SOFTWARE DEFINED RADIO ATTENUATION CONTROL IN 5G COMMUNICATION SYSTEM

Abstract

Millimeter wave (mmWave) frequencies are under research for providing high speed data spectrum for upcoming 5th Generation (5G) communicationsystem. The mmWave frequencies suffer with high propagation attenuation due to various channel obstruc tions. There is a demand and need to have an effective method for efficient mmWave propagation. The focus of this thesis is to identify the various channel obstructions along with their effect on mmWave propagation in terms of attenuation and channel capacity to have desired controlled user data rate. In this research, we propose an innovative method for intelligently selecting a grid or group of grids having minimal mmWave propagation attenuation to overcome the effect of obstructions at the transmission end. The proposed 5G ML transmitter system keeps learning mmWave propagation vegetation attenuation values. The ML unit pre dicts the vegetation attenuation values using regression mode with the algorithms like K-Nearest Neighbors, Decision Tree and Random Forest and also predicts Shannon channel capacity (SCC). Further, we present mmWave propagation losses dataset for four Indian major urban cities like Delhi, Mumbai, Kolkata and Chennai locations in the presence of atmospheric impairments. The simulations have also been performed for the mmWave frequencies 28 GHz, 37 GHz and 39 GHz. The thesis research and results can be used for transmission power loss budget analysis, transmission power control, beamforming, SCC analysis and etc. It significantly improves the 5G system performance by saving the transmitter power radiation and provides multigigabit data rates.