[Top] | [Previous] | [Next]

5.2.1 Planning Performance Drivers

Stressing cases were devised in each planning area, Orbit, Attitude, and Mission Planning. Results of these benchmark runs were used to ensure that the computer power assigned to the planning workstation is sufficient to handle the load.

In the Orbit area, the FRD requirements indicated that the orbit determination process must be completed within 20 minutes of completion of a tracking pass. We assumed that the orbit data preprocessing functions would be sufficiently automated to allow the human intervention required for orbit determination to be limited to 15 minutes. This implies a highly automated tracking data processing capability with some data editing capability. This leaves 5 minutes for the computer to process the data. A stressing benchmark was run (see Appendix D, Section D-6), involving 40.5 days of actual DSP tracking data and a subsequent 32-day ephemeris and events generation. This combination took 5.1 minutes on a Sun Sparc20. Similar benchmarks were done for realistic medium altitude cases to verify that the DSP case was indeed the worst case. Given the expected performance growth of this class of workstation, there should be no problem satisfying the requirement with a high-powered desktop workstation.

The performance requirement in Attitude is for attitude determination to be completed within 10 minutes of a successful orbit determination run. Several benchmarks were run on a 486 AST Premmia PC (see Appendix E, Section E-6). This PC is rated at about 1/6th of the floating point processing performance of the Sun Sparc20. The most stringent attitude determination benchmark, followed by an attitude ephemeris and events generation, took 6.5 minutes on the PC. This is equivalent to a little over 1 minute on a Sparc 20, leaving over 8 minutes for human intervention in the process. Sufficient automation of processing should be achievable provided that observation data can be processed and edited automatically.

The Mission Planning area did not have any temporal performance requirements. This is because mission planning is assumed to be a non-real-time function and the associated SOH activities are generally not time-critical. Nonetheless, several maneuver planning and ephemeris generation benchmarks were run (see Appendix A, Section A-6). These benchmarks took roughly 1 and 2 minutes respectively on the 486 PC and are no cause for concern. The one area where a high-powered desktop workstation will be needed is in automatic scheduling of room and network resources (see [2]). The mission planners are expected to employ an automatic scheduler and a brief survey of scheduling benchmarks [3] has indicated that the Sparc20 class machine is needed for a responsive scheduling system.