Rectangular Cumulant Matrix for Direction Finding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current direction finding techniques are limited in processing sources using rectangular arrangements of higher-order statistics and arbitrary antenna arrays, lacking the ability to optimize performance and source processing capacity effectively.
Innovation Solution
A method involving the arrangement of 2q-order statistics into a rectangular matrix, utilizing circular cumulants and a MUSIC algorithm to estimate directions of arrival, while reducing complexity and redundancy, allowing for increased source processing capacity and improved spatial resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If 2q-order statistics are arranged in a square matrix as in prior art, then the algorithm can process sources using higher-order statistics, but the number of sources that can be processed is limited and the complexity is high
Solution Approach 1:
The patent transforms the traditional square matrix arrangement of 2q-order statistics into a rectangular matrix arrangement. This dimensional change allows the statistics to be organized in an (Nv x N2q-v) format rather than a square Nq x Nq format, enabling the system to process more sources by utilizing the rectangular structure's additional degrees of freedom while reducing computational complexity through optimized matrix operations
Solution Approach 2:
The patent introduces new parameters v and (2q-v) that define the dimensions of the rectangular matrix, allowing flexible adjustment of the matrix shape based on the number of sources and sensors. By changing these parameters, the system can optimize the balance between the number of processable sources and computational complexity for different operational scenarios
2Measurement precision
If more sensors are added to increase spatial resolution, then the ability to process more sources improves, but the system cost and complexity increase
Solution Approach 1:
The patent segments the higher-order statistics into different rectangular matrix components that can be processed independently. By dividing the 2q-order statistics into manageable rectangular blocks and processing them through segmented algorithmic steps, the system achieves high spatial resolution equivalent to larger sensor arrays while maintaining a smaller physical sensor network
Solution Approach 2:
The patent creates virtual sensor arrays through mathematical transformations of the rectangular matrix statistics, effectively copying the resolution capabilities of large sensor networks without physically deploying additional sensors. The rectangular arrangement generates virtual aperture information that mimics larger arrays, providing high spatial resolution with reduced hardware complexity
3Productivity
If rectangular arrangements of higher-order statistics are used, then source processing capacity increases, but the computational complexity and redundancy increase
Solution Approach 1:
The patent extracts and eliminates redundant information from the rectangular matrix of higher-order statistics by identifying and removing duplicate computational operations. The method selectively processes only the independent elements of the rectangular arrangement, taking out the essential information needed for source processing while discarding redundant calculations, thereby reducing computational complexity while maintaining enhanced source processing capacity
Data Source
AI summary
Method and device for estimating the directions of arrival of sources of incidence on a network comprising N narrowband antennas each receiving the contribution of P sources, the statistics of the observations received on each antenna being able to be put in the form of a matrix C2q,x(I) of cumulants of order 2q, with q>1, whose coefficients are the circular cumulants of order 2q, Cum[xi1(t),..., xiq(t), xiq+1(t)*, ..., xi2q(t)*], , characterized in that the statistics of order 2q are arranged according to a rectangular arrangement, and the information contained in the rectangular matrix of order 2q of dimension (Nv x N2q-v), whose coefficients are the circular cumulants of order 2q of the observations, is exploited to find the direction of arrival of a source.