Neural Network Configuration Messaging for Adaptive Wireless Recovery

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Evolving wireless communication systems face challenges in processing higher frequency ranges due to increased signal distortions and complex hardware requirements, leading to inefficiencies and higher costs.

Innovation Solution

Implementing dynamic neural network configurations using a neural network table to quickly reconfigure deep neural networks (DNNs) based on changing operating conditions, reducing data transmissions and improving information recovery.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If higher frequency ranges are used to increase data capacity, then data throughput is improved, but signal quality deteriorates due to increased susceptibility to multipath fading, scattering, atmospheric absorption, and interference

Engineering Contradiction:
Improvedata throughputVSAvoidsignal quality
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent implements dynamic neural network configuration where the DNN architecture and parameters are adjusted in real-time based on changing operating conditions such as signal quality metrics. This allows the system to adapt to varying channel conditions at higher frequencies, maintaining reliable information recovery while utilizing high-throughput frequency bands

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes neural network parameters (architecture, layers, filters) based on operating conditions to optimize performance. By dynamically modifying these parameters, the system can compensate for signal quality deterioration at higher frequencies while maintaining high data throughput capability

Inventive Principle:
Principle #35Parameter changes

2Productivity

If complex hardware capable of transmitting and receiving higher frequencies is deployed, then data capacity increases, but processing costs and device complexity increase

Engineering Contradiction:
Improvedata capacityVSAvoidhardware complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent replaces complex hardware-based signal processing with software-based deep neural network processing. This substitution allows the system to handle higher frequency signals and compensate for their vulnerabilities through intelligent algorithms rather than requiring increasingly complex hardware architectures

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system uses neural network models that can be configured and reconfigured through software to handle different frequency bands and signal conditions, avoiding the need for dedicated hardware for each frequency range and reducing overall hardware complexity

Inventive Principle:
Principle #26Copying

3Ease of operation

If traditional configuration methods are used for DNN, then configuration is simple, but responsiveness to changing operating conditions is slow

Engineering Contradiction:
Improveconfiguration simplicityVSAvoidresponsiveness
Core Design Contradiction:
Ease of operationVSSpeed

Solution Approach 1:

The system implements dynamic configuration where the DNN is reconfigured based on real-time operating conditions. The network entity monitors conditions and sends reconfiguration signals to update the DNN parameters, enabling rapid adaptation while maintaining operational simplicity through automated processes

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system pre-configures multiple DNN configurations that can be quickly switched between based on operating conditions. This preliminary preparation of multiple configurations allows rapid response to changing conditions without requiring complex real-time optimization calculations

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3997619B1Communicating a neural network formation configuration
Publication Date: 2025.12.10 GOOGLE LLC
  • EP3997619B1 patent drawingFigure 1
  • EP3997619B1 patent drawingFigure 2
  • EP3997619B1 patent drawingFigure 3

AI summary

Techniques and apparatuses are described for enabling base station-user equipment messaging regarding deep neural networks. A network entity (base station 121, core network server 320) determines a neural network formation configuration (architecture and/or parameter configurations 1208) for a deep neural network (deep neural network(s) 604, 608, 612, 616) for processing communications transmitted over the wireless communication system. The network entity (base station 121, core network server 302) communicates the neural network formation configuration to a user equipment (UE 110). The user equipment (UE 110) configures a first neural network (deep neural network(s) 608, 612) based on the neural network formation configuration. In implementations, the user equipment (UE 110) recovers information communicated over the wireless network using the first neural network (deep neural network(s) 608, 612). This allows the wireless communication system to adapt to changing operating conditions and improve information recovery.