Abstract:
In recent years, there has been a growing interest in understanding how rain affects frequency bands in 4G/5G networks, especially in outdoor industrial environments. Mos...Show MoreMetadata
Abstract:
In recent years, there has been a growing interest in understanding how rain affects frequency bands in 4G/5G networks, especially in outdoor industrial environments. Most studies focus on the use of frequencies above 10 GHz, while few address this issue using frequencies typical of 4G/5G systems. This study investigates the effects of rain on band variation in 4G/5G mobile radio networks. In particular, by hypothesizing a value of eNBID, the number of times a band change occurs with and without rain was calculated. From these data, an increment factor (\Delta F), normalized over time, was calculated to define the occurrence of band changes with and without rain. Two CNN network trainings were conducted with two different datasets, the first characterized by the absence of band variation information and the second characterized by the presence of such information. In the first case, an accuracy of about 75% was obtained, and in the second case, the obtained result was around 71%. These findings suggest that the addition of band variation information seems to be not relevant in terms of network performance. Finally, a CTGAN network was introduced for the generation of synthetic data for smaller classes, which resulted in an increase in performance of about 10%, reaching an average accuracy of 85%
Published in: 2024 IEEE 8th Forum on Research and Technologies for Society and Industry Innovation (RTSI)
Date of Conference: 18-20 September 2024
Date Added to IEEE Xplore: 26 November 2024
ISBN Information: