Vision-Based Autonomous MM-Wave Reflector Using ArUco-Driven Angle-of-Arrival Estimation

Reliable millimeter-wave (mmWave) communication in non-line-of-sight (NLoS) environments remains a major challenge for next-generation wireless networks, particularly in indoor and infrastructure-constrained scenarios. This paper presents a proof-of-concept vision-guided passive reflector system that enhances mmWave connectivity by autonomously steering signal reflections using a motorized metallic plate. A lightweight monocular camera detects fiducial ArUco markers on transmitter and receiver nodes, estimates their relative orientation, and aligns the reflector in real time, without relying on RF feedback, external infrastructure, or Reconfigurable Intelligent Surfaces (RIS). Beyond dynamic beam alignment, the use of uniquely identifiable markers enables selective targeting of authorized users, addressing the user identification problem largely overlooked in prior vision-aided RIS research. A low-cost prototype, implemented on a Raspberry Pi 5, demonstrates operation in dynamic 60 GHz NLoS indoor environments, achieving received power gains of up to 17 dB (average ≈10 dB) compared to the no-reflector baseline. The results establish vision-guided reflectors as a practical, low-power alternative to active relays or RIS panels, with strong potential for short-range Industry 4.0/5.0 applications such as IoT networks, AR/VR systems, autonomous robotics, and adaptive factory communications.