Modulation Recognition and Detection System

📌 Overview

This project focuses on automatic recognition and classification of modulated waveforms using machine learning. It generates, processes, and classifies signals like QPSK, 8PSK, 16APSK, 32APSK, and 64APSK using a RandomForestClassifier with feature extraction and visualization.

🚀 Features

✔ Automatic Modulation Recognition – Detects modulation type from input signals.
✔ Machine Learning Model – Uses RandomForestClassifier to classify waveforms.
✔ Feature Extraction – Extracts statistical and frequency-domain features from signals.
✔ Signal Visualization – Includes constellation diagrams, scatter plots, and confusion matrices.
✔ Interactive CLI – Allows users to train models, test signals, and view results.

🛠️ Technologies Used

• Python
• NumPy – Signal processing and numerical computations.
• Matplotlib & Seaborn – Data visualization.
• Scikit-learn – Machine learning model training.
• Joblib – Model saving and loading.

📂 Project Structure

Modulation_Recognition/ ├── 📄 modulation_signal_generator.py # Generates modulation signals and saves them to files ├── 📄 modulation_recognition.py # Trains ML model & classifies signals ├── 📂 sample_inputs/ # Stores generated signals ├── 📄 modulation_model.pkl # Trained model file ├── 📄 README.md # Documentation

📌 Project Description

This project focuses on Automatic Modulation Recognition (AMR) for DVB-S2X waveforms using machine learning.
It includes signal generation, training a classification model, and recognizing modulation types.

📜 Files and Directories

• modulation_signal_generator.py – Generates various modulation signals and stores them as files in sample_inputs/.
• modulation_recognition.py – Trains a machine learning model for modulation recognition and classifies signals.
• sample_inputs/ – Stores generated modulation signal files.
• modulation_model.pkl – The trained machine learning model for modulation classification.
• README.md – Project documentation.

🚀 How to Run

1. Generate modulation signals
   python modulation_signal_generator.py

🔧 Installation

git clone https://github.com/yourusername/Modulation_Recognition.git cd Modulation_Recognition pip install -r requirements.txt

🎯 Usage

1. Generate Modulation Signals bash Copy Edit python modulation_signal_generator.py 👉 Saves QPSK, 8PSK, 16APSK, 32APSK, and 64APSK signals as text files.

2. Train & Test the Model bash Copy Edit python modulation_recognition.py Option 1: Train a new model with custom sample size. Option 2: Test a new waveform signal file and get predictions.

📊 Model Performance

The trained RandomForestClassifier achieves an accuracy of 80% on test data.

📜 License

This project is open-source and licensed under the MIT License.