS addition of UERE components is justified
under the assumption that the errors can be treated as independent random vari-
ables such that the variances add or equivalently the 1-sigma total error is the RSS of
the individual 1-sigma values.
Tables 7.3 and 7.4 show estimates of typical contemporary UERE budgets
based on the data presented in Sections 7.2.1-7.2.7. Table 7.3 describes a typical
UERE budget for a dual-frequency P(Y) code receiver. Table 7.4 shows the UERE
budget for a single-frequency C/A code receiver. For a single-frequency user, the
dominant pseudorange error source is the residual ionospheric delay after applying
the broadcast ionospheric delay corrections. Dual frequency users can use the tech-
nique described in Section 7.2.4.1 to nearly completely remove the error due to
ionospheric delays.