Efficient feedback scheme based on compressed sensing in MIMO wireless networks

In multiple-input–multiple-output (MIMO) broadcast channel the throughput can be enhanced by channel state information (CSI) feedback, but it is resources and feedback expensive. We propose a compressed sensing (CS) feedback scheme for zero-forcing beamforming (ZFBF) in MIMO broadcast channel, which can reduce the feedback load and resource consumption. The feedback channels are shared and opportunistically accessed by users who are self pre-selected based on their channel vectors’ norm and correlation. Multiple measurement vectors (MMV) CS is used for the CSI feedback. Orthogonal matching pursuit and reduced MMV and boost algorithms are applied for CS recovery to get the CSI. Semi-orthogonal user selection and ZFBF are implemented at the base station to achieve spatial multiplexing gain. Both the analog and digital CS feedback schemes are analyzed. Simulations show that the proposed CS feedback has good performances compared with traditional feedback schemes in points of feedback load and throughput due to user self pre-selection algorithm and CS feedback.

Figure optionsDownload as PowerPoint slideHighlights
► Compressed sensing is applied into MIMO broadcast communication.
► Multiple measurement vectors CS is used for CSI recovery.
► Self pre-selection scheme based on channel vector norm and correlation is proposed.
► CS feedback scheme is feedback efficient compared with traditional feedback.