Switches and Inverters Based Energy-Efficient Dynamic Hybrid Precoding Design for Millimeter-Wave Massive MIMO Systems

碩士 === 國立中正大學 === 通訊工程研究所 === 107 === Consider a hybrid precoding design for a typical millimeter wave multi-user massive MIMO downlink communication system. A large number of high-resolution phase shifters (PSs) are replaced by a small number of energy-efficient switches and inverters in the analog...

Full description

Bibliographic Details
Main Authors: YANG, SHIH-HSIEN, 楊士賢
Other Authors: HU, CHIA-CHANG
Format: Others
Language:zh-TW
Published: 2019
Online Access:http://ndltd.ncl.edu.tw/handle/9a677t
Description
Summary:碩士 === 國立中正大學 === 通訊工程研究所 === 107 === Consider a hybrid precoding design for a typical millimeter wave multi-user massive MIMO downlink communication system. A large number of high-resolution phase shifters (PSs) are replaced by a small number of energy-efficient switches and inverters in the analog precoder part. This SI-based hybrid precoding architecture significantly reduces the hardware complexity and energy consumption of the system, and thus the transmission energy efficiency is increased. On the basis of the achievable sum-rate (ASR) maximization, the SI-based overlapping subarray (OSA) hybrid precoding design is proposed. The existing coordinate update algorithm (CUA) with slightly modified is utilized to find the best precoder. Compared with the current SI-based partially connected architecture, the more the number of overlapping antenna arrays, the better the transmission performance, and the closer to the sum-rate of the PS-based precoding designs. Although the overlapping sub-array design reduces the transmission energy efficiency, it is still better than the PS-based precoding designs. On the other hand, a Gram–Schmidt (GS) orthogonalization based dynamic antenna selection (AS) algorithm for the SI-based architecture is proposed. By eliminating part of the transmission antenna corresponding to the less uncorrelation channel and maintaining the channel transmission capacity, the hardware cost is reduced and the transmission energy efficiency is improved.