| 摘要 |
<p>A method and apparatus for providing feedback from the decoder to the encoder to improve performance in a predictive speech coder under frame erasure conditions includes notifying an encoder (100, 106) in a receiving speech coder if a decoder (104, 110, 200, 300) in the receiving speech coder fails to receive a frame transmitted by an encoder (100, 106) in a transmitting speech coder. A modified packet is transmitted from the encoder (100, 106) in the receiving speech coder to a decoder (104, 110, 200, 300) in the transmitting speech coder in response to notification. The decoder (100, 106) in the transmitting speech coder notifies the encoder (100, 106) in the transmitting speech coder when the modified packet is received. The encoder (100, 106) in the transmitting speech coder then encodes a packet with a modified encoding format. The modified encoding format may be either a low-memory encoding format or a memoryless encoding format. The modified packet may have an erasure indicator bit set to a digital value of one.</p> |
