Methods and Systems to Diminish False-alarm Rates in Multi-hypothesis Signal Detection through Combinatoric Navigation
In satellite-based navigation, a navigation solution is computed from navigation signals received from multiple navigation satellites. A received navigation signal may have a relatively low signal-to-noise ratio, which may make it difficult to distinguish the signal from noise. Additionally, motion of a receiver or a satellite may impart a Doppler frequency shift to the signal. Atmospheric and/or other environmental conditions may impart propagation delay to the signal. These issues may impact the detection accuracy of the navigation signal, which may impact the accuracy of a navigation solution.
APL researchers have developed a range-Doppler correlation grid that may be generated for each of multiple received navigation signals. Energy peaks of the grids may be compared to a relatively low first threshold to identify up to multiple range-Doppler hypotheses per grid. The first threshold may be less than a conventional per-signal or per-grid false-alarm threshold. The relatively low signal-detection threshold permits detection of weaker signals. Subsequent evaluations to identify a plausible and a most plausible navigation solution effectively transform a per-navigation-signal false-alarm rate to a per-navigation-solution false-alarm rate.
Where a grid energy peak exceeds a second threshold that is higher than the first threshold, a corresponding hypothesis may be identified as a sole hypothesis for the corresponding navigation signal.
This invention could be integrated into high-sensitivity Global Navigation Satellite System (GNSS) receivers—even in the short term. As computers, application specific integrated circuits, etc., become faster, the invention becomes increasingly valuable.
Patent Status: U.S. patent(s) 8,325,086 issued.CONTACT:
Dr. G. R. Jacobovitz
Phone: (443) 778-9899