Search

In this paper, we propose and experimentally demonstrate HyperSight, which is a precise image decoding algorithm for the mobile-phone camera based visible light communication (VLC) system. The conventional algorithms have to select an appropriate column matrix, and then decode the image according to the fitting curve. The selection of the column matrix and the fitting of the function are not only time-consuming but also error-prone, leading to low decoding efficiency and high bit error rate (BER). While HyperSight calculates characteristic grayscale matrix to avoid column matrix selection and uses critical grayscale to replace the function fitting to decode the image. As expected, our algorithm not only reduces the BER, but also improves the processing performance. The experimental results show that, the BER can be reduced by two orders of magnitude on average compared with the conventional scheme. For the $4.8\,$ kbps data rate, our algorithm can achieve the transmission distance of $50\,$ cm under the $7\%$ forward error correction (FEC) threshold with the low illuminance of $702\,$ lux.

In this article, a novel method for constructing a pre-equalizer by optimized filter synthesis with vector fitting is proposed to broaden the bandwidth of a visible light communication (VLC) system. By using the proposed method, an active-passive hybrid equalization circuit is constructed, and it is demonstrated that the 3-dB bandwidth of the VLC system with commercially available phosphorescent white LED is extended from 30 MHz to 600 MHz. A 1.35 Gb/s NRZ-OOK VLC transmission over 12 cm free space transmission distance is experimentally realized, achieving a bit error rate (BER) of $\text{8.8} \times \text{10}^{-\text{5}}$ . To the best of the authors knowledge, this is the largest reported 3-dB bandwidth and real-time data rate achieved in the VLC system with a single commercially phosphorescent white LED.