The Iterative Sequential Polynomial Fitting Algorithm for a Mobile 4CASK-OCC System

We propose and experimentally validate the adaptive detection method for a 4CASK optical camera communication (OCC) system that supports user mobility. In mobile scenarios, lateral movement shifts the brightest received position in the image, while vertical movement causes the change in received light intensity. These effects lead to grayscale variations for the same symbol across different frames, making empirical or trained thresholds ineffective. To address this, we propose an iterative sequential polynomial fitting (ISPF) method for detecting multi-level symbols in a 4-compound amplitude shift keying (4CASK) OCC system. By iteratively fitting each symbol level, ISPF adapts to the inherent nonlinearity of LED modulation and derives time-varying thresholds per frame, ensuring a fixed transmission rate under varying distances and movement speeds. Additionally, we design an outlier point identification algorithm to filter erroneous symbols caused by the illumination unevenness and LED instability at the edges of the fitted curve. Experiments were conducted on a 4CASK-OCC system with movement speeds of the camera as 40, 60, and 80 cm/s, covering vertical distances from 100 cm to 400 cm and lateral distances up to 50 cm. Results show that movement speed has minimal impact on the system performance, and the proposed method achieves 6.8 kbps throughput under low illumination (70 lx). Our approach requires no training and is robust to channel variations.