Eigenvector-Based Living Body Estimation Device
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for estimating the number and positions of living bodies using radio signals face accuracy issues when the number of living bodies is large, due to reduced differences in eigenvalues corresponding to living bodies and noise, leading to deteriorated estimation accuracy.
Innovation Solution
An estimation device that extracts living body information from received signals, calculates eigenvectors of a living body correlation matrix, and uses a combination of position estimation methods like the Capon method and MUSIC to accurately estimate positions and numbers of living bodies, distinguishing between actual and false images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If eigenvalue analysis is used to estimate the number and positions of living bodies, then the estimation can be performed using wireless signals, but the estimation accuracy deteriorates when the number of living bodies is large due to reduced difference between eigenvalues
Solution Approach 1:
The patent changes the parameter from eigenvalue analysis to eigenvector analysis. By using eigenvectors instead of eigenvalues, the system can distinguish between living bodies and noise more effectively, maintaining accurate estimation even when the number of living bodies is large and eigenvalue differences become minimal.
Solution Approach 2:
The patent introduces a correlation matrix as an intermediary structure. By constructing a correlation matrix from the received signals and performing eigenvector decomposition on this matrix, the system creates a mathematical representation that separates living body information from noise, enabling accurate estimation without direct reliance on eigenvalue magnitude differences.
2Area of stationary object
If the number of living bodies increases, then the detection range expands, but the difference between eigenvalues corresponding to living bodies decreases, leading to worsened estimation accuracy
Solution Approach 1:
The patent transitions from using eigenvalue magnitudes to using eigenvector directions for identification. This parameter change allows the system to maintain discrimination capability regardless of the number of living bodies, as eigenvectors provide angular information that remains distinct even when signal strengths (eigenvalues) become similar.
Solution Approach 2:
The patent moves the discrimination dimension from magnitude (eigenvalue) to direction (eigenvector). By utilizing the directional information contained in eigenvectors rather than the magnitude information from eigenvalues, the system achieves robust performance in detecting multiple living bodies without suffering from eigenvalue convergence.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The solution improves estimation accuracy by combining eigenvector-based methods with position estimation techniques, effectively addressing the limitations of existing technologies in large populations and enhancing the detection range and noise resistance.
Implementation Method 1
an eigenvector calculation unit configured to calculate one or more eigenvectors of a living body correlation matrix obtained from the living body information
Data Source
AI summary
An estimation device includes: a living body information extraction unit that extracts living body information which is a component corresponding to one or more living bodies in a space; an eigenvector calculation unit that calculates one or more eigenvectors of a living body correlation matrix obtained from the living body information; a first position estimation unit that estimates, using the living body correlation matrix, positions of the one or more living bodies and at least one false image, according to a predetermined position estimation method; a second steering vector output unit that extracts, from first steering vectors stored in a storage, and outputs as second steering vectors, first steering vectors corresponding to the positions estimated; and a second position estimation unit that estimates at least one of the position and the number of the one or more living bodies, using the one or more eigenvectors and the second steering vectors.