| With the development of intelligent vehicles technology,intelligent sensors such as millimeter-wave radar are widely used in advanced driver assistance systems such as automotive adaptive cruise control(ACC)and automatic emergency braking(AEB),which provides real-time,high accuracy target perception capability and ensures the safety of autonomous vehicles.Meanwhile,due to the increase of electromagnetic equipment in millimeter-wave frequency band and the widespread of automotive millimeter-wave radars,jamming is divided into human deliberate jamming and mutual jamming between multiple vehicle radars.Abovementioned electromagnetic interference will inevitably lead to misidentification or loss of detection targets,which causes serious safety hazards.Accordingly,the research on the interference mechanism and related protection system of automotive millimeter-wave radar is of great significance.A millimeter-wave RF front-end receiving system based on the alldielectric metasurface is proposed to solve the human deliberate jamming in radar system.Compared with the traditional metal structure,the alldielectric metasurface can make the millimeter wave component immune to high-power millimeter wave electromagnetic pulse(HPMMW)damage more effectively.Plus,the metasurface structure provides a new possibility for the miniaturization of the receiving system.The proposed system consists of laser,all-dielectric metasurface antenna,electro-optic resonator,photodetector,and electronic circuitry.We exploit a 3×3 cells array metasurface antenna to capture and deliver the 80 GHz millimeter waves signal to an electro-optic field sensor,then the photodetector carries out the conversion from optical signal to the demodulated radio wave signal which can be further processed by the latter electronic circuitry.The proposed system has a compact size of 7.7×7.7 mm2 and high receiver sensitivity of-52 dBm,which makes it available to be integrated with other electronic circuitries and realize miniaturization design of telecommunication devices.Additionally,due to the complex relationship between metasurface performance and parameters in the synthesis of all-dielectric metasurface antennas,the geometric structure parameters are difficult to be decided through formulas.It requires a large number of electromagnetic simulation iterations,which consumes a lot of time.Accordingly,a metasurface structure parameter prediction method based on stacking ensemble learning model is proposed in this paper.The recommended geometric parameters can be provided according to the user’s performance parameter requirements.The ensemble learning model is built by combining five primary learners to give full play to the advantages of each type of learners.The system shows a great ability of parameter prediction with a mean absolute percentage error(MAPE)of 3.73%.In addition,this paper compares the proposed model with the traditional data-driven model,which fully verifies the performance of the model.For the mutual j amming between multiple vehicle radars problem,this paper analyzes the radar interference mechanism in detail and establishes a probability calculation model.This model can be used to analyze the influence of radar setting parameters on the interference in actual traffic conditions and provide model guidance for the interference protection of automotive millimeter-wave radars.The actual traffic flow of Beijing is taken as an example to verify the results of the model.The model can provide an assessment of the security of millimeter-wave radars and provide guidance for the security configuration of millimeter-wave radars. |
PDF Full Text Request