Low-Cost Low-Rank Channel Estimation for TDD/FDD massive MIMO Systems--A Unified Approach via Array Signal Processing
Feifei Gao received the Ph.D. degree from National University of Singapore in 2007. He was a Research Fellow with the Institute for Infocomm Research (I2R), A*STAR, in 2008 and an Assistant Professor with the School of Engineering and Science, Jacobs University, Germany from 2009 to 2010. In 2011, he joined the Department of Automation, Tsinghua University, Beijing, China, where he is currently an Associate Professor. Prof. Gao's research areas include communication theory, signal processing for communications, array signal processing, and convex optimizations. He has authored/ coauthored more than 70 refereed IEEE journal papers, with 3500 plus citations from Google Scholar. Prof. Gao has served as an Editor of IEEE Trans. on Wireless Commun., IEEE Wireless Commun. Lett., International Journal on Antennas and Propagations, and China Commun. He has also served as the symposium co-chair for 2015 IEEE ICC, 2014 IEEE GLOBECOM, 2014 IEEE VTC.
We present a new transmission strategy for the multiuser massive MIMO systems, including uplink/downlink channel estimation and user scheduling for data transmission. This scheme exploits the array structure as well as the spatial information of the incoming signal at BS, and generates an alternative low rank signaling model, named as spatial basis expansion model (SBEM). The basis vectors of SBEM are formulated from discrete Fourier Transform (DFT) vectors and can be efficiently deployed by the fast Fourier transform (FFT). With SBEM, both uplink and downlink channel estimation of multiusers can be carried out with very few amount of training resources, which significantly reduces the training overhead and the feedback cost for the downlink training. Moreover, the channel estimation of different users could be spatially separated from orthogonal beams, which immediately relieve the pilot contamination problem. Compared to the existing low rank modeling, the newly proposed SBEM does not require knowledge of massive channel statistics and does not suffer from high complex eigenvalue decomposition of the massive covariance matrices. Another key benefit of SBEM is that due to reciprocity of electromagnetic waves, the beam directions are also reciprocal for uplink and downlink transmission. Hence, SBEM is not only applicable for TDD system but may also present a low complexity solution for the FDD systems. Lastly, some efficient user scheduling algorithms for data transmissions are designed based on different optimization criterions.