DIGITAL SIGNAL PROCESSING SOLUTION FOR BEAMFORMING IN MIMO COMMUNICATION

Abstract

This dissertation report presents methodologies and digital signal processing solutions to optimize the beamshaping and beamformation. Implementation of these methodologies is presented in this report using LTE signal as an input to the system which lays foundation for future technologies and upcoming 5G technology in millimeter wave band. The results provided can act as building blocks for upcoming millimeter wave communication systems. Analysis of beamforming techniques is applied for compensation of noise as well as in- terference using MATLAB 2017a. The report also presents windowing techniques which is implemented for the suppression of side lobes of the beam. Analysis of directivity with respect to number of antenna elements for windowing techniques is also presented. Results of window techniques for large number of antenna elements are used as a motivation for the Genetic Algorithm based side lobe suppression. Analysis of results in connection with suppression of side lobes using Genetic Algorithm is observed for number of metrics using LTE signal. In order to have one optimal solution and higher side lobe suppression than observed in GA and windowing techniques, Convex Optimization approach was implemented, whose results eventually showed that there is higher side lobe suppression and better directivity. It was noted that computational time in convex optimization was dependent on number of antenna elements, hence a new approach and novel technique named Relaxed Methodology using Convex Optimization (RMCO) is proposed, whose results show that the computa- tional time is not directly dependent on the number of antenna elements however there is some minor decrement in SLS performance, thus acts as a trade-o between computational time and performance