Research On Adaptive Location-based CSI Feedback Compression

Recent years,the rapid development of mobile internet has brought great opportunities and challenges to network engineers and researchers.High-speed,low-cost and stable internet access are urgent needed by network equipment.MU-MIMO proposed in 802.11 ac was applied in WLAN,WiMAX and LTE cellular networks to solve these problems.MU-MIMO enables concurrent transmissions,has potential to influence deployment of network and increase efficiency of spectrum,enhances the network throughput and capacity directlyHowever,transmission supported by beamforming in MU-MIMO is highly depends on channel state information feedback from receivers.Experiment evaluation shows that overhead on CSI feedback increase with the number of users,sometimes even overwhelm the real transmission.Hence,efficient CSI feedback scheme is the key point of MU-MIMO network.Exist schemes solve this problem to some degree,however,leaving some loopholes of scalability and fairness,unreasonable use of the characteristics cause the ineffective of the MUMIMO network.Hence,an Adaptive Location-based CSI Feedback Compression Scheme in MU-MIMO Networks is proposed in this paper.LFC effectively use the location information of receivers,and enables receivers have similar SINR in a nearby area could share a same CSI matrix,LFC feedback in terms of group instead of per-packet scheme.LFC provides a matric to balance the tradeoff between overhead of CSI feedback and system throughput.Simulation and experiment are designed to evaluate LFC,result shows that LFC could achieve at most 67.9% and average 30% higher throughput than traditional per-packet scheme.LFC is a low-complexity,high robustness and scalable scheme with consideration of fairness and energy consumption.LFC could also applied in wider range and higher dimension with the help of multiple APs and hardware support.