Machine Learning Phase Ambiguity Limitation Wireless Systems

Solution Overview

Problem

Next-generation wireless communication networks face complexity in modeling, analysis, and management due to challenging requirements such as higher data rates, lower latency, and energy efficiency, particularly with the trend of densification, where traditional methods are inadequate, and phase ambiguity in wireless systems hinders efficient machine learning processing.

Innovation Solution

Implementing a machine learning system that limits phase ambiguity to reduce the number of necessary teaching steps and improve processing efficiency, allowing for better control and optimization of wireless communication systems by mapping inputs and outputs to specific phase formats, thereby enhancing the quality of solutions and reducing processing efforts.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional machine learning processing is used without phase ambiguity limitation, then the system can handle all possible phase variations, but the number of teaching steps increases and processing efficiency decreases

Engineering Contradiction:

Improveprocessing efficiencyVSAvoidnumber of teaching steps

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The patent extracts and eliminates the phase ambiguity component from the input signal processing. By removing the ambiguous phase information that does not contribute to the classification task, the machine learning system processes only the essential features, thereby reducing the number of teaching steps required and improving processing efficiency without sacrificing classification accuracy.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies parameter transformation by changing the phase representation of the input signal. Specifically, it transforms the phase-ambiguous complex signal into a phase-invariant representation (e.g., using magnitude only or phase differences), which reduces the dimensionality of the input space and accelerates the machine learning training process while maintaining the discriminative information needed for accurate classification.

Inventive Principle:

Principle #35Parameter changes

2Adaptability or versatility

If phase ambiguity is not limited, then all phase variations are processed, but the complexity of the machine learning system increases

Engineering Contradiction:

Improvehandling of phase variationsVSAvoidmachine learning system complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent extracts and removes the phase ambiguity component from the input signal processing. By removing the ambiguous phase information that does not contribute to the classification task, the machine learning system processes only the essential features, thereby reducing the number of teaching steps required and improving processing efficiency without sacrificing classification accuracy.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies parameter transformation by changing the phase representation of the input signal. Specifically, it transforms the phase-ambiguous complex signal into a phase-invariant representation (e.g., using magnitude only or phase differences), which reduces the dimensionality of the input space and accelerates the machine learning training process while maintaining the discriminative information needed for accurate classification.

Inventive Principle:

Principle #35Parameter changes

3Measurement precision

If full phase space is processed in machine learning, then comprehensive analysis is achieved, but processing speed decreases

Engineering Contradiction:

Improveanalysis comprehensivenessVSAvoidprocessing speed

Core Design Contradiction:

Measurement precisionVSSpeed

Solution Approach 1:

The patent extracts and removes the phase ambiguity component from the input signal processing. By removing the ambiguous phase information that does not contribute to the classification task, the machine learning system processes only the essential features, thereby reducing the number of teaching steps required and improving processing efficiency without sacrificing classification accuracy.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies parameter transformation by changing the phase representation of the input signal. Specifically, it transforms the phase-ambiguous complex signal into a phase-invariant representation (e.g., using magnitude only or phase differences), which reduces the dimensionality of the input space and accelerates the machine learning training process while maintaining the discriminative information needed for accurate classification.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20240380455A1Machine learning for wireless communication

Publication Date: 2024.11.14 TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

AI summary

There is disclosed a machine learning system. The machine learning system is configured to provide an output based on an input, the input representing a status of a wireless communication system comprising a plurality of radio nodes, the output representing an action for the wireless communication system, the machine learning system being configured for a phase ambiguity limitation regarding the output. The disclosure also pertains to related devices and methods, for example radio nodes anda wireless communication system.