We proposed and evaluated a speech packet loss concealment method which predicts lost segments from speech included in packets either before, or both before and after the lost packet. The lost segments are predicted recursively by using linear prediction both in the forward direction from the packet preceding the loss, and in the backward direction from the packet succeeding the lost segment. Predicted samples in each direction are smoothed by averaging using linear weights to obtain the final interpolated signal. The adjacent segments are also smoothed extensively to significantly reduce the speech quality discontinuity between the interpolated signal and the received speech signal. Subjective quality comparisons between the proposed method and the the packet loss concealment algorithm described in the ITU standard G.711 Appendix I showed similar scores up to about 10% packet loss. However, the proposed method showed higher scores above this loss rate, with Mean Opinion Score rating exceeding 2.4, even at an extremely high packet loss rate of 30%. Packet loss concealment of speech degraded with G.729 coding, and babble noise mixed speech showed similar trends, with the proposed method showing higher qualities at high loss rates. We plan to further improve the performance by using adaptive LPC prediction order depending on the estimated pitch, and adaptive LPC bandwidth expansion depending on the consecutive number of repetitive prediction, among many other improvements. We also plan to investigate complexity reduction using gradient LPC coefficient updates, and processing delay reduction using adaptive forward/bidirectional prediction modes depending on the measured packet loss ratio.

Subjective quality tests have proven that embedded adaptive differential PCM (ADPCM), known to tolerate information loss through bit dropping, does not maintain sufficient speech quality when directly applied to asynchronous transfer mode (ATM) due to the fixed-length cell transmission scheme unique to ATM. We propose a coding and transmission scheme which enhances the performance by adjusting the embedded ADPCM coding rate according to input speech characteristics, thereby taking advantage of the ATM environment, where the transmission of variable rate sources is feasible. By varying the number of code bits of an embedded ADPCM coder from 6bits per sample, or 48kbps, for blocks of speech with a high prediction gain, to 2bits, or 16kbps, for silent blocks, a good compromise between coding bit rate and speech quality with gradual degradation due to information loss is achieved. The results of subjective evaluation tests showed the speech quality of the proposed scheme to be over 3.5 mean opinion score (MOS) on a scale of 1 to 5 at a cell loss rate of 10%. A prototype of the codec and the ATM cell assembly/disassembly functions were also fabricated using 3 conventional digital signal processors (DSPs) for real-time conversation tests.