Research On Visible Light Communication Handover Technology Based On Distributed Computing And Positoning

With the rapid development of wireless communication technology,people put forward higher requirements for high-speed,low-cost,lowpower and secure communication methods.In this context,visible light communication(VLC)technology has emerged.Visible light communication uses visible light for communication,which has the advantages of high security,no electromagnetic pollution and high reliability,and gradually becomes a research hotspot.However,there are some challenges in visible light communication.Due to the propagation characteristics of light,visible light communication is subject to interference from ambient light and requires a light source facing the communication device for communication.Therefore,it becomes critical to realize stable connection of visible light communication in different environments.Visible light positioning(VLP)is a positioning technology based on visible light signals that enables indoor positioning,navigation and other applications.In visible light communication,visible light positioning technology can be used to obtain the location information of communication devices,so that VLC access point(AP)switching and optimization can be performed to improve communication quality and stability.Therefore,it is important to study the visible light communication switching technology based on visible light positioning.To address the above issues,this thesis conducts a study on visible light communication switching technology based on distributed computing and positioning,and the main work and innovations of this thesis are as follows:(1)Both visible light communication and infrared communication have the property of being immune to electromagnetic interference in conventional wavelengths,and the characteristics of the two types of communication can complement each other.Therefore,this thesis gives a visible light and infrared heterogeneous networking communication architecture that can be used in electromagnetically sensitive situations such as warehouses in the Industrial Internet of Things(IIoT)to effectively guarantee user communication quality.(2)A VLP scheme combining image sensors and artificial neural networks is proposed and implemented,and experimental tests are completed.The results show that the scheme can support the positioning accuracy of walking mobile users in the cm scale,and the positioning errors of 90%of the samples are still within 1.49 cm,1.61 cm and 1.92 when the positioning target moves at 1m/s,2m/s and 3m/s,respectively cm or less.(3)A VLP and reinforcement learning(RL)phase assisted indoor visible communication AP point switching scheme in IIoT automatic storage is proposed and implemented,and its performance is optimized by using multiple intelligences combined with distributed computing.In this thesis,we simulate and analyze the performance of the scheme with the assistance of three reinforcement learning algorithms,Q-learning,Sarsa and Deep Q Network(DQN),and the results show that the DQN-assisted switching scheme has the best convergence performance and the lowest convergence value under the same user terminal mobility conditions,and the downlink rate can reach 93.14 Mbps at a single AP bandwidth of 100 MHz The downlink rate can reach 93.14Mbps at a single AP bandwidth of 100MHz,which is 11.3%and 11.4%higher than that of Q-learning and Sarsa,respectively.