Design Of Indoor Target Detection System For Millimeter Wave Radar Based On FPGA

With the rapid development of Io T technology,the acquisition of indoor location information is becoming increasingly important.Although global positioning systems can help us obtain accurate outdoor location information,there still remains some issues in indoor location detection.Currently,domestic and foreign scholars have proposed various indoor detection technologies,mainly applying sensors such as WIFI,RFID,Bluetooth,infrared wire,etc.These technologies had demerits such as low detection accuracy,complex engineering,and be prone to interference.In this paper,millimeter wave radar sensor is employed for indoor detection.Millimeter wave radar detection technology has advantages such as high accuracy,low power consumption,anti-interference,and all-weather performance,etc.However,the issue of 400 Mbps high-speed data acquisition and caching in the RF front-end of millimeter wave radar was a challenge in the application and development of millimeter wave radar systems.This thesis proposes the use of FPGA+DDR3 data acquisition systems for acquisition because FPGA acquisition technology has advantages of flexible development,short development cycle,and fast execution speed compared to CPU and DSP data acquisition technologies.This thesis uses TI’s millimeter wave radar as the RF front-end sensor to build an FPGA based indoor target detection system for millimeter wave radar,achieving indoor target detection.The main work of the paper is as follows:1.A millimeter wave radar indoor detection system based on FPGA is built,and the FPGA toplevel system architecture is designed,which mainly includes clock frequency division module,RF front-end configuration module,RF front-end interface module,DDR3 memory module and USB module.In the thesis,signal analysis and timing analysis on each sub module are carried out,and Verilog is used to design and simulate the functionality on the Vivado platform.Finally,Vivado was used for board level verification,ILA module was used to debug internal signals,and FPGA internal resource usage report,timing report and power consumption report were analyzed after board level synthesis,layout and wiring.In the design of the RF front-end configuration module and RF frontend interface module,in order to address the configuration issues of the RF front-end,this thesis adopts the idea of nested dual state machines.The two state machines respectively benchmark the internal registers and data packets of the RF front-end,and carry out state transfer through different RF front-end internal registers calling configuration data packets.The working parameters of the RF front-end millimeter wave radar can be configured through the SPI interface.At the same time,for the high-speed LVDS data of 400 Mbps in the RF front-end,this thesis designs dual asynchronous FIFO for cross-clock domain data acquisition of high and low level effective data respectively.DDR3 memory is employed to achieve high-speed data caching in this thesis,with an I/O port clock rate of800 MHz,a data bit width of 16 bit,and a data transmission rate of 3.125GB/s,which can achieve high-speed data collection and caching in the RF front-end.2.The indoor target detection scene was built in the thesis.When detecting indoor targets in different scenes,the upper computer Matlab receives RF front-end raw data through a USB interface,and parses single frame raw data and 1024 frames raw data to obtain target position information.Finally,the point cloud data of the target was obtained through three models of single detection,double detection and single walking.By comparing the position information of the target point cloud data with the actual position information,the error of the test data can be obtained between-0.1 and0.1,and the test accuracy can reach centimeter level.The feasibility of the millimeter wave radar indoor target detection system based on FPGA was verified.