Global Positioning System Reference
In-Depth Information
arising in land surveying, where one has the luxury of time. More recently, for appli-
cations where the user is moving with respect to the fixed reference station and
real-time positioning is required (i.e., a kinematic environment), rapid resolution of
carrier-cycle integer ambiguities is highly desirable and an absolute must if
centimeter-level accuracies are to be achieved. The ambiguity function technique is
thus no longer used in most current systems.
Advantage can be taken by combining the L1 and L2 frequencies to speed the
ambiguity resolution process, and this approach has been the subject of a number of
articles in the literature (e.g., Hatch [18]). After the P code was encrypted (becoming
the Y code) by the DOD, a number of receiver manufacturers were successful in
recovering the full carrier phase and pseudorange Y code observables. This has
allowed the continued use of the dual-frequency property of the GPS signal struc-
ture. These dual-frequency receiver measurements can be combined to produce the
sum and difference of the L1 and L2 frequencies. The result is sum and difference
wavelengths of 10.7 cm and 86.25 cm, respectively. Using the difference wavelength
(known as the wide lane) makes the integer ambiguity search more efficient. A
change of one wide-lane wavelength results in virtually a fourfold increase in dis-
tance over that of one wavelength at either the L1 and L2 frequencies alone. Obvi-
ously, the search for the proper combination of integer ambiguities progresses more
quickly using wide-lane observables, but the requirements on the receiver for simul-
taneous dual-frequency tracking—here, the P(Y) code is generally used—are more
stringent. In particular, the noise factor for the wide-lane processing goes up by a
factor of nearly six [19]. These matters aside, wide-lane techniques offer great
advantage for obtaining rapid, on-the-fly, integer ambiguity resolution, and the
methodology will be presented later in this chapter.
8.4.1 Precise Baseline Determination in Real Time
Determination of the carrier-cycle ambiguities on the fly is key to any application
where precise positioning at the centimeter level, in real time, is required. Such tech-
niques have been successfully applied to aircraft precision approach and automatic
landing for approach baselines extending to 50 km in some instances [20-23]. They
are equally applicable, however, to land-based or land-sea applications (e.g., precise
desert navigation or off-shore oil exploration). In contrast, land-surveying applica-
tions and the like, often involving long baselines, have had the luxury of the postpro-
cessing environment and, as a result, accuracies at the millimeter level are
commonplace today. Techniques applied in such instances involve resolution of car-
rier cycle ambiguities on the data sets collected over long periods of time (generally
an hour or more). In addition, postprocessing of the data lends itself to recognition
and repair of receiver cycle slips. Precision can be further enhanced by use of precise
satellite ephemerides. These topics, while of interest, are beyond the scope of this
book. Texts such as [24] ably cover these applications.
The following discussion focuses on an integer ambiguity resolution technique
first proposed in [25], which capitalizes on some concepts from [26] to resolve the
inconsistencies between redundant measurements. The latter work maintains that
“all information about the 'inconsistencies' resides in a set of linear relationships
known as parity equations .” While these techniques were originally applied to iner-
 
Search WWH ::




Custom Search