Nonorthogonal Time-Frequency Training-Sequence-Based CSI Acquisition for MIMO Systems

作者：Ding, Wenbo; Yang, Fang^*; Liu, Sicong; Wang, Xianbin; Song, Jian

来源：IEEE Transactions on Vehicular Technology, 2016, 65(7): 5714-+.

DOI：10.1109/TVT.2015.2463716

• Channel state information (CSI) acquisition
• multiple-input multiple-output (MIMO)
• nonorthogonal time-frequency training sequence (TS)
• spatial-temporal correlations
• structured compressed sensing (SCS)

In this paper, a spectrally efficient nonorthogonal time-frequency training sequence (TS)-based channel state information (CSI) acquisition approach is proposed for multiple-input-multiple-output orthogonal frequency-division multiplexing (MIMO-OFDM) systems under the framework of structured compressed sensing (SCS). The scheme first relies on a time-domain TS that is identical for all transmit antennas to acquire the partial channel common support by utilizing the spatial correlation property of the MIMO channels and then uses the frequency-domain TS for accurate CSI recovery based on the proposed adaptive spatially-temporally joint simultaneous orthogonal matching pursuit algorithm. Here, the obtained partial channel common support can be utilized to reduce the complexity of the classical SCS algorithm and improve the signal recovery probability. Simulation results show that the proposed scheme could significantly reduce the TS overhead and demonstrate better performance than the existing MIMO-OFDM systems, which might be suitable for massive MIMO systems in future fifth-generation communications.

• 出版日期2016-7
• 单位清华大学; 深圳清华大学研究院