Research on blind equalization algorithm of high-speed visible light communication signal based on machine learning and Marketing

Issue  JCB, Volume 26, Number 1, 2021  Research Article  

Yali Li
Department of Network and Communication Engineering, Shijiazhuang Information Engineering Vocational College, Shijiazhuang, 052160, China
Ning Chen
Department of Network and Communication Engineering, Shijiazhuang Information Engineering Vocational College, Shijiazhuang, 052160, China
Hui Li
Department of Network and Communication Engineering, Shijiazhuang Information Engineering Vocational College, Shijiazhuang, 052160, China

DOI:https://doi.org/10.5912/jcb1050


Abstract:

In order to solve the problem of high communication error rate caused by inter symbol interference in high-speed visible light communication signal blind equalization algorithm, a high-speed visible light communication signal blind equalization algorithm is designed based on machine learning. The visible light communication channel model is constructed, and a fusion length limited coding mode is set to make the signal transmitted by visible light communication more in line with the overall response of the channel. The decision feedback equalizer is selected as the equalizer, and the channel cascade impulse response condition is set to compensate the distorted output of the channel. The blind equalization algorithm is designed to classify the data by unsupervised learning, and the extreme point of the error function is found through iterative calculation to optimize the equalizer. The simulation results show that under the same SNR test conditions, the bit error rate of the proposed algorithm is significantly lower than that of the blind equalization algorithm based on LMS and BP neural network. Under the condition of SNR = 20dB, the bit error rate is significantly reduced to the order of 10-5, with high communication quality. The VLC system uses an LED backlight display with a CMOS camera. In general, this panel is used to display marketing information. By adjusting the brightness of the CMOS camera's rolling shutter effect (RSE), additional data, such as dynamic content, may be wirelessly sent. NR (noise to signal) ratio of VLC can't keep up with the variety of material being shown on the screen. It is possible to significantly enhance the CMOS RSE pattern decomposition by using a grayscale value distribution (GVD) and a machine learning method (MLA).

Download PDF