Citation
Abstract
This paper describes an investigation leading to techniques to reduce the error present in the DSN station measurement of doppler phase-noise variance and standard deviation. The error source of concern is the doppler counter resolver, which sorts continuous phase data into finite group intervals. The intervals are 3.5 deg (1 MHz bias) or 18 deg (5 MHz bias) wide. Counter operation is covered to define parameters, and then appropriately limited model functions for the variance, its estimate, and the error are stated. The model introduces linear and sinusoidal mean-sweep functions to investigate their effect on the error. Machine-programming of the model yields results which indicate that meansweep significantly reduces the measurement error when phase-noise standard deviation is low (less than 2 deg at 1 MHz bias, or 10 deg at 5 MHz bias). Linear sweep is the most accurate technique, but sinusoidal sweep is recommended as the more feasible; errors with the latter do not exceed 0.15 deg rms (1.0 MHz bias) in the region of interest. Finally, the recommended parameters and the expected residual error are stated for use in test program configuration and data reduction; the error is asymptotic to a universal constant variance offset which can be appropriately subtracted to compensate for the group interval width at all noise levels.
Details
- Volume
- 42-25
- Published
- February 15, 1975
- Pages
- 163–171
- File Size
- 974.0 KB