Investigation on efficient hybrid precoding techniques for mimo system

Abstract

Guglielmo Marconi successfully tested wireless telegraphy in 1897 by transmitting electromagnetic waves over modest distances of 100 meters. All telephone-related communications were conducted via wired networks in their initial state. But there is a growing trend toward using wireless mobile communication instead of the sophisticated landline telephone system. Without the need for physical means like wires or cables, wireless communication is a technique for sending data from one location to another. There have been a lot of wireless systems and approaches that have flourished and gone during the evolution of wireless communications. The best examples of these are television broadcasts and telephone communication. A billion individuals can now be connected via the internet due to the modern digital economy. The finest illustration of this is the ability of mobile phones to be used anywhere in the world. newlineIn wireless communication mm wave massive MIMO is seen as a next-generation solution for emerging communications. Analog and digital precodings are the basic blocks of MIMO system. In current MIMO system hybrid analog and digital precoding blocks are used. This hybrid structure reduces hardware complexity and energy consumption. The system performance namely SNR value, system capacity, number of users and data rate are maximized. However, this hybrid precoding method has many fundamental limitations. This means that it has high computational complexity and does not use this peer-to-peer approach. In this research work deep learning based techniques are proposed to overcome some of the limitations. newlineThe first proposal is a deep learning-based hybrid precoder. This hybrid precoder with use of deep neural network (DNN) improves the spectral efficiency and reduces BER value. The proposed system performance is not good in of multi user interference environment. newline