Related Experiment Videos
On-line adaptive chaotic demodulator based on radial-basis-function neural networks
1Department of Electronic and Information Engineering, Hong Kong Polytechnic University, Hong Kong, China. fenghc@swnu.edn.cn
Summary
This study introduces an adaptive chaotic demodulator using a radial-basis-function (RBF) neural network for spread spectrum communication. The system effectively retrieves message signals from noisy channels by tracking chaotic carrier dynamics.
Area of Science:
- Electrical Engineering
- Signal Processing
- Chaos Theory
Background:
- Chaotic modulation offers advantages for spread spectrum communication systems.
- Developing robust demodulators for chaotic signals in noisy environments is crucial.
- Radial-basis-function (RBF) neural networks provide powerful approximation capabilities.
Purpose of the Study:
- To propose and design an on-line adaptive chaotic demodulator.
- To leverage RBF networks and extended Kalman filters for adaptive learning and tracking.
- To estimate the modulating parameter carrying message information from noisy spread spectrum signals.
Main Methods:
- Implementation of an on-line adaptive learning algorithm for the demodulator.
- Utilizing the approximation capability of RBF networks and the tracking ability of the extended Kalman filter.
- Employing the Henon map as the chaos generator and least-square fit for parameter estimation.
- Evaluating performance with square-wave, sine-wave, speech, and image signals.
Main Results:
- The proposed demodulator successfully tracks spread spectrum signals in additive white Gaussian noise.
- The modulating parameter, encoding message data, is accurately estimated.
- The system demonstrates the ability to track chaotic carrier dynamics.
- Message signals are effectively retrieved from noisy channels.
Conclusions:
- The on-line adaptive chaotic demodulator based on RBF networks is effective for spread spectrum communication.
- The integration of RBF networks and extended Kalman filters enables robust signal tracking and parameter estimation.
- The method is validated for various signal types, confirming its practical applicability in retrieving information from chaotic carriers in noisy conditions.