Vector Comparator System for Peak Value Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing signal processing applications face challenges in efficiently finding maximum or minimum numbers in a set due to sequential processing, limited data size capabilities, and high processing complexity, which hinders real-time detection and location identification of peak values.
Innovation Solution
A vector comparator system with a parallel architecture, comprising a vector comparator unit and an index generation unit, that automatically adapts operational modes to process varying data sizes using page and element modes, enabling efficient reduction of the search set and tracking of the winning index through vector comparison and tree reductions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If sequential processing is used to find peak values, then processing complexity is reduced, but processing speed and real-time detection capability deteriorate
Solution Approach 1:
The patent divides the set of N numbers into multiple groups and uses multiple comparator units to process different groups simultaneously in parallel. This segmentation approach enables concurrent processing of multiple data elements, significantly improving processing speed while maintaining manageable complexity through modular architecture
Solution Approach 2:
The patent transitions from sequential one-dimensional processing to multi-dimensional parallel processing by introducing multiple comparator units operating simultaneously. This dimensional expansion allows the system to evaluate multiple candidates in each clock cycle, achieving real-time detection capability without proportionally increasing overall system complexity
2Device complexity
If fixed data size is used with comparator unit width limitation, then device complexity is reduced, but adaptability to varying data sizes deteriorates
Solution Approach 1:
The patent employs dynamic configuration where the number of active comparator units and grouping strategies can be adjusted based on the actual data size N. This dynamic adaptability allows the system to optimize its operational parameters for different input sizes without requiring a complete redesign, maintaining both complexity efficiency and versatility
Solution Approach 2:
The patent creates a universal comparator system that can handle various data sizes by configuring multiple comparator units to work in parallel. The same architectural framework adapts to different N values by adjusting the distribution of data across comparators, providing multi-functionality without requiring separate dedicated circuits for each data size
3Speed
If parallel processing with multiple comparator units is used, then processing speed and real-time detection capability are improved, but processing complexity and resource requirements increase
Solution Approach 1:
The patent segments the processing task across multiple comparator units, each handling a specific group of candidates. This segmentation distributes the computational load and allows independent operation of each unit, improving processing speed while keeping individual unit complexity low and manageable through modular design
Solution Approach 2:
The patent introduces control logic as an intermediary that coordinates the parallel comparator units, manages candidate selection, and determines winning candidates. This intermediary layer simplifies the overall system by providing centralized control over the parallel operations, reducing the complexity burden that would otherwise arise from managing multiple independent comparators
4Quantity of substance
If sequential processing of one data point at a time is used, then resource requirements are reduced, but productivity and processing throughput deteriorate
Solution Approach 1:
The patent divides the input data into segments that can be processed simultaneously by multiple comparator units. Each comparator unit processes a portion of the data in parallel, increasing the number of data points evaluated per clock cycle and thereby improving processing throughput while using a proportional increase in comparator resources
Solution Approach 2:
The patent transitions from single-thread sequential processing to multi-thread parallel processing by introducing multiple comparator units operating concurrently. This dimensional change in processing architecture enables multiple data points to be evaluated simultaneously, dramatically improving productivity and throughput without requiring exponential resource increases
Data Source
AI summary
A comparator (231) for determining a peak number, representing a maximum or minimum of a set of numbers, includes a multi-element comparator (232) for comparing different pages of the set of numbers in a page comparison mode to output a candidate set of winning numbers, and for automatically switching to a leaf/tree search of the candidate set of winning numbers in an element comparison mode. Operating in parallel with the multi-element comparator (232), an index generation unit (233) processes flag/sign bits from the multi-element comparator in conjunction with state machine control logic (230) to keep track of the index/indices for the peak value. Upon completion of final stage, the index generation unit returns the absolute index (235) of the peak value.