Geolocation Accuracy via Delay-Doppler Signal Discrimination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current navigation technologies, such as GNSS, face challenges in indoor and urban environments due to low signal strength and severe multipath interference, where existing methods are ineffective in discriminating between line-of-sight (LOS) and non-line-of-sight (NLOS) signals, leading to corrupted position estimates.
Innovation Solution
The introduction of the Synthetic Aperture Line of Sight Assessment (SALSA) method, which uses the direction of arrival to discriminate between LOS and NLOS signals, and the Genetic Algorithm for Multipath Elimination (GAME) to identify and isolate incorrectly identified LOS signals, along with Weighted Average Functionality For Limiting Error Sources (WAFFLES) to mitigate timing and calibration errors, and LEAF to estimate pseudorange in environments with signal obstruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If narrow correlators or strobe correlators are used to mitigate multipath interference, then multipath signals can be effectively isolated or removed, but these techniques are only effective when multipath delays are on the order of or larger than the inverse signal bandwidth (at least 0.1 chip length, 30 meters for GPS civilian signal)
Solution Approach 1:
The patent transitions from one-dimensional delay domain processing to two-dimensional delay-Doppler domain processing. By incorporating Doppler frequency information as an additional dimension, the system can distinguish LOS signals from multipath signals even when their delays are identical or very close, enabling effective multipath mitigation in indoor and urban environments where traditional delay-only methods fail.
2Measurement precision
If MEDLL receiver with many correlators is used to integrate satellite signal at different delays, then LOS and multipath components can be discriminated, but this method is effective only if there are few dominant signal components and if one of the dominant components is the LOS signal
Solution Approach 1:
The patent extends MEDLL by adding Doppler frequency discrimination to the delay domain analysis. This creates a two-dimensional delay-Doppler profile that enables reliable LOS identification even when multiple dominant components exist or when LOS is weak, by exploiting the characteristic that LOS and multipath signals typically have different Doppler shifts.
Solution Approach 2:
The patent changes the processing parameters from delay-only to delay-Doppler joint processing. By transforming the signal representation into the delay-Doppler domain and analyzing the spectral characteristics in both dimensions, the system can reliably distinguish LOS from multipath components under severe multipath conditions where traditional single-parameter methods fail.
3Reliability
If direction of arrival information is used to discriminate between LOS and NLOS signals, then reliable LOS identification can be achieved in environments where multipath signal amplitude exceeds LOS signal amplitude, but this requires processing multiple signal dimensions
Solution Approach 1:
The patent utilizes direction of arrival (DOA) information as an additional dimension for signal discrimination. By combining delay, Doppler, and spatial domain information, the system achieves robust LOS identification in severe multipath environments. The DOA difference between LOS and reflected signals provides an independent discriminant that enhances reliability even when signal amplitudes are comparable.
Data Source
AI summary
A method for increasing accuracy of geolocation based on timing measurements of RF signals. The method includes: (a) receiving at a user one or more RF signals transmitted over time, (b) correlating the received signals with a plurality of replica waveforms, (c) creating a first set of estimates of the time or arrivals of the signals at the user, (d) using the first set of estimates to create a plurality of models, wherein each model assumes that some of the measurements are affected by multipath or interference, (e) testing the models to estimate the likelihood of timing measurements for each model, and (f) determining the set of LOS measurements as one associated with a model having the largest likelihood.