In order to improve accuracies to the sub-meter 
level, the carrier phase observable is generally 
required. One exception may be the use of narrow 
correlator technology on baselines of up to 20 km 
with the addition of chokering groundplanes to 
minimize multipath. However, in most cases 
differential carrier phase processing is 
implemented which adds significant complexity to 
the operation in terms of data acquisition and 
especially data processing due to the issues of 
carrier phase cycle slips and ambiguity resolution. 
The benefit is that ultimate accuracies in the order 
of a few cm can be achieved for static and 
kinematic applications within 10 km from a 
monitor station. 
As previously mentioned, GPS data can be acquired 
using a variety of techniques which are generally 
dictated by the application as well as the required 
accuracy. Table 3 summarizes the major techniques 
that are used for the collection of GPS positioning 
information as well as their general features. 
For static applications there are generally three 
techniques which are used, namely conventional 
static, rapid static and semi-kinematic. The first 
technique is the most robust and reliably gives 
results at the 1-3 ppm level (i.e. 1-3 cm per 10 km of 
monitor-remote separation). Its main 
disadvantage is the low productivity caused by 
the long observation span. In recent years rapid 
static surveying has become feasible due to the 
increased satellite constellation, P codeless 
receivers as well as improved processing 
algorithms. In general this technique can give cm- 
level accuracies in minutes, but is limited to 
baselines where integer ambiguities can be 
resolved, i.e. generally less than about 15 km. If 
rapid static surveying is implemented using a 
receiver that does not output the carrier phase 
(only the carrier-smoothed code), the accuracy is 
limited to about 1-3 m. Semi-kinematic positioning 
is an alternative to rapid static in that it does not 
depend on P codeless technology. In this case the 
carrier phase integer ambiguities are initialized 
at a starting point and then the receiver is 
transported to other points of interest. As long as 
lock is maintained to at least four satellites during 
these movements, cm-level accuracies can be 
achieved. However, if less than four satellites are 
tracked, the ambiguities must be re-initialized. It 
is expected that the rapid static technique will 
fully replace semi-kinematic positioning once the 
cost of P codeless technology decreases further. For 
more information on semi-kinematic surveying see 
Cannon (1991). 
Table 3: GPS Data Collection Techniques 
  
  
Technique General Features 
- carrier phase processing VIE 
Conventional - highest achievable accuracy 
Static - long observation span (30-60 min) 
- low productivity 
- carrier phase processing 
Rapid - accuracy from cm to sub-metre 
Static - Short observation span (5-15 min) 
(also called - may need P codeless receivers 
Fast Static) - high productivity 
- generally for baselines « 15 km 
  
- carrier phase processing 
- accuracy at the cm-level 
  
  
  
memi. - must maintain lock to 4 SVs 
kinematic ders 
- high productivity 
- generally for baselines « 15 km 
; - code or code-smoothed processing 
Kinematic 
- accuracy at the one to several m 
(also called ee ; F 
- easily implemented in real-time 
Dynamic) hi ios 
- high productivity 
- carrier phase processing 
Kinematic - static initialization required 
with - accuracy at the several cm level 
Static - generally for baselines < 50 km 
Initialization - must maintain lock to 4 SVs 
- high productivity 
- carrier phase processing 
- no static initialization required 
On-the-fly - accuracy at the several cm level 
Kinematic - generally for baselines < 50 km 
(OTF) generally need P codeless units 
high productivity 
  
The need for real-time positioning is usually 
driven by navigation or guidance requirements, 
however knowledge of high quality results during 
the observation span is also beneficial. Of the 
techniques presented in Table 3, the kinematic 
(dynamic) mode of operation is the easiest to 
implement in real-time. This is due to the use of 
the code observable which simplifies the 
processing requirements as wells as data 
transmission. Corrections from the monitor receiver 
can be transmitted at a rate as low as 50 bps using 
an internationally accepted standard format, i.e. 
RTCM-SC104. Accuracies in the real-time mode of 
operation are generally within those achieved in 
post-mission, assuming no significant breaks in the 
correction transmission. The U.S Coast Guard has 
initiated differential GPS (DGPS) correction 
transmission from its existing marine radiobeacons 
and the current coverage includes part of the US 
east coast as well as the Gulf of Mexico. Future 
166 
  
plans are 
coastline 
is also in 
in Canad 
To achie 
requirem 
2000 bps 
transmitt 
algorithr 
differenti 
of real-tir 
code dif 
systems 
feasibilit 
for high t 
4. 
41 Sub- 
Many GI 
few metre 
occupied 
on sub-m 
constraint 
that must 
occupy a 
standard 
m accurac 
a metre 
observabl 
a longer p 
satellite £ 
A study { 
the sub-rr 
Motorola 
can track 
output th 
collected « 
hours of c 
post-proc 
SEMIKIN 
processed 
were com 
in order 
Figure 3 
occupatio 
baseline. 
baseline. 
The two f 
accuracy 
Occupatioi 
different f