利用数据和导频的多小区Massive MIMO系统信道估计

发布时间：2019-02-27 10:23

【摘要】：本文提出一种通过同时利用上行数据和导频进行信道估计来对抗时分双工大规模多输入多输出(Massive multi-input multi-output,Massive MIMO)系统中导频污染的方法。考虑到Massive MIMO系统下空间相关信道在角度域具有近似稀疏性,期望用户信道和干扰信道的多径分量因而可分辨。同时上行数据子帧中的数据符号数远大于导频序列的长度。基于以上两点观察,首先利用上行导频对信道进行粗估计,然后将解码得到的上行数据符号视为导频序列,用于得到角度域信道估计中能量较大分量所处的位置。通过抽取信道粗估计中相应位置上的信道系数,可以得到信道的精估计。仿真表明,无论相邻小区采用同一组正交导频序列、还是不同组但具有相关性的导频序列,所提出的信道估计方法都能有效地消除导频污染。
[Abstract]:This paper presents a method of channel estimation using uplink data and pilot to combat pilot pollution in time-division duplex large-scale multi-input multiple-output (Massive multi-input multi-output,Massive MIMO) systems. Considering that the spatial correlation channel in the Massive MIMO system has approximate sparsity in the angle domain, the multipath components of the user channel and interference channel are expected to be discernible. At the same time, the number of data symbols in the uplink data subframe is much longer than that of the pilot sequence. Based on the above two observations, the uplink pilot is first used to rough estimate the channel, and then the decoded uplink data symbol is regarded as the pilot sequence, which is used to obtain the position of the larger energy component in the angle domain channel estimation. The precise channel estimation can be obtained by decimating the channel coefficients at the corresponding position in the rough estimation of the channel. Simulation results show that the proposed channel estimation method can effectively eliminate pilot pollution regardless of whether the adjacent cells use the same set of orthogonal pilot sequences or different groups of pilot sequences with correlation.
【作者单位】：北京航空航天大学;华为北京研究所;
【基金】：国家自然科学基金海外及港澳学者合作研究基金(61429101) 国家863计划(2014AA01A703)
【分类号】：TN919.3

【相似文献】