Voice Data Path Selection for Bandwidth Optimization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing call systems face challenges in maintaining high call quality due to increased network bandwidth and data traffic consumption, which slows down voice data transmission speeds.

Innovation Solution

A method where a server selects and transmits only the most relevant voice data paths based on voice activity detection and energy levels, reducing the number of paths transmitted to the receiving terminal, thereby optimizing network bandwidth usage and improving call quality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the server transmits all voice data paths to the receiving terminal, then complete voice communication is achieved, but network bandwidth occupation increases and transmission speed decreases

Engineering Contradiction:
Improvevoice communication completenessVSAvoidvoice data transmission speed
Core Design Contradiction:
ReliabilityVSSpeed

Solution Approach 1:

The patent extracts only the necessary voice data paths from the complete set of voice data. The server identifies and transmits only those voice paths that contain actual speech content, filtering out redundant paths where no one is speaking. This extraction approach maintains communication completeness while reducing transmission volume and improving speed.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

Instead of transmitting all voice data paths (excessive action), the system transmits only the necessary subset of paths containing speech content (partial action). This partial transmission approach is sufficient for achieving complete voice communication while significantly reducing network bandwidth occupation and improving transmission speed.

Inventive Principle:
Principle #16Partial or excessive action

2Loss of information

If the server transmits all voice data paths, then no voice information is lost, but data traffic consumption increases

Engineering Contradiction:
Improvevoice information completenessVSAvoiddata traffic consumption
Core Design Contradiction:
Loss of informationVSLoss of substance

Solution Approach 1:

The system extracts and transmits only the essential voice data paths that contain speech content, filtering out redundant paths. This extraction ensures that no actual voice information is lost while significantly reducing the total data traffic consumed during transmission.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system discards redundant voice data paths where no speech is occurring, and recovers/transmits only the necessary paths containing speech content. This selective discarding and recovering approach maintains voice information completeness while reducing overall data traffic consumption.

Inventive Principle:
Principle #34Discarding and recovering

3Reliability

If all voice data paths are transmitted, then comprehensive voice mixing is achieved, but network bandwidth occupation increases

Engineering Contradiction:
Improvevoice mixing qualityVSAvoidnetwork bandwidth occupation
Core Design Contradiction:
ReliabilityVSLoss of energy

Solution Approach 1:

The patent extracts only the necessary voice data paths containing speech content for transmission, rather than transmitting all paths. This extraction maintains sufficient voice mixing quality for accurate speech reproduction while significantly reducing network bandwidth occupation.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system transmits only the partial set of voice paths that are necessary for quality voice mixing, rather than transmitting all paths excessively. This partial transmission approach achieves adequate voice mixing quality while conserving network bandwidth resources.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP4016956B1Calling method, device and system, server and storage medium
Publication Date: 2024.07.31 TENCENT TECHNOLOGY (SHENZHEN) CO LTD
  • EP4016956B1 patent drawingFigure 1~3
  • EP4016956B1 patent drawingFigure 4
  • EP4016956B1 patent drawingFigure 5~6

AI summary

A calling method. Said method comprises: acquiring voice data sent by at least three first terminals, the voice data carrying indication information; selecting, according to the acquired indication information of at least three paths of voice data, at least two paths of target voice data from the at least three paths of voice data; sending the at least two paths of target voice data to a second terminal, the second terminal being used to decode the at least two paths of target voice data, mixing the decoded at least two paths of target voice data, and playing back the mixed voice data.