Citation

Abstract

Current plans for supporting Voyager's encounter at Neptune include the arraying of the DSN antennas at Goldstone, California, with the National Radio Astronomy Observatory’s Very Large Array (VLA) in New Mexico. Not designed as a communications antenna, the VLA’s signal transmission facility suffers a disadvantage in that the received signal is subjected to a “gap” or blackout period of approximately 1.6 msec once every 5/96 sec control cycle. Previous analyses showed that the VLA data gaps could cause disastrous performance degradation in a VLA stand-alone system and modest degradation when the VLA is arrayed equally with Goldstone. These basic conclusions were independent of whether Voyager was using its convolutional code alone or the convolutional code concatenated with its Reed-Solomon outer code. New analysis indicates that the earlier predictions for concatenated code performance were overly pessimistic for most combinations of system parameters, including those of Voyager-VLA, The periodicity of the VLA gap cycle tends to guarantee that all Reed-Solomon codewords will receive an average share of erroneous symbols from the gaps. The number of gapped symbols is not subject to the same kind of statistical fluctuations that govern the ordinary random errors the code must also overcome, However, large deterministic fluctuations in the number of gapped symbols from codeword to codeword may occur for certain combinations of code parameters, gap cycle parameters, and data rates, In this article, several mechanisms for causing these fluctuations are identified and analyzed. Fortunately, the Voyager-VLA parameters do not produce wild fluctuations in the number of gapped symbols from codeword to codeword. The result is graceful degradation of concatenated code performance due to the VLA gaps, even for a VLA standalone system. The magnitude of the deterioration at a constant concatenated code bit error rate of 10-> is 0.5 dB to 0.6 dB for a VLA stand-alone system and 0.3 dB to 0.4 dB for the VLA arrayed equally with Goldstone. Even though graceful degradation is predicted for the Voyager-VLA parameters, catastrophic degradation greater than 2 dB can occur for a VLA stand-alone system at certain non-Voyager data rates inside the range of the actual Voyager rates. Thus, it is imperative that all of the Voyager-VLA parameters be very accurately known and precisely controlled.

Details

Volume
42-95
Published
November 15, 1988
Pages
112–133
File Size
1.1 MB