TAP™ Composite Study Speed
Q: How can I make my Composite studies run faster?
A: The Composite Study function in TAP version 5.0.1069 and later with a Maintenance Subscription date of August 31, 2005, and later, implement a completely new method of searching the computed field strength databases for the component studies. The performance of the program is faster by orders of magnitude (see below) and the results are more well-defined.
First, a brief refresher on TAP Aggregate and Composite studies.
□ The original Aggregate study function released several years ago, enabled TAP users to combine two or more coverages studies to compute various results, such as the maximum field strength at each location, a best-server study, simulcast studies, etc. Aggregate studies could only combine Tile studies, and all of the component studies from different sites had to be on exactly the same rectangular tile grid.
□ Composite studies, introduced in 2002, provided more flexibility, using either radial or tile studies, and the tile studies do not need to be on the same points as is required by the Aggregate study. The flexibility required more sophisticated searching of the individual component studies to find the points to use at the desired target grid locations. The consequence was significantly increased execution times for the studies.
The latest version of the Composite study function uses a new method for finding the points from each study with markedly faster results. For example, a large Composite study using five component studies that took over 30 hours in the earlier version, now runs in approximately 90 seconds.
You may notice differences in the results for a study run under both the original Composite function and the new version. These differences result from a refinement in the process used for selection the points from each study.
For example, suppose your target grid (the point locations where the composite value, such as the maximum signal, is to be computed from all the component studies) has an increment of one-mile between points. Consider a small region of the study, with a particular target point in the center, and the points immediately surrounding it on a half-mile grid.
The earlier version of the Composite study searched the entire region around each point, all the way to the next point in each direction. In this example, that means the value assigned to the target point in the center of the diagram would be 30dBu, the maximum value of the points being considered.
The new version considers the “cell” around each target point, and only searches for points within the cell, or half-way to the next adjacent target point.
In this case, the value assigned to the target point in the center of the drawing would be 20dBu, since the 30dBu point is outside of the target point cell.
Since the selection of points is different in the new version, the computed results (maximum value, best server, etc.) may be different. The “cell” approach around each target point gives a better representation of the computed field strength values that affect that point.
 |
Copyright 2005 by SoftWright LLC