Key Technology Research For Massive MIMO Systems Based On Symbol-Level Precoding

As a key support in Beyond Fifth Generation(B5G),Massive Multiple Input Multiple Output(MIMO)systems provide higher communication rates,larger capacity,and better coverage,where precoding is an indispensable physical layer signal processing technology to achieve these metrics.The precoding technique suppresses interference in advance and achieves directional transmission of beam-space by acquiring Channel State Information(CSI).Symbol-Level Precoding(SLP)provides excellent performance gain and energy efficiency compared to traditional precoding techniques,and is in line with the development trend of green communication.Focused on the Interference Exploitation Precoding(IEP)of SLP,the thesis proposes various enhanced schemes for improving performance or reducing complexity,then simulates and verifies these schemes on 5G channels.Firstly,the thesis conducts an extensive study of Massive MIMO systems,analyzing the system characteristics and associated indicators.Several linear precoding methods are also investigated.Additionally,as the starting point to study SLP,Vector Perturbation(VP)precoding is researched,highlighting the significance of reducing transmission power to enhance the performance throughout the whole thesis.Subsequently,the thesis investigates the IEP of SLP by in-depth analysis of principles and analogy to VP,and achieves the algorithm design and implementation of IEP in different modulations by additive Constructive Interference(CI),where the simulation and evaluation of IEP under different interference alignment strategies are given.Then,based on the block fading scenario,the thesis investigates SLP schemes of multiple-symbol joint optimization,including Gyre Precoding(GP)and the IEP with joint angle rotation,namely Rotated IEP.The thesis studies the design and implementation of these algorithms and assesses their performance and complexity in accordance with the system models and guiding principles of these schemes.Essentially,both schemes need extra feed-forward overhead for performance gain.Finally,based on the aforementioned schemes,the thesis enhances the IEP scheme with decoupling phase rotation,termed DPR-IEP,and the codebook-assisted version of it,termed CB-DPR-IEP.With the simulation of proposed schemes under different conditions,the thesis completes the performance comparisons and complexity analysis.In addition,the thesis also evaluates the performance of proposed schemes for different symbol block lengths.The simulation results show that DPR-IEP can significantly reduce complexity while maintaining performance gain,and CB-DPR-IEP achieves a trade-off and balance between the complexity,the performance,and the overhead.