CLS recommends the new algorithm based on Kalman filtering for all applications because it introduces significant improvements in the number of positions and their accuracy, especially for applications where just a few messages are received per satellite pass or for platforms operating in difficult transmission conditions.
This is true even when a platform is likely to have big gaps in average speed.
This is also true when there are changes in frequency :
For moderate frequency changes (<400 Hz) due for instance to temperature variation,
For more important frequency changes (reinitialization or oscillator fluctuation).
However, for those users who need very long time-series of homogenous data (several years), we recommend to continue using the least squares method for location processing.
|
Least-squares analysis |
Kalman filtering |
Raw data |
Frequency measurements |
|
Initialization procedure (computation of 1st position) |
Four messages required in a single satellite pass |
|
Number of messages needed per satellite pass to calculate a position |
Two |
One |
Accuracy estimation |
Error estimates available as an ellipse error with at least 4 messages. Partial information with 2 or 3 messages. |
Error estimates available as an ellipse error with at least 1 message. |
Number of solutions provided |
Two (nominal and mirror solutions) |
One (nominal solution) |
Digital elevation model |
USGS GTOPO30 |
|
This value-added service provides users with complementary information about transmitter performance. It also distributes non-standard locations, including locations calculated with less than four messages (Locations classes A, B) and locations that fail plausibility tests (Class Z). This service is very useful in certain cases, and is thus activated by default for a number of applications, including animal tracking.