An Empirical Study of Deep Learning Models for LED Signal Demodulation in Optical Camera Communication

Optical camera communication is an emerging technology that enables communication using light beams, where information is modulated through optical transmissions from light-emitting diodes (LEDs). This work conducts empirical studies to identify the feasibility and effectiveness of using deep learning models to improve signal reception in camera communication. The key contributions of this work include the investigation of transfer learning and customization of existing models to demodulate the signals transmitted using a single LED by applying the classification models on the camera frames at the receiver. In addition to investigating deep learning methods for demodulating a single VLC transmission, this work evaluates two real-world use-cases for the integration of deep learning in visual multiple-input multiple-output (MIMO), where transmissions from a LED array are decoded on a camera receiver. This paper presents the empirical evaluation of state-of-the-art deep neural network (DNN) architectures that are traditionally used for computer vision applications for camera communication.

BER performance study for optical OFDM of optical camera communication

In this article, different forms of optical orthogonal frequency division multiplexing (OFDM) were observed which were suitable for optical camera communication (OCC) systems. This research aims to establish the bit error rate (BER) versus signal-to-noise ratio (SNR) of the OCC system. This research will focus on OCC systems and the design that produces the noise of the clipping but will gain SNR as a whole if an optimum clipping factor is chosen. The BER versus SNR analysis was investigated for the different clipping factors 0.7, 1.4, and 2.6. The BER performance of the asymmetrically clipped optical OFDM (ACO-OFDM) was also compared with the direct current optical OFDM (DCO-OFDM) to show the suitable effectiveness of the proposed approach. ACO-OFDM was considered to be better due to lower bit loading, but DCO-OFDM was efficient for higher SNR values. This was because the DC bias used was inefficient in terms of optical capacity, while ACO-OFDM used only half of the subcarriers to transmit the information. Moreover, ACO-OFDM two-dimensional half-subcarriers of mapping rule would introduce the clipping noise to its unused 2D subcarriers, although further data can be provided by the 2D DCO-OFDM mapping rule.