Modified User Grouping and Hybrid Precoding for Information Decoding and Energy Harvesting in Hardware Impaired Point-To-Point MM-Wave Massive MIMO NOMA

Abstract

The fifth generation (5G) and beyond 5G race covers both spectral-efficient and energy-efficient mechanisms such as simultaneous wireless information decoding and energy harvesting (SWIDEH).Structured as full or sub-connection of hybrid-precoded mm-wave massive multiple-input multiple-outputnon-orthogonal multiple access (NOMA) systems. However, the more practical hardware-impaired (HI)functionality yet to be considered for point-to-point (p-to-p) MIMO is conceived from an existing system.A hardware impairment-based power splitting is proposed for energy harvesting. The energy of the HI signalis harnessed through the joint optimization of power allocation (PA) for information decoding (ID) and thepower splitting (PS) factor for energy harvesting (EH) using zero-forcing (ZF) precoding and minimummean square error (MMSE) downlink (DL) detection. Based on the simulation results, an adaptive clusterhead selection (CHS) criterion is proposed for analog precoding to mitigate intracluster interferences ata high signal-to-noise ratio (SINR). A successive interference cancellation (SIC) is carried out to enabledynamic channel overhead estimation and user signal detection. An investigation is done by simulating thesystem performance at varied numbers of users, the number of radio-frequency (RF) chains, and channelcharacteristics for modelling the user group threshold coupled with channel size reduction via an antennaselection scheme. Observing the execution time and some graphical plots shows that the spectral and energyefficiency can be improved by a random (R) or maximum norm average (MNA) channel vector formationfrom each user’s channel matrix combined with the best adaptive correlation threshold during CHS and aregularized ZF (RZF) precoder.