Digital Signal Processing Pipeline ? DSP Case Study

> Case Study Code: ASCOOS-OS-CASESTUDY-SEC02786

This case study demonstrates how Ascoos OS and the TDigitalCircuitHandler class implement a complete digital signal processing (DSP) pipeline in pure PHP, without external libraries.

The pipeline includes:

• generation of clean and noisy signals
• Gaussian or Uniform noise injection
• FIR & IIR low?pass filtering
• Hann?windowed FFT
• SNR, THD, SINAD, ENOB estimation
• latency estimation
• JSON & CSV export
• logging and debugging

The result is a complete, educational, and production?ready DSP implementation suitable for IoT, embedded simulation, audio pipelines, and academic use.

Purpose

• Generate a clean 440 Hz sine wave
• Add noise (Gaussian or Uniform)
• Apply FIR & IIR low?pass filters
• Perform FFT using a Hann window
• Detect peak frequency
• Compute: - SNR (input, FIR, IIR) - SNR improvement (gain) - THD (Total Harmonic Distortion) - SINAD (Signal?to?Noise And Distortion) - ENOB (Effective Number Of Bits)
• Export results to JSON & CSV
• Log all metrics for analysis

Core Ascoos OS Classes

DSP Core

• TDigitalCircuitHandler Main DSP class: filters, FFT, SNR, Gaussian noise, THD/SINAD/ENOB.

DSP Utilities

• gaussianNoise() Generates Gaussian noise using Box?Muller with caching
• findClosestBin() Finds the nearest FFT bin within tolerance
• estimateThdSinadEnob() Computes THD, SINAD, and ENOB from FFT data

Export & Logging

• toJSONFile() Exports the full DSP pipeline to JSON
• toCSVFile() Exports FFT spectrum & time?domain signals
• Logger Records all metrics and file paths

File Structure

The implementation is located in:

• digital_circuit_dsp.php

This file includes:

• signal generation
• noise injection (Gaussian/Uniform)
• FIR/IIR filtering
• Hann?windowed FFT
• SNR, THD, SINAD, ENOB
• latency estimation
• JSON/CSV export
• logging

Architecture Diagram

????????????????????????????????????????????????????????????????????????????????
?                              ASCOOS OS KERNEL                                ?
?                     (Electronics ? DSP ? Math Engine)                        ?
????????????????????????????????????????????????????????????????????????????????
                                      ?
                                      ?
                    ????????????????????????????????????????
                    ?      TDigitalCircuitHandler          ?
                    ?  (Filters ? FFT ? Noise ? Metrics)   ?
                    ????????????????????????????????????????
                                      ?
                                      ?
         ????????????????????????????????????????????????????????????????????
         ?                         DSP Pipeline                             ?
         ????????????????????????????????????????????????????????????????????
                                      ?
                                      ?
         ????????????????????????????????????????????????????????????????????
         ? 1. Signal Generation (Clean + Noise)                             ?
         ? 2. FIR Filtering                                                 ?
         ? 3. IIR Filtering                                                 ?
         ? 4. Hann Window                                                   ?
         ? 5. FFT (Magnitude Spectrum)                                      ?
         ? 6. SNR / THD / SINAD / ENOB                                      ?
         ? 7. Latency Estimation                                            ?
         ? 8. JSON / CSV Export                                             ?
         ? 9. Logging                                                       ?
         ????????????????????????????????????????????????????????????????????
                                      ?
                                      ?
????????????????????????????????????????????????????????????????????????????????
?                           Unified DSP Output Renderer                        ?
?      (Console Summary ? JSON Data ? CSV Files ? Logs)                        ?
????????????????????????????????????????????????????????????????????????????????

Requirements

1. PHP ? 8.3
2. Installed Ascoos OS or AWES 26 Pro
3. Enabled classes `TDigitalCircuitHandler`, `TCircuitHandler`, and `TElectronicsHandler` via the Extras Classes Manager
4. Access to: - `$AOS_TMP_DATA_PATH` - `$AOS_LOGS_PATH`

Execution Flow

1. Initialize the DSP Handler Loads `TDigitalCircuitHandler` with logging enabled.
2. Generate clean signal 440 Hz sine wave, 1024 samples.
3. Inject noise - Gaussian (Box?Muller) - or Uniform
4. Apply FIR filter 3?tap moving average.
5. Apply IIR biquad Butterworth filter 2nd?order, cutoff 1800 Hz.
6. Apply Hann window Reduces spectral leakage.
7. Perform FFT Computes magnitude spectrum.
8. Detect peak frequency Finds the dominant spectral component.
9. Compute SNR Input, FIR, IIR + SNR gain.
10. Compute THD, SINAD, ENOB Using tolerance to handle leakage.
11. Export JSON Full dataset + metadata.
12. Export CSV FFT spectrum + time?domain signals.
13. Logging Records all metrics and file paths.

Code Example

$dsp = new TDigitalCircuitHandler([], $properties);

$noise = $dsp->gaussianNoise(0.03);
$fir   = $dsp->applyFIRFilter($firCoeffs, $noisySignal);
$iir   = $dsp->applyBiquadFilter($biquadCoeffs, $noisySignal);

$fft   = $dsp->fft($windowed, true);
$mag   = array_map([$dsp, 'complexMagnitude'], $fft);

$metrics = $dsp->estimateThdSinadEnob($fftFreqs, $fftMags, 440.0);

Expected Output

=== DSP Simulation Results ===

Peak frequency (FFT): 440.0 Hz
Max FFT magnitude   : 0.9981

SNR (input noisy)   : 29.52 dB
SNR (after FIR)     : 33.87 dB
SNR (after IIR)     : 35.12 dB

THD (approx)        : -42.1 dB
SINAD (approx)      : 33.5 dB
ENOB (approx)       : 5.27 bits

JSON exported to: /tmp/signal_dsp_demo.json
FFT CSV exported to: /tmp/signal_dsp_fft_spectrum.csv
Time CSV exported to: /tmp/signal_dsp_time_signals.csv

Resources

• Ascoos OS Documentation (Under Construction)
• Ascoos OS Website (Under Construction)
• Ascoos
• AWES Studio
• GitHub Repository

License

This case study is covered under the Ascoos General License (AGL). See LICENSE.md.