Global Positioning System Reference
In-Depth Information
global integrity concept is the answer to the needs of different types of users who are all
looking for different services in terms of signal and performance.
A new concept of Integrity will be introduced in the following paragraphs. In particular,
starting from the Galileo Integrity concepts, we will illustrate a few solutions to the integrity
problem and describe a new one, in which data of different constellations (GPS/EGNOS
and Galileo) are combined in order to improve the accuracy and the availability of the
navigation data.
2. Galileo integrity
The integrity concept developed in Galileo has the aim of ensuring the correct computation
of the user's position and provide a valid alarm to the user if the error in the position
solution has exceeded a fixed threshold - the Alert Limit - relative to the specific application
(Martini, 2006). The user can be in one of the following conditions (Table 1):
System
Alert for
Satellite
Satellite User
Msg.
Case
System Case
System State
Comment
1
Fault-Free
Nominal
NO
OK
2
Fault-Free
Nominal
YES
NOT-OK
False Alert
3
Faulty
Non-nominal
YES
NOT-OK
True Alert
4b
Faulty
Non-nominal
YES
NOT-Monitored
True Alert
Error below
Threshold
4c
Faulty
Non-nominal
NO
OK
Table 1. Examples of integrity
In order to estimate all the errors that might occur in different situations, we have adopted a
Gaussian model (J. Rife et al., 2004), whose standard deviation derives from the standard
deviation of the error distribution and from the accuracy of the system. Moreover, each
Gaussian distribution might have a bias, representing the presence of a faulty condition. The
following Figure (Figure 1) shows the system's estimate of the error distribution, illustrating
the situations displayed in Table 1. The first two cases concern a faulty free condition: the
error is modelled with a zero-mean Gaussian distribution. In this case, the system only has
an estimation of the error. This estimation could be considered as a sample of the above-
mentioned Gaussian distribution, and this sample could be above (1) or below (2) the
specific threshold. In case 1, the system is working in nominal condition, whereas case 2
concerns a False Alarm condition. The failure is modelled as the presence of a bias in the
error distribution. This bias could be higher than the threshold (case 3), and in that case the
system would certainly detect it.
Otherwise, the mentioned bias could be higher than the threshold, but the sample of the
distribution could be below this limit (case 4). This case is referred as Missed Detection
condition (Martini, 2006).
The Galileo system provides three elements to preserve user integrity:
•
Signal-in-Space Accuracy (SISA): this is the expectation of the errors relative to the SW's
clock and ephemerides, based on long term observations.
Search WWH ::
Custom Search