FMCW Radar Target Classification with Multifeature AI at Ultra-Short Range
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing FMCW radars face challenges in detecting and distinguishing targets at ultra-short ranges of several centimeters due to limitations in resolution and precision, making it difficult to identify the type and status of targets accurately.
Innovation Solution
A target detection device that extracts range profiles, magnitude variance, phase variance, scatter plot, and spectrogram data from radar signals and inputs them into an AI model to detect target types, enhancing classification performance by combining numerical and image data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If FMCW radar is used to detect targets, then detection capability at medium ranges is improved, but detection precision at ultra-short ranges deteriorates
Solution Approach 1:
The patent segments the radar detection process into multiple stages: signal reception, range profile extraction, feature extraction (magnitude variance, phase variance, scatter plot, spectrogram), and AI-based classification. This segmentation allows specialized processing for ultra-short range detection while maintaining overall detection capability across different ranges.
Solution Approach 2:
The patent transforms the radar signal from one-dimensional time-domain data into multi-dimensional feature space by extracting range profiles, magnitude variance, phase variance, scatter plot data, and spectrogram data. This dimensional transformation enables the AI model to capture complex target characteristics that are not visible in the original signal, thereby improving ultra-short range detection precision.
2Device complexity
If traditional radar signal processing is used, then system complexity is reduced, but target classification accuracy deteriorates
Solution Approach 1:
The patent introduces an AI model as an intermediary between the radar signal processing and target classification. The AI model receives multiple extracted features (range profiles, magnitude variance, phase variance, scatter plot, spectrogram) and performs sophisticated pattern recognition to classify target types and motion states, achieving high classification accuracy without requiring complex hardware modifications.
Solution Approach 2:
The patent changes the parameter representation from raw radar signal parameters to derived feature parameters including magnitude variance, phase variance, scatter plot coordinates, and spectrogram frequency-time characteristics. These transformed parameters provide richer information for classification while being computationally tractable for the AI model to process.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The solution enables accurate classification of targets at short ranges, including differentiation between objects and people, and detailed classification of object types and motion, improving detection capabilities.
Implementation Method 1
A Frequency Modulated Continuous Wave (FMCW) radar is a sensor that detects and tracks targets using millimeter waves
Implementation Method 2
The FMCW radar may transmit electromagnetic waves toward a target and then receive signals reflected by the target
Data Source
AI summary
A detection device extracts a plurality of range profiles from a radar signal detected during a time period from a target; determines obtain magnitude variance data defined with respect to a plurality of magnitude components corresponding to a target index in the plurality of range profiles, phase variance data defined with respect to a plurality of phase components corresponding to the target index, scatter plot data defined with respect to a plurality of In-phase Quadrature (IQ) components corresponding to the target index, and spectrogram data defined with respect to the target index and a plurality of adjacent indices adjacent to the target index; and inputs the magnitude variance data, the phase variance data, the scatter plot data, and the spectrogram data into an Artificial Intelligence (AI) model to detect a type of the target.