Design of robust communication systems using genetic algorithms

This paper presents a novel genetic algorithm for jointly optimizing source and channel codes. The algorithm uses a channel-optimized vector quantizer for the source code, and a rate-punctured convolutional code for the channel code. The genetic algorithm enhances the robustness of the rate-distortion performance of the channel-optimized vector quantizer, and reduces the computational time for finding the best rate-punctured convolutional code. Numerical results show that the algorithm attains near optimal performance while having low computational complexity.