SDR/model/surveillance_radar.py

import numpy as np
from scipy.signal import find_peaks
# ==================== 侦查雷达类 ====================
class SurveillanceRadar:
    #初始化函数，传入usrp，输入的通道，输出的通道
    def __init__(self, usrp, rx, tx):
        self.usrp = usrp
        self.rx = rx
        self.tx = tx

        # 封装一个发送信号的函数
    def send_signal(self, tx_signal, duration, center_freq, sample_rate, gain):
        # 发送信号
        self.usrp.send_waveform(tx_signal, duration, center_freq, sample_rate, self.tx, gain)
        print('侦查雷达已发送信号')

    # 封装一个接收信号的函数
    def recv_signal(self, num_samples, sample_rate, center_freq):
        rx_signal = self.usrp.recv_num_samps(num_samples, center_freq, sample_rate, self.rx);
        print('侦查雷达已接收信号')
        return rx_signal
    # 分析信号，获取目标距离
    def analyze_signal(self, rx_signal: np.ndarray, sample_rate: float,
                     cfar_threshold: float = 20.0) -> list:
        """
        分析接收信号并返回目标距离列表
        :param rx_signal: 接收信号（复数形式）
        :param sample_rate: 采样率（Hz）
        :param cfar_threshold: CFAR检测阈值（dB）
        :return: 目标距离列表（米）
        """
        # 1. 去斜处理（Dechirping）
        mixed = rx_signal * np.conj(self.tx_signal[:len(rx_signal)])

        # 2. 加窗处理（Hamming窗）
        window = np.hamming(len(mixed))
        windowed = mixed * window

        # 3. 距离FFT
        range_fft = np.fft.fft(windowed)
        spectrum_db = 20 * np.log10(np.abs(range_fft) + 1e-10)  # 转换为dB

        # 4. 计算距离轴
        freq_bins = np.fft.fftfreq(len(spectrum_db), 1/sample_rate)
        ranges = (self.c * self.T * freq_bins) / (2 * self.B)

        # 5. CFAR目标检测（简化版）
        noise_floor = np.median(spectrum_db)
        peaks, _ = find_peaks(spectrum_db, height=noise_floor + cfar_threshold)

        # 6. 提取目标距离（取前向部分）
        valid_peaks = peaks[peaks <= len(ranges)//2]
        print(f"检测到目标距离：{[abs(ranges[p]) for p in valid_peaks]} 米")
        return [abs(ranges[p]) for p in valid_peaks]