Multi-tone Signal Discriminator Using Goertzel Algorithm
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing signaling technologies face challenges in distinguishing between machine-generated dual-tone multi-frequency (DTMF) signals and voice-simulated DTMF signals, particularly due to variations in frequency and amplitude caused by voice compression techniques like Code Excited Linear Prediction (CELP), which can mimic machine-generated signals.
Innovation Solution
A dual-tone signal discriminator apparatus and method that compares output profiles from filters preset to identified multi-tone frequencies, using the Goertzel algorithm and Schwarz Inequality to determine if an input sample stream contains machine-generated or voice-simulated DTMF signals by assessing the inequality degree between filter outputs and adjusting thresholds based on signal persistence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If voice compression techniques like CELP are used to generate digital signals, then voice signal transmission efficiency is improved, but the ability to distinguish between machine-generated DTMF signals and voice-simulated signals deteriorates
Solution Approach 1:
The patent segments the analysis of DTMF signals by separating machine-generated signals from voice-simulated signals through spectral characteristics. It divides the frequency spectrum into tone pairs and analyzes the spectral envelope shape, allowing the system to distinguish between the two signal types even when both produce similar audible tones.
Solution Approach 2:
The patent uses spectral envelope shaping to create distinctive 'color' or fingerprint characteristics for different signal types. Machine-generated DTMF signals have a different spectral distribution compared to voice-simulated signals, and the system exploits these spectral differences to achieve reliable discrimination.
2Quantity of substance
If voice signals with constant pitch frequency and low modulation are transmitted, then transmission bandwidth is reduced, but the ability to differentiate from machine-generated DTMF signals deteriorates
Solution Approach 1:
The patent changes the analysis parameters from simple frequency detection to spectral envelope shape analysis. By examining the distribution of spectral energy across frequency bins and comparing it against expected patterns for machine-generated versus voice-simulated signals, the system achieves accurate identification even with limited bandwidth signals.
3Reliability
If DTMF signal discrimination is implemented to distinguish machine-generated from voice-simulated signals, then signal authentication reliability is improved, but system complexity increases
Solution Approach 1:
The patent applies partial action by focusing analysis only on the spectral envelope characteristics of the DTMF signal rather than analyzing the entire voice signal. This selective approach provides sufficient discrimination capability while keeping computational complexity manageable.
Solution Approach 2:
The patent creates simplified spectral models or templates representing machine-generated and voice-simulated DTMF signals. By comparing the actual signal against these pre-established models, the system achieves reliable authentication without requiring complex real-time analysis of all signal parameters.
Data Source
AI summary
In one embodiment, a method for discriminating between a machine generated multi-tone signal and a simulated voice multi-tone signal is provided. The method may comprise comparing output profiles generated from sampled outputs of a plurality of filters. The plurality of filters may have a single input sample stream applied to them and each filter may be preset at a measured multi-tone frequency associated with an identified prospective multi-tone signal. Based on the comparison of the output profiles, an inequality degree between the output profiles is generated and compared to an inequality threshold, thereby to determine whether the input sample stream comprises a machine generated multi-tone signal.