Occlusion Detection In Visible Light Communication: Algorithm Design And System Implementation

To meet the wireless communication requirement of the explosive mobile services,the wireless radio spectrum is nearly occupied and become more and more congested.Visible Light Communication(VLC)is a new technique transferring data through the blinks of light emitting diodes(LEDs)which are used for illumination simultaneously.It is a green technology,secure,radiation-free and with high throughput.VLC is regarded as a promising solution to mitigate wireless spectrum shortage.However,VLC relies heavily on the direct link between the transceivers,which may be blocked frequently in real application.Therefore,to improve the stability of VLC systems,it is critical to detect occlusions.But there is little research on the barrier and stability of the direct link of VLC transmission,and no perfect solution at present.In this paper,we study the occlusion detection in VLC systems through efficient analysis of the physical layer data during the transmission.With the combination of the improved K-means cluster algorithm and the compressive sensing method,we propose an approach to on-line check and sometimes predict the situation of occlusions in VLC.We can store the data to guide the scheduling of the transmitting end for data packages by showing the situation of receiving end's channel.With the same technology,the data flow statistic information can be well conserved to trace the malfunction reasons.Implementation and simulations in real VLC systems demonstrate that our approach can efficiently detect occlusion online.