Speech Packet Multiplexing With Redundant Coding for Frame Loss
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional speech coding methods experience significant quality deterioration and additional transmission delay when consecutive packet losses occur, as the internal states of the coder and decoder become mismatched, leading to propagated errors in decoded speech.
Innovation Solution
The system transmits redundant coded data to synchronize the internal states of the coding and decoding apparatus after a frame loss, preventing error propagation and improving subjective speech quality without additional delay by multiplexing and packetizing first and second coded information, and using this redundant information for decoding processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If coded data of two consecutive frames are multiplexed into one packet, then transmission efficiency is improved, but transmission delay increases by one frame
Solution Approach 1:
The coded data is segmented into two types: normal coded data (first coded information) and reset coded data (second coded information). The normal coded data is transmitted in multiplexed packets for efficiency, while reset coded data is transmitted separately to synchronize internal states without adding delay to the main data stream.
Solution Approach 2:
Reset coded data acts as an intermediary mechanism. It is transmitted separately from the main multiplexed data stream to serve as a synchronization signal that resets the internal states of the speech coding/decoding apparatus, preventing error propagation without disrupting the efficient multiplexed transmission of normal data.
2Device complexity
If packet loss concealment processing is carried out independently on the decoder side, then processing complexity on the transmitter is reduced, but quality of decoded speech deteriorates when consecutive packets are lost
Solution Approach 1:
The transmitter performs preliminary action by generating and transmitting reset coded data in advance. This reset coded data is prepared and sent with the multiplexed data stream, so that when packet loss occurs, the receiver can use the pre-transmitted reset coded data to synchronize internal states and prevent error propagation, improving speech quality without adding complexity to the receiver's concealment processing.
3Reliability
If reset coded data is transmitted for every frame, then synchronization reliability is improved, but transmission bandwidth is wasted
Solution Approach 1:
Instead of uniformly transmitting reset coded data for every frame, the system applies local quality by selectively transmitting reset coded data only when necessary. The transmitter determines whether to generate reset coded data based on the current frame's characteristics and transmission conditions, optimizing the balance between synchronization reliability and bandwidth utilization.
Data Source
AI summary
A speech signal transmission apparatus multiplexes, packetizes, and sends first coded information coded in a normal state and second coded information used for improving the quality of decoded speech when a frame loss occurs. A first error calculating section calculates a first error signal between a target signal and a synthesized signal generated by an adaptive codebook, and a second error calculating section calculates a second error signal between the target signal and a synthesized signal generated by a fixed codebook. An error signal ratio calculating section calculates the ratio of the first error signal to the second error signal. A speech frame classifying section classifies a speech frame according to the magnitude of the ratio, and a decision section decides whether or not to multiplex the second coded information based on the classification result.