Radar Target Set Generation Using Neural Network Encoding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current radar systems face challenges in accurately generating target sets with precise coordinates in millimeter-wave frequency regimes, particularly in distinguishing targets from noise and ambiguities, which affects detection and tracking performance.

Innovation Solution

A method and radar system utilizing a deep neural network with convolutional encoders and fully-connected layers to process range-Doppler images, combined with a distance-based loss function for training, to generate accurate target sets with associated coordinates, and employing non-uniform discrete Fourier transform coefficients for feature extraction and noise rejection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional radar signal processing methods are used to generate target sets, then the system complexity remains low, but the measurement precision of target coordinates deteriorates due to noise and ambiguities

Engineering Contradiction:
Improvetarget coordinate estimation accuracyVSAvoidneural network processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces traditional mechanical signal processing methods with a neural network-based deep learning system. The neural network learns complex patterns in radar data to accurately estimate target coordinates, substituting conventional algorithms with an intelligent system that can handle noise and ambiguities more effectively.

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

Solution Approach 2:

The patent transforms radar data through multiple parameter transformations including range-Doppler mapping, angle estimation, and coordinate system conversions. The neural network processes these transformed parameters to extract accurate target positions, utilizing parameter changes to enhance measurement precision.

Inventive Principle:
Principle #35Parameter changes

2Reliability

If multiple antennas are used to implement directional beams, then the detection capability improves, but the device complexity increases due to phased array techniques

Engineering Contradiction:
Improvetarget detection capabilityVSAvoidphased array system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The neural network is designed to perform multiple functions including target detection, coordinate estimation, and noise filtering within a single unified system. This multi-functional approach maintains improved detection capability while avoiding the complexity of separate processing systems for each function.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Measurement precision

If machine learning techniques are applied to resolve angular ambiguities, then the measurement precision improves, but the loss of time increases due to data processing requirements

Engineering Contradiction:
Improveangular position estimation accuracyVSAvoidprocessing time for target set generation
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The neural network is pre-trained on extensive radar data to learn patterns of angular ambiguities and their resolutions. This preliminary training allows the system to quickly resolve ambiguities during actual operation without requiring extensive real-time computation, thereby reducing processing time while maintaining high measurement precision.

Inventive Principle:
Principle #10Preliminary action

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 enhances target detection and tracking by improving the accuracy of target coordinate estimation, reducing noise interference, and increasing the robustness of the radar system in identifying multiple targets within a wide field of view.

Implementation Method 1

the distance between the radar and a target is determined by transmitting a frequency modulated signal, receiving a reflection of the frequency modulated signal (also referred to as the echo), and determining a distance based on a time delay and/or frequency difference between the transmission and reception of the frequency modulated signal

Methodology Applied
Scientific EffectRadar: Radar

Implementation Method 2

receiving a reflection of the frequency modulated signal (also referred to as the echo)

Methodology Applied
Scientific EffectEcho: Echo

Implementation Method 3

mixing a replica of the plurality of transmitted radar signals with the plurality of received reflected radar signals to generate an intermediate frequency signal

Methodology Applied
Scientific EffectMixing: Heterodyne

Data Source

PatentEP3992661B1Radar-based target set generation
Publication Date: 2024.10.23 INFINEON TECHNOLOGIES AG
  • EP3992661B1 patent drawingFigure 1
  • EP3992661B1 patent drawingFigure 2~4
  • EP3992661B1 patent drawingFigure 5A

AI summary

In an embodiment, a method for generating a target set using a radar includes: generating, using the radar, a plurality of radar images; receiving the plurality of radar images with a convolutional encoder; and generating the target set using a plurality of fully-connected layers based on an output of the convolutional encoder, where each target of the target set has associated first and second coordinates.