Packet Header Multiplexing for AI/ML Data in Mobile Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication systems face inefficiencies in managing and transmitting AI/ML data, leading to increased traffic load and data size due to frequent reporting and redundant data transmission, which can impact latency and accuracy in network management.
Innovation Solution
Implementing a method and apparatus that utilize header configuration information to include network management and user service provision information in packet headers, allowing efficient transmission of AI/ML data through a multi-tier AI/ML model structure, optimizing data processing and reducing redundancy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If AI/ML data is frequently reported and transmitted through existing wireless communication systems, then network management and service provision can be maintained, but traffic load increases and latency increases
Solution Approach 1:
The patent merges AI/ML data with existing packet headers by placing AI/ML data in the header field of packets transmitted between base stations and core network. This integration allows AI/ML data to be transmitted concurrently with regular traffic, eliminating separate reporting mechanisms and reducing latency while maintaining network management accuracy.
Solution Approach 2:
The patent makes packet headers universal by enabling them to carry both traditional network management information and AI/ML data. The header configuration information allows flexible multiplexing of different data types in the same transmission channel, reducing the need for dedicated AI/ML data channels and lowering overall traffic load.
2Loss of information
If AI/ML data is transmitted through existing packet structures, then data can be conveyed, but data size increases due to redundant transmission
Solution Approach 1:
The patent extracts redundant AI/ML data from separate transmission channels and integrates it directly into existing packet headers. By removing the need for dedicated AI/ML data transmission channels and leveraging existing header structures, the system eliminates redundant data transmission while maintaining complete information conveyance.
Solution Approach 2:
The patent changes the structural parameters of packet headers by introducing header configuration information that enables AI/ML data multiplexing. This parameter change allows the same header structure to serve multiple purposes - carrying both traditional network management data and AI/ML data - thereby reducing overall data size without losing information.
3Productivity
If separate AI/ML data transmission channels are used, then data can be transmitted efficiently, but device complexity increases
Solution Approach 1:
The patent makes existing packet headers universal by enabling them to carry both traditional network management information and AI/ML data. The header configuration information allows flexible multiplexing of different data types in the same transmission channel, eliminating the need for separate AI/ML data channels and reducing system complexity.
Solution Approach 2:
The patent merges AI/ML data transmission with existing packet header transmission mechanisms. By combining what were previously separate transmission functions into a unified header-based system, the patent maintains transmission efficiency while significantly reducing device and system complexity.
Data Source
AI summary
A data transmission method and apparatuses in a wireless communication system are provided. More particularly, a method performed by a first node in a wireless communication system is provided. The method includes receiving header configuration information from a second node, if header configuration information indicates to contain first data including one or more of network management information and user service provision information into a header, transmitting a packet including the first data and additional information for parsing the first data to the second node, and receiving a response to the first data from the second node which receives the packet.