该文提出了一种便携式伪随机编码超宽带人体感知雷达的设计方法。人体感知雷达主要包括对运动人体目标进行跟踪以及对静止人体目标的呼吸频率进行提取。为了获得较强的穿透能力与较好的距离分辨率，采用了载波为800 MHz的m序列作为雷达发射信号。为了减小雷达系统尺寸，采用高速DAC与FPGA直接合成m序列调制信号，发射信号的平均功率为5 dBm。接收机具有两个混合采样接收通道，第1通道用于实时获得参考码，第2通道用于接收雷达回波。为了提高雷达系统扫描率，采用FPGA内部的多个DSP内核级联来实现脉冲压缩的并行计算。另外，动目标跟踪算法与生命探测算法在Intel处理器内运行，将探测结果传给头盔上的微型显示器进行显示。最后，通过穿墙实验表明，该雷达能实时跟踪墙后16 m内的动目标，同时能提取墙后14 m内静止人体的呼吸频率。%This paper presents the design of a handheld pseudo random coded Ultra-WideBand (UWB) radar for human sensing. The main tasks of the radar are to track the moving human object and extract the human respiratory frequency. In order to achieve perfect penetrability and good range resolution, m sequence with a carrier of 800 MHz is chosen as the transmitting signal. The modulated m-sequence can be generated directly by the high-speed DAC and FPGA to reduce the size of the radar system, and the mean power of the transmitting signal is 5 dBm. The receiver has two receiving channels based on hybrid sampling, the first receiving channel is to sample the reference signal and the second receiving channel is to obtain the radar echo. The real-time pulse compression is computed in parallel with a group of on-chip DSP48E slices in FPGA to improve the scanning rate of the radar system. Additionally, the algorithms of moving target tracking and life detection are implemented using Intel’s micro-processor, and the detection results are sent to the micro displayer fixed on the helmet. The experimental results show that the moving target located at less than 16 m far away from the wall can be tracked, and the respiratory frequency of the static human at less than 14 m far away from the wall can be extracted.
展开▼
