Bias Estimation for TOA Wireless Geolocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for bias estimation and correction in time-of-arrival (TOA)-based wireless geolocation systems, particularly in indoor and urban environments, face challenges due to multipath propagation and non-line-of-sight (NLOS) issues, leading to inaccurate range and distance measurements, which degrade localization and tracking accuracy.
Innovation Solution
A method that recursively estimates biases in distance measurements between a mobile device and base stations using differential bias changes, allowing for dynamic estimation and locking onto the lowest experienced bias, without assuming zero bias in line-of-sight (LOS) conditions or requiring LOS/NLOS identification, and can be integrated with a Kalman filter framework for position estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If statistical bias correction techniques are used that process measurements over a window, then bias correction is achieved, but the approach requires a priori statistical characterization of biases which is difficult to maintain in highly dynamic environments
Solution Approach 1:
The patent implements a dynamic bias correction approach using a sliding window that adapts to changing environments. The window size and parameters are adjusted based on observed bias variations, allowing the system to maintain accuracy in highly dynamic indoor and urban environments without requiring fixed a priori statistical characterization. The bias estimate is continuously updated as new measurements become available within the sliding window period.
Solution Approach 2:
The system performs preliminary bias estimation and correction within the sliding window framework before final position calculation. By pre-processing the range measurements to remove biases using the dynamic window approach, the system prepares corrected measurements that can be directly used for accurate localization without requiring complex post-processing or a priori statistical models.
2Measurement precision
If NLOS identification and mitigation approaches are used, then NLOS measurements can be classified and mitigated, but perfect NLOS identification is highly unrealistic and rarely achieved
Solution Approach 1:
The patent extracts and corrects NLOS biases directly from the range measurements using the sliding window statistical approach, without requiring explicit NLOS identification. By separating the bias component from the total measurement error and correcting it independently, the system achieves NLOS mitigation without relying on unreliable NLOS classification algorithms.
Solution Approach 2:
The sliding window statistical analysis acts as an intermediary that processes the raw range measurements to extract bias information. This intermediate processing step transforms the problematic NLOS-contaminated measurements into corrected measurements, serving as a mediator between the raw data and the final position calculation without requiring direct NLOS identification.
3Adaptability or versatility
If bias tracking algorithms incorporating Kalman Filter are used, then dynamic bias tracking is achieved, but the approach requires a priori knowledge of the bias covariance matrix which is difficult to obtain due to rapid fluctuations
Solution Approach 1:
The system performs self-service by automatically estimating the bias covariance characteristics through the sliding window statistical analysis. Rather than requiring external provision of bias covariance matrices, the algorithm derives these parameters directly from the observed measurement variations within the window, enabling adaptive bias tracking without external knowledge input.
Solution Approach 2:
The sliding window approach provides continuous feedback on bias characteristics by analyzing the distribution of recent measurements. This feedback mechanism allows the system to dynamically adjust its bias estimates and correction strategy based on observed patterns, effectively tracking rapid bias fluctuations without requiring pre-specified covariance information.
4Measurement precision
If weighted least squares or weighted constrained optimization algorithms are used, then location optimization is improved, but computational complexity increases
Solution Approach 1:
The patent applies partial correction by focusing on removing the dominant bias component through the sliding window statistical approach, rather than performing full optimization. This partial action on the bias term alone achieves significant accuracy improvement without the computational burden of complete weighted least squares or constrained optimization, maintaining high productivity.
Data Source
Figure 1~2
Figure 3~4(b)
Figure 5(a)~5(b)
AI summary
This invention relates to methods and devices for bias estimation and correction, particularly for time-of-arrival (TOA) –based wireless geolocation systems. Multipath and non-line-of- sight (NLOS) biases can cause distance estimation errors in the range of tens-hundreds of meters and is particularly problematic in urban and indoor environments. The behaviour of the biases dynamically changes depending on the clutter and/or obstructions between the base station and the mobile device. Aspects of the present invention provide practical real- time bias estimation and correction techniques for TOA-based systems and are based on inferring and estimating the biases from dynamic time differential measurements. The