Machine Learning Passive Intermodulation Cancellation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Passive intermodulation caused by non-linearities in antennas and equipment is difficult to diagnose and costly to resolve, as it requires site visits by skilled technicians and is challenging to differentiate from adjacent channel interference.

Innovation Solution

A machine learning-based system that remotely identifies and diagnoses passive intermodulation sources by analyzing transmission signals, using a cloud-based approach to model and cancel non-linearities, adapting to various interference scenarios and incorporating antenna beamforming techniques.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If site visits by skilled technicians are used to diagnose passive intermodulation, then diagnostic accuracy is improved, but operational cost and time loss increase

Engineering Contradiction:

Improvediagnostic accuracyVSAvoidtime loss

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system enables self-service diagnosis through automated machine learning models that independently identify and diagnose passive intermodulation sources without requiring skilled technicians to physically visit the site. The cloud-based processing automatically analyzes signal data, detects intermodulation products, and locates sources using algorithms that mimic expert diagnostic capabilities.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces the mechanical system of physical site visits with a digital signal processing system. Machine learning models analyze received signal data to identify passive intermodulation sources, substituting the need for technicians to travel to and physically inspect equipment with remote automated analysis of electromagnetic signals.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If site visits by skilled technicians are used to diagnose passive intermodulation, then diagnostic accuracy is improved, but operational cost increases

Engineering Contradiction:

Improvediagnostic accuracyVSAvoidoperational cost

Core Design Contradiction:

Measurement precisionVSEase of manufacture

Solution Approach 1:

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Reliability

If traditional methods are used to detect passive intermodulation, then detection capability is achieved, but difficulty of detecting and measuring increases

Engineering Contradiction:

Improvedetection capabilityVSAvoiddifficulty of detection

Core Design Contradiction:

ReliabilityVSDifficulty of detecting and measuring

Solution Approach 1:

The system employs feedback mechanisms where the machine learning model continuously analyzes received signals, compares detected intermodulation products against expected patterns, and refines its detection algorithms. The cloud-based platform accumulates data from multiple measurements and uses this feedback to improve detection accuracy and reduce false positives over time.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

4Productivity

If cloud-based machine learning is used to cancel passive intermodulation, then productivity is improved, but device complexity increases

Engineering Contradiction:

ImproveproductivityVSAvoiddevice complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent extracts the complex machine learning processing functions from the local equipment and places them in the cloud. The base station or user equipment only needs to capture and transmit signal data, while the heavy computational burden of running machine learning models, training algorithms, and generating cancellation signals is performed remotely in the cloud environment with sufficient computing resources.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The system introduces a cloud-based processing platform as an intermediary between the signal sources and the cancellation execution. This intermediary handles the complex machine learning operations, signal analysis, and model training, while communicating only essential data and control signals to the local equipment, thereby simplifying the local device architecture.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS10601456B2Facilitation of passive intermodulation cancelation via machine learning

Publication Date: 2020.03.24 AT&T INTELLECTUAL PROPERTY I L P

AI summary

A passive intermodulation detection system is provided to remotely identify passive intermodulation at a base station site and diagnose the type of intermodulation and location of the non-linearity that is the source of the passive intermodulation. A passive intermodulation cancelation system can generate an equivalent signal to a received interference signal and use the equivalent signal to generate an error signal. The error signal can then be used to reinforce a learning system and converge on a steady state of the interference signal to cancel other interference signals.