Classification and Comparison of Massive MIMO Propagation Channel Models
Rui Feng, Cheng‐Xiang Wang, Jie Huang, Xiqi Gao, Sana Salous, Harald Haas
Abstract
Considering great benefits brought by massive multiple-input–multiple-output (MIMO) technologies in the Internet of Things (IoT), it is of vital importance to analyze new massive MIMO channel characteristics and develop corresponding channel models. In the literature, various massive MIMO channel models have been proposed and classified with different but confusing methods, i.e., physical versus analytical method and deterministic versus stochastic method. To have a better understanding and usage of massive MIMO channel models, this work summarizes different classification methods and presents an up-to-date unified classification framework, i.e., artificial intelligence (AI)-based predictive channel models and classical nonpredictive channel models, which further clarify and combine the deterministic versus stochastic and physical versus analytical methods. Furthermore, massive MIMO channel measurement campaigns are reviewed to summarize new massive MIMO channel characteristics. Recent advances in massive MIMO channel modeling are surveyed. In addition, typical nonpredictive massive MIMO channel models are elaborated and compared, i.e., deterministic models and stochastic models, which include the correlation-based stochastic model (CBSM), geometry-based stochastic model (GBSM), and beam-domain channel model (BDCM). Finally, future challenges in massive MIMO channel modeling are given.