Entropy-Based Channel Identification for NLOS Geolocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Geolocation technologies face significant challenges in urban and indoor environments due to non-line of sight (NLOS) propagation issues, leading to localization errors and reduced accuracy, as existing methods rely on line of sight assumptions and are not robust enough to differentiate between various NLOS conditions.
Innovation Solution
A method that estimates the entropy of the channel impulse response (CIR) to identify the transmission channel condition, distinguishing between direct and non-direct paths, and using this information to improve location estimation in NLOS environments through entropy-based channel identification and correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If geolocation systems use TOA, TDOA, or AOA techniques assuming line of sight conditions, then localization can be achieved in open environments, but significant localization errors occur in urban and indoor environments with NLOS propagation
Solution Approach 1:
The patent changes the parameter used for channel identification from traditional statistical moments (mean, variance) to higher-order statistics (skewness, kurtosis). This parameter change enables the system to better distinguish between LOS and NLOS channel conditions by capturing the asymmetric and heavy-tailed characteristics of NLOS impulse responses, thereby improving localization accuracy in urban and indoor environments
Solution Approach 2:
The patent performs preliminary channel condition identification using higher-order statistics before executing the localization algorithm. By pre-classifying channels as LOS or NLOS based on skewness and kurtosis values, the system can select appropriate processing strategies in advance, improving overall reliability without adding significant complexity to the main localization workflow
2Measurement precision
If traditional statistical methods (mean, variance) are used for channel identification, then the approach is simple to implement, but NLOS detection accuracy is insufficient
Solution Approach 1:
The patent transitions from using first and second-order statistical parameters (mean, variance) to third and fourth-order parameters (skewness, kurtosis). This parameter escalation captures the distinctive asymmetric and heavy-tailed nature of NLOS impulse responses, significantly improving NLOS detection accuracy despite the increased computational requirements
Solution Approach 2:
The patent applies higher-order statistics selectively to identify NLOS conditions rather than processing all signals with maximum complexity. By using skewness and kurtosis only when needed for NLOS detection and mitigation, the system achieves superior detection accuracy while managing computational complexity through targeted application rather than universal processing
3Reliability
If all received signals are treated as LOS in position estimation, then the localization process is straightforward, but NLOS corrupted measurements cause large positioning errors
Solution Approach 1:
The patent employs higher-order statistical parameters (skewness and kurtosis) to detect and identify NLOS channel conditions. By measuring these parameters from the received signal's impulse response, the system can distinguish NLOS from LOS conditions, enabling accurate classification and mitigation of NLOS-corrupted measurements in position estimation
Solution Approach 2:
The patent implements a feedback mechanism where channel identification results from higher-order statistics are fed back into the position estimation process. This feedback loop allows the system to adjust its processing based on detected channel conditions, improving reliability by preventing NLOS measurements from degrading position estimates
Data Source
AI summary
This invention relates to methods and devices for channel identification. The invention is particularly concerned with techniques for non-line of sight channel identification. In embodiments of the invention the methods and devices are used for channel identification in wireless geolocation systems. Embodiments of the invention make use of an entropy estimation of the channel to distinguish channel conditions and in particular to identify line-of-sight and non-line-of-sight channels and which can be used to solve the NLOS problem of determining relative distances between transmitter and receiver. In particular embodiments an entropy estimation of the channel impulse response (CIR) is used to construct a robust entropy-based channel identification technique. As a result, more accurate localization in indoor and other multipath environments may be possible.