题 目：Codebook design for Grassmannian sources and applications to Massive MIMO
主讲人：Dr. Maxime Guillaud（欧洲华为研究院）
摘要: I will discuss the problem of quantizing a Grassmannian source with practical implementation constraints. In particular, we focus on high-resolution quantization of Grassmannian variables of arbitrary dimensions, which can lead to extremely large codebooks for which quantization complexity and storage requirements are intractable using conventional approaches. We propose a solution based on a structured codebook, which affords extremely low quantization complexity and zero storage, for a negligible loss of optimality in terms of the rate-distortion trade-off. I will discuss the possible uses of these techniques in Massive MIMO communications.
He is a principal engineer in Huawei Technologies' Mathematical and Algorithmic Sciences Laboratory in Paris, France, where I head the Signal and Information Processing team. I previously worked at Vienna University of Technology, FTW Telecommunications Research Center Vienna, EURECOM (PhD in 2005), and Bell Labs. His research interest is in physical layer design for wireless (in particular cellular massive MIMO) communication networks, including channel reciprocity modeling and calibration, efficient channel state acquisition, representation and feedback, coordinated beamforming in multi-user wireless systems.