Radio Wave Anomaly Detection Using Multi-Timescale Features

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing radio wave anomaly detection methods fail to accurately detect anomalies with periods exceeding the predetermined time point, as they rely solely on amplitude features extracted at fixed intervals, missing anomalies that may occur outside these intervals.

Innovation Solution

A radio wave anomaly detection system that extracts both short-term and long-term features from received data, using a first feature every first predetermined time period and a second feature every second predetermined time period longer than the first, employing machine learning to identify anomalies using these features.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If amplitude features are extracted every predetermined time point, then the detection process is simple and fast, but radio interference with periods exceeding the predetermined time point cannot be detected

Engineering Contradiction:

Improveanomaly detection accuracyVSAvoidfeature extraction complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The feature extraction process is segmented into multiple types: amplitude features extracted at predetermined time points, and spectral features extracted at different time points. This segmentation allows the system to detect anomalies at multiple time scales without requiring a single complex extraction mechanism, thereby improving detection reliability while managing complexity through modular processing

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system transitions from single-dimension time-based feature extraction to multi-dimensional feature extraction by incorporating both time-domain amplitude features and frequency-domain spectral features. This dimensional expansion enables detection of anomalies with various periods by analyzing signals from multiple feature spaces simultaneously

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Reliability

If only amplitude features are used for anomaly detection, then the detection method is simple, but anomalies with features other than reception level cannot be detected

Engineering Contradiction:

Improveanomaly detection accuracyVSAvoidradio wave feature information

Core Design Contradiction:

ReliabilityVSLoss of information

Solution Approach 1:

The system merges multiple feature types including amplitude features, spectral features, and their combinations. By combining these different feature representations of the same radio wave data, the system preserves comprehensive information about the signal characteristics while enabling detection of diverse anomaly types that would be invisible to single-feature approaches

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The feature extraction system is designed to be universal by extracting multiple types of features from the same received signal data. This multi-functional approach allows the system to detect various anomaly types (interference, failures, distortions) using a unified framework that processes the input data through different feature extraction pathways simultaneously

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Reliability

If a single threshold (spectrum mask) is used for anomaly detection, then the detection process is straightforward, but cases where reception level is low or other features cause anomaly are not detected

Engineering Contradiction:

Improvedetection coverageVSAvoiddetection system complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The detection system is segmented into multiple detection pathways: one using amplitude features with traditional thresholding, another using spectral features with spectral masks, and combined approaches. This segmentation enables comprehensive anomaly detection across different signal characteristics while maintaining relatively simple individual detection modules that can be independently configured and managed

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Spectral features serve as intermediaries that bridge the gap between simple amplitude-based detection and complex multi-parameter analysis. By introducing spectral features as an intermediate layer, the system enhances detection capability for subtle anomalies without requiring direct complex analysis of all raw signal parameters, thus improving detection coverage while controlling overall system complexity

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS12375952B2Radio wave anomaly detection system, radio wave anomaly detection method, and non-transitory computer readable medium storing radio wave anomaly detection program

Publication Date: 2025.07.29 NEC CORP

AI summary

A radio wave anomaly detection system outputs a detection result about a radio wave anomaly that is included in received data and has been detected using a first feature that is a feature extracted from the received data every first predetermined time period, and a second feature that is a feature extracted from the received data every second predetermined time period longer than the first predetermined time period.