Predictive Level-Crossing Sampling for Fiducial Point Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing signal processing methods for wearable and miniaturized devices face challenges in detecting fiducial points of input signals while maintaining low circuit complexity, and are susceptible to noise and unnecessary sampling due to high-amplitude low-frequency baseline wandering and low-amplitude high-frequency noise.
Innovation Solution
A second-order level-crossing sampling method that predicts key sampling points using digital feedback, applies dynamic thresholds, and includes neighbor amplitude and slope filters to remove unnecessary samples, reducing computing overhead and noise effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If first-order level-crossing sampling is used, then power consumption is reduced and unnecessary sampling is avoided, but the method cannot detect fiducial points and is susceptible to noise and baseline wandering
Solution Approach 1:
The patent implements dynamic threshold adjustment where the sampling threshold adapts to the local characteristics of the input signal. The threshold is no longer fixed but changes based on the signal's amplitude and slope variations, allowing the system to maintain low power consumption while accurately detecting fiducial points across different signal conditions.
Solution Approach 2:
The patent changes the sampling parameters dynamically by introducing slope-based threshold adjustment. The sampling threshold is modified according to the slope of the input signal, enabling the system to distinguish between significant signal changes (fiducial points) and minor variations (noise), thus improving detection accuracy without increasing power consumption proportionally.
2Measurement precision
If second-order delta modulation is used, then fiducial points can be detected, but the method is susceptible to high frequency noise and increases circuit complexity
Solution Approach 1:
The patent extracts only the essential information needed for fiducial point detection by using slope-based threshold comparison rather than full second-order delta modulation. This approach captures the critical slope variation information while discarding redundant data, achieving fiducial point detection with reduced circuit complexity.
Solution Approach 2:
The patent uses simple comparator circuits and basic arithmetic operations instead of complex second-order delta modulation circuits. The solution employs readily available low-complexity components to achieve the desired functionality, avoiding the need for expensive and complex circuit implementations.
3Loss of information
If Nyquist sampling is used, then all signal information is captured, but unnecessary data is generated increasing processing workload
Solution Approach 1:
The patent applies partial sampling by using slope-based thresholds to sample only when significant signal changes occur. Instead of continuous Nyquist sampling, the system performs sampling selectively based on signal characteristics, capturing essential information while generating fewer data points for processing.
Solution Approach 2:
The patent performs preliminary slope calculation and threshold comparison before actual sampling occurs. This preliminary action identifies which samples are worth taking, allowing the system to capture necessary signal information while avoiding unnecessary sampling and subsequent processing of redundant data.
4Productivity
If slope-tracking method is used, then sampling can be optimized, but the method introduces accumulated errors and involves complicated circuit structures
Solution Approach 1:
The patent implements feedback mechanisms where the sampling decision is based on comparing the current slope with dynamically adjusted thresholds. This feedback loop prevents accumulated errors by continuously adapting to signal characteristics and correcting sampling decisions, ensuring reliable fiducial point detection while maintaining sampling efficiency.
Data Source
AI summary
A method and apparatus that provides a digital approximation of an input analog continuous-time signal using piece-wise linear waveform from key sampling points selected by a dynamic predictive sampling method. The apparatus includes a processing system or block to generate a dynamic prediction of the input signal as well as an upper threshold and a lower threshold to form a tracking window, a comparator that compares the input signal with the upper threshold and the lower threshold to determine if the prediction is successful, a counter to record timestamps between the unsuccessful predictions which are the selected key sampling points, and a processing block controlling predictions and sampling states of the system. The processing of the sampling points can include a neighbor amplitude filter and a slope filter.