Channel coding of FED STD MELP vocoder utilizing residual redundancy features

The US government has developed, adopted, and promulgated a new Federal Standard vocoder algorithm which operates at 2400 bps and is called MELP-mixed excitation linear prediction. This vocoder has reasonably good voice quality under benign error channel conditions. However, when it is subjected to high error conditions as may be experienced in mobile vehicular applications, correction schemes may be employed which take full advantage of the inherent inter-frame residual redundancy of the MELP parameters. This describes experiments conducted on the vocoder algorithm in combination with Viterbi convolutional error decoding, and enhanced with maximum a posteriori techniques which capitalize on these redundancy characteristics. Both hard and soft Viterbi decoding implementations are investigated.