FPGA Harmonic Radar for Multi-Target Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Harmonic radar systems face challenges with high linearity requirements in transmitting channels, weak target harmonic signals, and limited target identification capabilities, primarily restricted to binary classification, leading to low accuracy and difficulty in distinguishing various types of targets.
Innovation Solution
A harmonic radar system utilizing an FPGA and deep learning, incorporating a signal processing module, RF transmitter, transceiver antenna, harmonic receivers, and a target identification module, employs band-pass filters, pulse compression, and deep learning algorithms to enhance signal processing and target classification, enabling identification of multiple target types.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional harmonic radar uses simple binary classification, then the system is easy to implement, but the target identification accuracy is low and cannot distinguish detailed target types
Solution Approach 1:
The patent replaces traditional mechanical signal processing methods with FPGA-based digital signal processing and deep learning algorithms. The FPGA implements complex signal processing operations (Fourier transform, autocorrelation, spectral analysis) and the deep learning model (CNN-LSTM hybrid architecture) achieves multi-class target identification with 90% accuracy, resolving the contradiction between implementation simplicity and identification precision.
Solution Approach 2:
The patent changes the classification approach from binary to multi-class by introducing 12 distinct target categories. The deep learning model processes multiple feature parameters (frequency, amplitude, phase, time characteristics) simultaneously, enabling detailed target differentiation while maintaining system feasibility through automated feature extraction and classification.
2Speed
If the harmonic radar uses conventional signal processing, then the processing speed is slow, but the real-time processing capability is insufficient for effective target detection
Solution Approach 1:
The patent substitutes conventional software-based signal processing with FPGA hardware implementation. The FPGA's parallel architecture executes signal processing operations (Fourier transform, autocorrelation, spectral analysis) simultaneously, achieving real-time processing capability and eliminating the time delay inherent in sequential software processing.
Solution Approach 2:
The patent implements periodic signal processing through the LFM (linear frequency modulation) waveform and pulse repetition sequence. The systematic periodic structure enables efficient correlation processing and spectral analysis, significantly accelerating the detection and processing speed while maintaining accuracy.
3Power
If the harmonic radar transmits strong signals to overcome weak target reflections, then the transmission power is high, but the linearity requirement of the transmitting channel becomes extremely high
Solution Approach 1:
The patent extracts and processes only the harmonic components (2nd and 3rd harmonics) generated by nonlinear targets, separating these useful signals from the fundamental transmission signal. By focusing processing on the harmonic components rather than requiring perfect linearity throughout the entire transmission chain, the system achieves effective target detection while relaxing the stringent linearity requirements.
Solution Approach 2:
The patent applies preliminary signal processing steps (bandpass filtering, amplitude modulation, frequency shifting) before the signal reaches the target. These preliminary actions prepare the signal in a way that enhances the nonlinear response from harmonic targets while reducing the impact of channel nonlinearities on the final detection accuracy.
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 system achieves improved target identification accuracy and type differentiation, overcoming limitations of traditional harmonic radars by using FPGA's parallel processing capabilities and deep learning to stabilize and expedite neural network design, achieving 90% identification accuracy for 12 target types.
Implementation Method 1
mixes the analog signal with a local carrier frequency signal to obtain an RF transmitting signal through modulation
Implementation Method 2
transmits the RF transmitting signal by using the transceiver antenna
Implementation Method 3
the nonlinear target also radiates a higher-harmonic component externally
Implementation Method 4
after undergoing low-noise amplification and filtering by the corresponding harmonic receiver
Implementation Method 5
the second-harmonic signal and a corresponding local oscillator signal are demodulated by an I/Q demodulator to obtain an echo baseband signal
Implementation Method 6
the signal processing module performs pulse compression, pulse accumulation, and identification and detection on signals
Data Source
AI summary
The present disclosure relates to radar signal processing, and a harmonic radar based on a field programmable gate array (FPGA) and deep learning. Specifically, a signal processing module generates a baseband linear frequency modulation (LFM) signal, converts the baseband LFM signal into an analog signal, and mixes the analog signal with a local carrier frequency signal to obtain a radio frequency (RF) transmitting signal. A second-harmonic signal/a third-harmonic signal is generated after the RF transmitting signal irradiates a target, and is transmitted to a second-harmonic receiver/a third-harmonic receiver correspondingly; after undergoing low-noise amplification and filtering by the corresponding harmonic receiver, the second-harmonic signal/the third-harmonic signal and a corresponding local oscillator signal are demodulated to obtain an echo baseband signal; and the echo baseband signal is amplified and sent to a signal processing module for processing after being quantified and converted by an analog-to-digital converter (ADC).