Indoor Localization Using Independent Component Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for indoor location estimation of user equipment in wireless networks using received signal strengths (RSS) measurements fail to accurately account for correlations and dependencies between signals from access points, leading to poor location accuracy.
Innovation Solution
The method employs independent component analysis (ICA) and a multivariate Gaussian-based approach to exploit spatial correlations in RSS measurements, transforming correlated signals into uncorrelated ones to improve location estimation by using second-order statistics and Kullback-Leibler divergence for position determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional RSS measurement methods are used for location estimation, then the implementation is simple, but the location accuracy is poor due to not accounting for correlations between signals
Solution Approach 1:
The patent transforms the RSS measurement problem by changing the parameter representation from individual correlated RSS values to uncorrelated independent components derived through ICA. This parameter transformation allows the system to capture the essential spatial correlation information while eliminating the redundancy and correlation among measurements, thereby improving location accuracy without proportionally increasing system complexity
Solution Approach 2:
The patent introduces independent component analysis (ICA) as an intermediary processing step between RSS measurement collection and location estimation. This intermediary transforms the correlated RSS measurements into uncorrelated independent components, which then serve as the basis for more accurate location estimation using multivariate Gaussian models and Kullback-Leibler divergence calculations
2Measurement precision
If correlated RSS measurements are used directly for location estimation, then the processing is straightforward, but the representation of RSS profiles is inaccurate
Solution Approach 1:
The patent applies parameter transformation by converting correlated RSS measurements into uncorrelated independent components through ICA. This change in parameter representation provides a more accurate RSS profile that captures the underlying spatial correlations while removing redundant information, enabling better location estimation
Solution Approach 2:
The patent replaces traditional signal processing approaches with independent component analysis, a more sophisticated mathematical framework that better captures the statistical dependencies in wireless signals. This substitution allows for more accurate RSS profile representation by modeling the complex correlations among multiple access point signals
Data Source
Figure 1
Figure 2
Figure 3.a
AI summary
Method, user equipment, system and computer readable medium for determining the actual position of an user equipment among a plurality of possible positions, each possible position being associated with a predetermined vector describing a power of signals received at this position from a plurality of base stations, the method comprising the following steps a step of determining a received vector describing a power of signals received from the plurality of base stations by the user equipment, and a step of independent component analysis of the received vector in order to obtain a uncorrelated vector, and a step of determining for each possible position a distance between the uncorrelated intersection and the predetermined vector associated with this position, and a step of determining the position of the user equipment by using at least one of the distances associated to the possible positions.