Radar Signal Processing for UAV Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Doppler type radar systems face challenges in effectively detecting and classifying small Unmanned Aerial Vehicles (UAVs) due to their low radar cross-section, varied motion, and high maneuverability, which complicates signal processing and differentiation from clutter and noise.
Innovation Solution
A Doppler type radar system with processing circuitry that generates a three-dimensional data array by cropping and combining range-Doppler matrices from multiple beam lines, optimizing the input for neural networks to enhance feature extraction and classification, particularly using convolutional neural networks for improved object classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional Doppler radar signal processing is used, then the system can detect objects with basic range and velocity information, but it cannot effectively distinguish small UAVs from clutter and noise due to their low radar cross-section and varied motion
Solution Approach 1:
The patent segments the radar detection space into multiple beam lines and further divides each beam line's data into range-Doppler matrices. By processing each segment separately and then combining results, the system can focus computational resources on identifying UAV-specific patterns in each localized region, improving classification accuracy while managing the complexity of distinguishing signals from clutter.
Solution Approach 2:
The patent transforms the traditional two-dimensional range-Doppler data into a three-dimensional data array by incorporating beam line dimension. This additional dimension allows the system to analyze spatial distribution patterns across multiple beam lines, providing more features for distinguishing UAVs from clutter and improving measurement precision without overwhelming the processing capability.
2Reliability
If the radar system processes data from all beam lines comprehensively, then complete coverage is achieved, but the processing complexity and computational load increase significantly
Solution Approach 1:
The patent divides the complete radar dataset into multiple independent range-Doppler matrices, one for each beam line. This segmentation allows parallel processing of individual beam line data, reducing the computational complexity of handling all data simultaneously while maintaining complete detection coverage through aggregation of results from all segments.
Solution Approach 2:
The patent processes data from multiple beam lines (excessive action) rather than focusing on a single beam line, but does so in a modular way that manages complexity. By generating separate range-Doppler matrices for each beam line and then combining them, the system achieves comprehensive coverage without the exponential complexity increase that would result from attempting to process all data in a single monolithic operation.
3Loss of information
If the system generates detailed range-Doppler maps for all range and velocity combinations, then complete velocity information is obtained, but the data processing time and computational resources increase
Solution Approach 1:
The patent segments the velocity analysis into separate range-Doppler matrices for different beam lines, allowing parallel computation of velocity information across multiple segments. This maintains complete velocity information coverage while reducing total processing time through parallelization, as each segment can be processed independently and simultaneously.
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 the radar system's capability to distinguish target signals from noise and clutter, increasing the classification range and accuracy for small UAVs, thereby improving detection and tracking performance.
Implementation Method 1
Doppler type radar systems, such as FMCW radar systems, are well-known and wide spread for use in the automotive sector and other industrial applications
Implementation Method 2
the received waveform will build up a delayed replica of the transmitted waveform, with the time delay as a measure of the target range. If the target is moving, the radar system will register a Doppler shift within the received signal
Data Source
AI summary
A Doppler type radar system holds processing circuitry configured to generate a data array based on received radar data to thereby provide an optimized input of radar data for further processing, such as processing by a neural network or convolutional neural network.