Theoretical Analysis On MIMO Capacity Under Correlated Fading Channel

Multiple input multiple output(MIMO) system is widely used in wireless communication for its outstanding advantages, and as an extension of the traditional MIMO, Massive MIMO has become a hot spot of research. It is difficult to satisfy the condition of independent fading in the real environment, so it is necessary to study the capacity of MIMO system under the correlated fading channel. The main research content of this paper is centered around it.First of all, two kinds of channel models are introduced. The “One-ring” model is applicable to traditional MIMO system, this paper summarizes its principle and research status. The “Multi-ring” model is applicable to Massive MIMO system, we focus on the spherical wave assumption and non-stationary characteristics of this model.Then, the channel capacity of MIMO system under correlated fading is studied in this paper. An improved scheme of "One-ring" model is proposed, and a new upper bound of channel capacity is proposed. Based on the "One-ring" model, the influence of fading correlation on the channel capacity of the traditional MIMO is studied, and the relationship between the different physical quantities of the system and the channel capacity is shown in detail. Classic “One-ring” model ignores the multiple reflections of waves between the scatters. Combined with the knowledge of graph theory, this paper puts forward the improvement scheme. Based on this improvement scheme, we re-study the influence of fading correlation on the channel capacity, and draw the conclusion. In addition, based on the "One-ring" model, a new upper bound of ergodic capacity, namely Geometry-Based Stochastic Models Double-sided Correlated(GBSMs DSC) upper bound, is proposed to analyze the effect of fading correlation on the MIMO capacity.Finally, for the Massive MIMO system, the effect of fading correlation on the performance of system is studied. The new upper bound of capacity is extended to Massive MIMO, and a simplified upper bound is proposed. The performance of the systems that take different precoding schemes is compared. When extending GBSMs DSC upper bound to single-user Massive MIMO system, we take some simplified measures, and get a simplified upper bound. Compared with the new upper bound, this simplified one reduces the computational complexity but also the performance. Through the new upper bound, the relationship between the channel capacity and different physical quantities is studied, including antenna spacing, tilt angles of Base Station’s antenna and Mobile Station’s antenna, and angular spread. For multi-user Massive MIMO system, the performance of the systems that take the Zero Forcing(ZF) precoding scheme and the Regularized Channel Inversion(RCI) precoding scheme is analyzed and compared.