An artificial neural network equalizer for constant power 4-PAM in optical camera communications

Abstract

In this paper, for the first time, we propose and validate a nonlinear artificial neural network (ANN) equalizer for constant power pulse amplitude modulation (CP 4-PAM) in an optical camera communication (OCC) system. Using the proposed equalizer, we demonstrate experimental based non-flickering transmission at a data rate of up to 7 kbps using a single light-emitting diode and an image sensor with 30 frames per second rate. The quality of the received signal is introduced based on the eye-diagram opening where we have proved an ability to transmit data with up to 232 bits per frame after the equalizer deployment.