Redundancy-Based Packet Error Recovery in Wireless Speech
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current speech coding systems face challenges in maintaining high voice quality at low bit rates, particularly in wireless communication systems, where packet loss and errors during transmission lead to degraded audio quality and user experience due to the limitations of existing error recovery mechanisms.
Innovation Solution
A system and method for redundancy-based packet transmission error recovery, which includes a destination device with an analyzer that utilizes partial copies of packets as error correction data to reconstruct speech signals, dynamically adjusting buffer depth and retransmission thresholds based on packet loss and redundancy metrics to minimize user-perceptible artifacts and delay.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If speech is transmitted by sampling and digitizing at standard rates, then speech quality comparable to analog telephone is achieved, but data rate consumption increases to 64 kbps
Solution Approach 1:
The patent extracts and transmits only the essential speech parameters (spectral envelope, pitch, gain) rather than the full speech signal. By identifying and transmitting only the critical elements that define speech quality, the system achieves high speech quality at much lower data rates, resolving the contradiction between quality and data rate consumption
Solution Approach 2:
The patent transforms the speech signal from time-domain samples to frequency-domain parameters through spectral analysis. By changing the representation parameters from raw samples to extracted speech features (formants, pitch contours, gain), the system reduces data rate while maintaining speech quality, as these parameters capture the essential characteristics of speech with fewer bits
2Productivity
If packet transmission is used in wireless systems, then data rate and capacity are improved, but packet loss and errors occur leading to degraded audio quality
Solution Approach 1:
The patent prepares error correction data and partial packet copies in advance before transmission. By pre-processing and embedding redundancy information within the packet structure beforehand, the system enables rapid error recovery at the receiver without requiring retransmission, thus maintaining high data rate while improving reliability and audio quality despite packet loss
Solution Approach 2:
The patent creates and transmits partial copies of packets containing error correction data. When packet loss occurs, these copied portions can be used to reconstruct the original speech parameters, maintaining audio quality without requiring retransmission of the entire packet, thus resolving the contradiction between high data rate and reliable audio delivery
3Reliability
If existing error recovery mechanisms are used, then some packet loss is handled, but user-perceptible artifacts and degradation persist
Solution Approach 1:
The patent implements adaptive error recovery that uses feedback from the received packets to determine the appropriate error correction strategy. By analyzing packet loss patterns and using feedback from partial packet reception, the system dynamically adjusts error correction application to minimize audible artifacts while maintaining speech quality, resolving the contradiction between error recovery effectiveness and artifact reduction
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A device includes a receiver, a buffer, and an analyzer. The receiver is configured to receive packets that correspond to at least a subset of a sequence of packets and that include error correction data. The error correction data of a first packet of the packets includes a partial copy of a second packet. The buffer is configured to store the packets. The analyzer is configured to determine whether a first particular packet of the sequence is missing from the buffer, to determine whether a partial copy of the first particular packet is stored in the buffer as error correction data in a second particular packet, to update a value based at least in part on whether the first particular packet is missing from the buffer and the partial copy of the first particular packet is stored in the buffer, and to adjust an error recovery parameter based at least in part on the value.