Nonlinear Transform for Parallel Transmission for Image-Sensor-based Visible Light Communication

A nonlinear transform scheme is proposed to convert M discrete samples of parallel transmission (PT) system into the dynamic range between the transmitter and receiver for image-sensor-based (IS-based) visible light communication (VLC). In the PT system reported in [1], although the M discrete samples incorporating N spatial frequencies through M-point inverse discrete Fourier transform (IDFT) are transmitted by M spatially located transmitters. However, the signal of the PT system is, in essential, orthogonal frequency division multiplexing (OFMD) signal, thus suffering from the drawback of the multi-carrier systems, i.e., high peak-to-average ratio (PAPR). With our scheme, the M discrete samples are transformed into the dynamic range between the liquid crystal display (LCD) transmitter and image-sensor (IS) camera receiver. Further, although the M discrete signals are expanded in transforming them, some of the expanded discrete samples are out of the dynamic range. Such out-of-the-range samples are truncated and replaced by the maximum or minimum value, i.e., 255 or 0. Our scheme is carried out for the PT system of the LCD transmitter and a 1000-fps high speed camera receiver which are at the distance of two meters. Experimental results show that our scheme enables the PT system to carry more spatial frequencies with achieving error-free transmission than without the scheme, thus increasing the number of bits per frame.