A comparative approach of unsupervised machine learning techniques for LTE network parameter clustering

Pasquino, Nicola; Zinno, Stefania; Cotugno, Federica; Petrocelli, Sofia

doi:10.1109/i2mtc43012.2020.9129223

Cited by 5 publications

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

On the Assessment of Exposure From Lte 800-MHZ Downlink Frequency Band Through the Time-Averaged and Integral-Based Measure

Suka

Simić-Pejović

Pejović

2022

Radiation Protection Dosimetry

View full text Add to dashboard Cite

Exposure assessment should consider the stochastic nature of the emitted signals. Therefore, the long-term temporal variability of E-field strength, obtained by band selective measurements for long-term evolution in a real mobile network environment, is considered in this paper. Intensive measurements were carried out 24 h a day for 2 weeks at an indoor location in the urban area. The repeatability of the measurement results was achieved through descriptive statistics of the 24-h instantaneous, time-averaged and integral-based values. More specifically, we addressed the uncertainties on the temporal variability of the instantaneous exposure levels and the effects of duration of averaging and time-integrating on it. To this aim, a huge amount of signal levels (120 960 samples) was collected during the measurement campaign by considering three operators in the 800 MHz downlink band. Results have shown how some factors, such as the time interval in which the measurements are taken, can significantly affect the measurement results and their repeatability. Furthermore, in some cases, such effects become comparable to, or even larger than typical measurement uncertainty. By applying both the 24-h time-averaged and integral-based measure, the variability of daily exposure levels could be reduced to values lower than the measurement uncertainty (i.e. lower than ± 20% of the mean week value obtained with either measure). Based on such results, an indicator depending on the exposure duration should better quantify the far-field exposure in each location.

show abstract