A STOCHASTIC APPROACH FOR ANALYSIS OF SHADOWING EFFECTS IN URBAN AREA

5G networks are started to be deployed in urban areas in the dynamic spectrum sharing mode. Therefore, the analysis of Non-Line-of-Sight (NLOS) shadowing effects behind buildings of different heights becomes particularly important. The shadowing of GSM (900 MHz), UMTS (2100 MHz) and LTE (1800 MHz) technologies was investigated in the urban area. The results are examined by using the log-normal path loss model and four different stochastic equations, specifically: normal, log-normal, Rayleigh, and Nakagami. The results showed that the fluctuations of path losses in the urban area correspond to the model of log-normal path losses. However, path losses for individual buildings do not fit this model. Therefore, stochastic methods were used to analyse shadowing effects. A path loss model based on a differential stochastic equation was proposed. This model includes the following values: free space losses, path losses disturbance due to the shadowing effect, standard deviation s, and a freely chosen parameter b. This model allows a very accurate estimate of the path losses of the three technologies. A clear correlation was observed between parameter b and standard deviation sigma.