DSSS Signal Detection Based on CNN.

Area of Science:

• Electrical Engineering
• Signal Processing
• Artificial Intelligence

Background:

• Direct sequence spread spectrum (DSSS) signals are widely used, complicating electronic reconnaissance in communication countermeasures.
• Traditional DSSS signal detection relies on autocorrelation algorithms, which are mature but have limitations.
• Deep learning methods are increasingly being applied to signal processing tasks.

Purpose of the Study:

• To propose and evaluate a novel deep learning-based method for DSSS signal detection.
• To compare the performance of the proposed method against traditional autocorrelation detection algorithms.
• To assess the effectiveness of the method under various signal conditions.

Main Methods:

• Development of a convolutional neural network (CNN) model for DSSS signal detection.
• Experimental analysis comparing the CNN model with the autocorrelation detection algorithm.
• Evaluation across different signal-to-noise ratios, spreading code lengths, spreading code types, and modulation methods.

Main Results:

• The proposed CNN model achieved higher detection performance compared to the traditional autocorrelation method.
• The overall performance improvement of the CNN model was found to be 4 dB.
• The model demonstrated robust estimation performance under diverse signal parameters.

Conclusions:

• Convolutional neural networks offer a promising alternative to traditional methods for DSSS signal detection.
• The developed CNN model provides enhanced electronic reconnaissance capabilities for DSSS signals.
• Deep learning significantly improves the detection performance in challenging signal environments.