Abstract: ‍ ‍As remote sensing technology develops， the arc synthetic aperture radar （ARCSAR） system has become an important tool for deformation monitoring and emergency rescue applications for landslides. This system acquires high-precision images of the surrounding scene by rotating the radar with a lever arm， which can achieve up to 360° of rotation. Deformation information can be obtained via repeat-pass interferometry. At present， the majority of ARCSAR systems are designed to operate at Ku band or even lower frequency bands， which often results in larger radar system hardware. Additionally， to achieve higher azimuth resolution at such bands， a longer lever arm is needed. This increases the complexity and cost of the system， making it less flexible and portable. Such a system is typically applied to long-distance scenarios， such as mining and emergency rescue， and it is not suitable for close-range scenarios， such as urban infrastructure （e.g.， tunnels）. Millimeter-wave radar has the advantages of wide beams， a small size， light weight， and low power consumption， and has been effectively utilized in many domains， such as vehicle assistance systems， intelligent driving， and traffic monitoring. Combining millimeter-wave radar with ARCSAR may achieve the advantages of a lightweight and high-resolution system， making it more suitable for urban application scenarios. However， currently no such practical ARCSAR systems exist. This paper describes the research and design of a millimeter-wave ARCSAR system. The system uses a millimeter-wave radar module as the radar unit. A lever arm and turntable unit is used to achieve the scanning function. The turntable part is equipped with a high-precision angle encoder for equal angular interval sampling， which reduces the difficulty of image processing. Finally， imaging is realized based on the range Doppler （RD） algorithm. This paper first presents the point target simulation results of the RD algorithm. Moreover， corner reflector indoor experiments and outdoor scene experiments were conducted. In the indoor corner reflector experiment， a comparison was made between the imaging results with and without the use of the angle encoder. The analysis indicated that the azimuth peak sidelobe ratio （PSLR） of the corner reflector was significantly higher with the encoder than that without it. The experimental results showed that the proposed system obtained well-focused images and has great potential for the monitoring of urban infrastructure.