3-Dimensional Manifold and Machine Learning Based Localization Algorithm for Wireless Sensor Networks

For upcoming futuristic communication systems, the Optical Wireless Communication (OWC) with its inherent advantages is becoming popular among service providers. The research in the past has primarily focused on Quality of service (QoS) aspect for OWC in the presence of atmospheric turbulence however, to assess the real time outcome of a service, the evaluation of both QoS and Quality of Experience (QoE) deliver a holistic approach. A much Less effort in the existing literature has been paid to this. Authors in this work attempt to determine the QoE for image transmission over a turbulent OWC link while considering the Structure Similarity Index (SSIM) as a visual performance indicator under varying turbulence strengths (regimes). A functional model to forecast the performance of SSIM practicable for all the regimes is proposed. The most suitable model is bimodal Gaussian mixture model which aptly describes the system performance. To improve the performance, spatial domain filters such as Median and Wiener filters have been employed. An increase of 125 m in the propagation distance and 5.88 dB in received SNR can be achieved while maintaining the SSIM at a 90% for median filter restoration in moderate turbulence regime for simulated values while predicted values suggest an increase of 115 m and 5.18 dB at same level. The results show that the proposed model is in good agreement with simulated values and median filter in moderate turbulence performs best out of all the situations.