Designing Low-Correlation GPS Spreading Codes with a Natural Evolution Strategy Machine Learning Algorithm
This work provides a novel machine learning approach for designing navigation spreading codes, which is significant for improving the performance of next-generation GPS systems.
This paper addresses the design of GPS spreading codes by developing a natural evolution strategy (NES) machine learning algorithm. The algorithm successfully constructs high-quality families of spreading code sequences, outperforming Gold codes and Weil codes of similar lengths (up to 1023 and 1031 bits) and family sizes (up to 31 codes) in minimizing the maximum between mean-squared auto-correlation and cross-correlation.
With the birth of the next-generation GPS III constellation and the upcoming launch of the Navigation Technology Satellite-3 (NTS-3) testing platform to explore future technologies for GPS, we are indeed entering a new era of satellite navigation. Correspondingly, it is time to revisit the design methods of the GPS spreading code families. In this work, we develop a natural evolution strategy (NES) machine learning algorithm with a Gaussian proposal distribution which constructs high-quality families of spreading code sequences. We minimize the maximum between the mean-squared auto-correlation and the mean-squared cross-correlation and demonstrate the ability of our algorithm to achieve better performance than well-chosen families of equal-length Gold codes and Weil codes, for sequences of up to length-1023 and length-1031 bits and family sizes of up to 31 codes. Furthermore, we compare our algorithm with an analogous genetic algorithm implementation assigned the same code evaluation metric. To the best of the authors' knowledge, this is the first work to explore using a machine learning approach for designing navigation spreading code sequences.