Continuous-Time Energy Estimation From Non-Uniform Analog Sampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for calculating the energy of an analog signal consume high power, especially in sparse signal conditions like voice signal processing, due to constant clock frequency-based discrete-time digital signal processing, and continuous-time non-uniform amplitude sampling lacks a time reference for energy calculation.
Innovation Solution
A device comprising a continuous-time analog-to-digital converter, filtering units, and energy estimators that convert the analog signal into a request signal with pulses and direction signals, performing finite impulse response filtering and combining delayed signals to calculate energy using a non-uniform time base, reducing power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If discrete-time digital signal processing with clock sampling is used to calculate energy from analog signal, then energy calculation accuracy is maintained, but power consumption increases
Solution Approach 1:
The patent uses periodic clock cycles to control when processing operations occur. The energy calculation is performed periodically based on signal energy thresholds rather than continuously, reducing power consumption while maintaining accuracy when needed. The system activates processing only during relevant time periods when signal energy exceeds predetermined thresholds.
Solution Approach 2:
The system dynamically adjusts its operation mode between continuous-time and discrete-time processing based on signal characteristics. When signal energy is high, it uses continuous-time processing for efficiency. When signal energy is low or below thresholds, it switches to discrete-time processing to maintain accuracy, optimizing the balance between power consumption and measurement precision adaptively.
2Use of energy by moving object
If continuous-time non-uniform amplitude sampling is used to decrease power consumption, then power consumption is reduced, but time reference is lost making energy calculation impossible
Solution Approach 1:
The patent introduces an intermediary energy calculation unit that bridges continuous-time sampling and energy measurement. This unit accumulates energy values from continuous-time samples and provides a computed energy metric that restores the time reference information needed for energy calculation, allowing the system to maintain continuous-time processing benefits while recovering measurement capability.
Solution Approach 2:
The system performs preliminary energy accumulation and threshold comparison operations on continuous-time samples before full discrete-time processing. By pre-calculating energy values and comparing against thresholds in advance, the system determines when detailed energy calculation is necessary, enabling continuous-time processing to reduce power while maintaining energy calculation accuracy when required.
3Measurement precision
If discrete-time digital signal processing is used for sparse voice signals, then energy calculation is possible, but power consumption is wasteful during silence periods
Solution Approach 1:
The patent implements periodic threshold-based activation where discrete-time processing is only activated when signal energy exceeds predetermined thresholds. During silence periods when energy remains below thresholds, processing is suspended entirely, eliminating power waste while maintaining energy calculation capability when speech occurs. This creates a periodic processing pattern aligned with actual signal relevance.
Solution Approach 2:
The system applies partial processing by performing only threshold comparison and energy accumulation during silence periods, rather than full discrete-time processing. When energy exceeds thresholds, full processing is activated. This partial action during low-activity periods reduces power consumption significantly while maintaining the capability to perform complete energy calculation when needed, avoiding excessive processing during unnecessary time periods.
Data Source
AI summary
Device (1), for continuous-time energy calculation of an analog signal, comprising:a continuous-time analog-to-digital converter which is configured to convert the analog signal into a request signal (REQ), and a direction signal (DIR);at least one filtering unit (11), configured to output a filtered output signal (Fout), and comprising a delaying module (12) and a calculating module (15), connected to the delaying module (12) and configured calculate the filtered output signal (Fout).According to the invention, the device (1) further comprises:at least one pulse combiner (16), connected to the delaying module (12) and configured to output a combined request signal (CREQ); andat least one energy estimator (17), connected to the filtering unit (11) and to the pulse combiner (16), configured to compute a stored energy value (Aout) associated with each pulse of the combined request signal (CREQ).