On the benefits of short-term weather forecasting for Ka-band (32 GHz)

TLDR: In this article, an analytical method for comparing the performance of an ideal forecasting algorithm to the best statistical methods in terms of average data return is derived, which is applied to two different cases.

Abstract: Due to spectrum limitations at lower frequencies, NASA's Deep Space Network is currently implementing Ka-band (32 GHz) tracking capabilities at all of its deep space communication complexes (DSCC's). Since weather effects and increases in the atmospheric noise temperature associated with them are the biggest uncontrollable factors in the performance of a Ka-band deep space telecommunications link, use of algorithms to forecast the atmospheric noise temperature for a pass is desirable. In this paper, an analytical method for comparing the performance of an ideal forecasting algorithm to the best statistical methods in terms of average data return is derived. This methodology is applied to two different cases. In the first case, the spacecraft cannot change its data rate during the pass. In the second case, the spacecraft can continuously vary its data rate. This methodology is applied to four different elevation profiles whose maximum elevation varies from less than 30 degrees to greater than 80 degrees for Goldstone, Madrid and Canberra DSCC's. This analysis shows that for the fixed data rate case, while the forecasting does not significantly increase the average data return on the link (between 0.2 dB and 0.4 dB, depending on the DSCC and the elevation profile) it does improve the reliability of the link significantly (in ideal case to 100%). For the continuously variable data rate case, forecasting improves both the average data return (by between 1 dB and 1.9 dB depending on the elevation profile and the DSCC) and the reliability of the link (in ideal case to 100%).