Digital Filter Tap Selection Using Multi-Resolution Vector Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Digital filters face a trade-off between achieving desired filter characteristics and computational expense due to the exponential increase in tap value combinations with resolution, making it inefficient to determine optimal tap values for high accuracy.
Innovation Solution
A method and system that iteratively refine tap values by generating vectors at increasing resolutions within a threshold distance of selected optimal vectors, using a performance metric to select the best vectors, significantly reducing computational expense while achieving optimal or near-optimal filter performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the resolution of tap values is increased to achieve higher filter accuracy, then the accuracy of filter characteristics is improved, but the computational expense increases exponentially
Solution Approach 1:
The patent segments the search process into multiple stages: first searching at a coarse resolution to identify a promising region, then refining the search within a threshold distance of the best vector found at each stage. This divides the exponentially large search space into manageable segments that can be searched sequentially, reducing overall computational expense while maintaining filter accuracy.
Solution Approach 2:
The patent performs preliminary action by conducting an initial search at a lower resolution to identify the general region containing the optimal tap values. This preliminary search establishes a threshold distance that guides subsequent higher-resolution searches, preventing exhaustive exploration of the entire high-resolution space while ensuring the optimal region is captured.
2Manufacturing precision
If the number of tap values is increased to improve filter characteristics, then the desired filter accuracy is achieved, but the time required to configure the filter increases
Solution Approach 1:
The configuration process is segmented into iterative stages where each stage searches at a specific resolution level. The search space is divided into regions based on the threshold distance from the best vector found at the previous stage. This segmentation allows the system to quickly identify the optimal region at coarse resolution and then efficiently refine the solution, significantly reducing total configuration time compared to exhaustive high-resolution search.
Solution Approach 2:
The patent performs preliminary configuration at lower resolution to establish the optimal region before committing to high-resolution configuration. This preliminary action at reduced computational cost guides the subsequent detailed configuration, ensuring that time is spent primarily in the most promising regions of the parameter space rather than uniformly across all possible tap value combinations.
Data Source
AI summary
Systems and methods are provided for determining effective tap values for a digital filter. A first plurality of vectors is generated, wherein each of the first plurality of vectors represents a set of tap values for the filter at a first resolution. A best vector is selected from the first plurality of vectors according to a first performance metric. A second plurality of vectors is generated within a threshold distance of the selected best vector, wherein each of the second plurality of vectors represents a set of tap values for the filter at a second resolution that is superior to the first resolution. A best vector is selected from the second plurality of vectors according to a second performance metric.