Radar Device Angle Estimation via Segmented Eigenvalue Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional ESPRIT methods for radar devices require long signal processing times, making it difficult for car-mounted radars to estimate angles quickly, especially when high-resolution processing is needed.
Innovation Solution
A radar device with a signal vector-forming unit, covariance matrix operation unit, submatrix-forming unit, regular matrix operation unit, eigenvalue decomposition unit, and angle calculation unit that performs a single eigenvalue decomposition on an (N-1)-dimensional matrix, reducing processing time by directly calculating the regular matrix from submatrices of the covariance matrix.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional ESPRIT method is used for high-resolution angle estimation, then measurement precision is improved, but processing time increases significantly
Solution Approach 1:
The patent segments the signal processing into two distinct stages: a first processing stage that performs eigenvalue decomposition to obtain signal subspace vectors, and a second processing stage that uses these vectors for angle estimation. This segmentation allows the computationally intensive eigenvalue decomposition to be performed once, with results reused across multiple angle estimation operations, thereby reducing overall processing time while maintaining high-resolution angle estimation precision.
Solution Approach 2:
The patent performs preliminary eigenvalue decomposition and obtains signal subspace vectors before the actual angle estimation process. These pre-computed signal subspace vectors are then utilized in the angle estimation stage, eliminating the need to perform eigenvalue decomposition repeatedly for each angle estimation operation. This preliminary action significantly reduces processing time while preserving measurement precision.
2Measurement precision
If higher resolution of angles is required, then measurement precision is improved, but processing time increases
Solution Approach 1:
The patent divides the processing into stages where signal subspace vectors are first obtained through eigenvalue decomposition, and then these vectors are used for high-resolution angle estimation. This segmentation enables the system to achieve high angle resolution without repeatedly performing the computationally intensive eigenvalue decomposition, thus reducing processing time while maintaining high resolution capability.
Solution Approach 2:
The patent performs preliminary computation of signal subspace vectors that are essential for high-resolution angle estimation. By having these vectors pre-computed and available, the system can perform rapid high-resolution angle estimation without the time penalty of repeated eigenvalue decomposition, thereby achieving high resolution with reduced processing time.
3Measurement precision
If multiple eigenvalue decomposition processing is performed, then measurement precision is improved, but processing time increases
Solution Approach 1:
The patent segments the processing to perform eigenvalue decomposition only once in the first stage to obtain accurate signal subspace vectors. These vectors are then reused in the second stage for angle estimation, eliminating the need for multiple eigenvalue decomposition operations. This segmentation maintains signal subspace accuracy while dramatically reducing processing time by avoiding redundant decomposition operations.
Solution Approach 2:
The patent performs the eigenvalue decomposition as a preliminary action to obtain signal subspace vectors that are then reused for angle estimation. This preliminary computation ensures accurate signal subspace representation without requiring multiple decomposition operations, thereby maintaining measurement precision while reducing overall processing time.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
The conventional ESPRIT method is accompanied by the problem of very long signal processing time. The radar device of the invention comprises a signal vector-forming unit for forming signal vectors based on waves reflected from an object and received by using a plurality of receiving antennas; a submatrix-forming unit for forming submatrices based on the signal vectors; a regular matrix operation unit for calculating a regular matrix from the submatrices; an eigenvalue decomposition unit for calculating an eigenvalue of the regular matrix; and an angle calculation unit for calculating an angle at where the object is present from the eigenvalue.