A Quantized CSI Acquisition Strategy Based on Position Information Feedback for Indoor VLC-Based IoT Systems

Channel State Information (CSI), characterizing real-time optical channel conditions, is critical for system optimization in visible light communication (VLC) systems. In VLC networks, CSI acquisition typically relies on uplink feedback mechanisms that require frequent updates due to rapid channel variations. As access point (AP) density increases, this feedback process incurs substantial overhead that scales with network size, becoming a major bottleneck for system scalability. ‌By decoupling static channel characteristics from dynamic user positions‌, this paper proposes a dual-step strategy decoupling offline channel quantization from online uplink positioning. This enables efficient downlink CSI reconstruction via pre-trained lookup tables, bypassing real-time computations. The feedback overhead scales as 1/M of conventional methods in M-AP systems, with experiments showing less than 3% spectrum efficiency loss for precoding-sensitive applications in high speed mobile scenarios. The framework’s adaptability to varying LED densities further supports practical deployment in large-scale networks.‌ This work resolves the CSI acquisition bottleneck, enabling scalable VLC systems with minimal performance degradation.