Voice Signal Encoding for Lossless Compression of Near-Zero PCM Frames
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current compression coding methods for signals encoded with non-uniform PCM, such as those used in VoIP systems, fail to provide sufficient compression during non-active speech periods due to high probability occurrences of specific code words like +0 and -0 or +8 and -8, leading to increased transmission capacity requirements.
Innovation Solution
Assigning unique code words to the two smallest quantization intervals and employing lossless encoding techniques like run-length encoding and integer encoding to compress frames containing these code words, allowing for reversible compression and reduced data transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If non-uniform PCM encoding is used for voice signals, then sound quality is maintained, but transmission capacity requirement increases
Solution Approach 1:
The patent applies parameter changes by transforming the encoding approach from standard non-uniform PCM to a modified encoding scheme that identifies and specially processes frames containing only the two smallest quantization interval code words. This parameter change in encoding strategy reduces the average number of bits per frame while maintaining the ability to reconstruct the original signal, thereby reducing transmission capacity requirements without sacrificing sound quality.
2Productivity
If compression coding is applied to reduce transmission capacity, then data transmission efficiency improves, but lossless encoding performance deteriorates
Solution Approach 1:
The patent segments the encoding process into distinct cases: frames containing only the two smallest quantization interval code words are processed with specialized lossless compression, while other frames use standard encoding. This segmentation allows the system to apply aggressive compression only where safe (when background noise is present and signal is near zero), thereby improving overall data transmission efficiency while maintaining lossless encoding performance for all frames.
Solution Approach 2:
The patent applies partial compression action by selectively compressing only those frames that meet specific criteria (containing only the two smallest quantization interval code words), rather than applying compression to all frames. This partial action approach ensures that compression is applied only when it will not compromise lossless reconstruction, thus improving transmission efficiency without sacrificing encoding fidelity.
3Productivity
If special handling is applied to frequent code words like +0 and -0, then compression ratio improves, but encoding complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-identifying and flagging frames that contain only the two smallest quantization interval code words before compression. This preliminary identification allows the encoder to prepare appropriate compression parameters and methods in advance, simplifying the overall encoding process despite the special handling required for these frequent code words. The complexity is managed by preparing compression strategies beforehand rather than computing them in real-time during encoding.
Data Source
AI summary
A frame formed of a plurality of code words encoded with an encoding mode in which two different types of code words are assigned one-to-one to two smallest quantization intervals is checked to determine whether it contains just the two types of code words assigned to the two smallest quantization intervals, and lossless encoding is applied to the frame containing just the two types of code words. A code obtained by this lossless encoding is decoded with a decoding method corresponding to the lossless encoding.