Characterization and modeling of MIMO wireless channels based on correlation tensor

This paper presents the characterization and modeling of several correlative channel models, namely, the Kronecker, Weichselberger, and structured models for multi-input multi-output (MIMO) systems. Due to its simplicity, the Kronecker model is by far the most popular MIMO channel model in the literature. It greatly simplifies channel analysis, as it holds that scatterers around the transmitter fade independently of those around the receiver. Despite its popularity, the Kronecker model has been shown to be inaccurate. The Weichselberger model uses the eigenvalue decomposition (EVD) of the channel as its parameters, and in most cases, is more accurate than the Kronecker model. The structured model is an extension of the Weichselberger model to the wideband MIMO channel. It uses tensors to express the wideband MIMO channel in an elegant fashion, often with fewer parameters than the Kronecker model.