Scalable Audio Codec Bit Allocation for Packet Loss

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current audio codecs in conferencing applications face challenges with scalability, computational resource usage, signal latency, and quality degradation due to varying network conditions and packet loss, especially in wireless networks and scenarios with limited computational resources.

Innovation Solution

A scalable audio codec dynamically allocates bits between frequency bands based on power levels and perceptual modeling, allowing for flexible bit reordering and stripping to maintain audio quality, and encodes once for both high and low bit rates, reducing computational resources and bandwidth requirements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If traditional audio codecs encode audio at fixed bit rates, then audio quality is maintained, but scalability to varying network conditions is poor

Engineering Contradiction:
Improvescalability to varying bit ratesVSAvoidaudio quality consistency
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The audio frequency spectrum is segmented into multiple bands (e.g., low band and high band), and bits are dynamically allocated to different bands based on network conditions and perceptual importance. This allows the codec to scale from low to high bit rates by selectively including or excluding certain frequency bands while maintaining audio quality.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The bit allocation between different frequency bands is dynamically adjusted based on network conditions, packet loss rates, and perceptual modeling. The codec can transition between different operating modes (e.g., wideband, super wideband, fullband) to adapt to varying network conditions while maintaining consistent audio quality.

Inventive Principle:
Principle #15Dynamics

2Adaptability or versatility

If audio is encoded separately for multiple bit rates, then adaptability to network conditions improves, but computational resource usage increases

Engineering Contradiction:
Improvebit rate flexibilityVSAvoidcomputational resource usage
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The codec merges multiple bit rate representations into a single encoding structure. By encoding the audio once with scalable bit allocation across frequency bands, the system generates bit streams that can be truncated to different bit rates without requiring separate encodings, thereby reducing computational complexity while maintaining bit rate flexibility.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The scalable audio codec provides multi-functionality by serving multiple bit rate requirements from a single encoding process. The same encoded bit stream can be used for different network conditions and receiver capabilities, making the codec universal across varying operational requirements without increasing device complexity.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Loss of energy

If bits are stripped from encoded audio data, then bandwidth requirements are reduced, but audio quality degrades

Engineering Contradiction:
Improvebandwidth consumptionVSAvoidaudio quality
Core Design Contradiction:
Loss of energyVSReliability

Solution Approach 1:

The codec applies local quality preservation by selectively allocating bits to frequency bands based on their perceptual importance and local signal characteristics. When bits need to be stripped to reduce bandwidth, the codec prioritizes removing bits from less important frequency regions while maintaining quality in critical bands, thereby reducing bandwidth consumption with minimal audio quality degradation.

Inventive Principle:
Principle #3Local quality

4Loss of energy

If transform coding is used to compress audio, then bandwidth efficiency improves, but handling packet loss and varying network conditions becomes more difficult

Engineering Contradiction:
Improvebandwidth efficiencyVSAvoidnetwork condition adaptability
Core Design Contradiction:
Loss of energyVSAdaptability or versatility

Solution Approach 1:

The transform coding parameters are made dynamic to adapt to network conditions. The codec can adjust the transform type, block size, and bit allocation in real-time based on network conditions and packet loss patterns, allowing efficient bandwidth utilization while maintaining adaptability to varying network environments.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP2402939B1Full-band scalable audio codec
Publication Date: 2023.04.26 HEWLETT PACKARD DEVELOPMENT COMPANY LP
  • EP2402939B1 patent drawingFigure 1A~1B
  • EP2402939B1 patent drawingFigure 2A~2B
  • EP2402939B1 patent drawingFigure 3

AI summary

A scalable audio codec for a processing device determines first and second bit allocations for each frame of input audio. First bits are allocated (318) for a first frequency band, and second bits are allocated (318) for a second frequency band. The allocations are made on a frame-by-frame basis based on the energy ratio between the two bands. For each frame, the codec transform codes (320) both frequency bands into two sets of transform coefficients, which are then packetized based on the bit allocations. The packets are then transmitted (324) with the processing device. Additionally, the frequency regions of the transform coefficients can be arranged (322) in order of importance determined by power levels and perceptual modeling. Should bit stripping occur, the decoder at a receiving device can produce audio of suitable quality given that bits have been allocated between the bands and the regions of transform coefficients have been ordered by importance.