A Channel Coding Scheme for 2400 bps Speech Coding Algorithm
Ye Li, Jie Zhang, Xiaofei Gong, Peng Zhang · 2019
In recent years, low-rate speech coding algorithms have been extensively and deeply studied, but as the coding rate decreases, the robustness of the speech coding algorithm to channel error is getting worse. In order to improve the robustness of low-rate speech coding algorithm to channel error, this paper proposes a channel coding algorithm based on source characteristics. The algorithm is based on MELP 2400 bps speech coding algorithm, and analyzes the importance difference between the coded bits and the synthesized speech quality, which are divided into important bits, sub-important bits and general bits. Different BCH channel coding schemes are employed for the three types of coded bits. Simulation experiments show that the channel coding scheme combined with bit importance has better anti-error effect under different random error rates.