大规模MIMO系统中单精度模数转换器的量化门限设置方案

Adaptive Quantization Threshold Setting Method for Massive MIMO Systems with 1-bit Analog-to-Digital Converter

针对配置单精度模数转换器(ADC)的大规模多输入多输出(MIMO)系统中信道估计误差和上行信号误符号率较高的问题,提出一种基于大规模MIMO信道相关性的自适应ADC量化门限设置方案,并据此给出了相应的信道估计及上行多用户检测方案。首先根据3GPP信道模型建模并生成信道数据,对该数据进行线性分类以获得与发射功率及用户数自适应的量化门限及相应的量化值,以此对上行导频及有用信号进行量化,然后设计基于该自适应量化门限的信道估计及信号检测矩阵,最终检测出上行信号。所提自适应门限设置方法将相关性强的天线进行组合,为同一组合内的多个单精度ADC设置不同的量化门限以提升量化精度。仿真结果表明:与传统固定量化门限的方法相比,所提出的自适应门限设置方法可充分利用信道相关特性,当天线的相关性较强(天线间距小于等于0.5λ)的时候,信道估计的最小均方误差降低到原来的50%左右,同时系统的误符号率提升了5dB以上。

A new adaptive threshold setting method using the channel correlation as well as its corresponding channel estimation and multiuser signal detection scheme is proposed to improve the performance of channel estimation and signal detection in massive multi-input multi-output systems with 1-bit analog-to-digital converter（ADC）. Channel data is firstly generated using the 3GPP channel model, and linearly categorized to obtain the quantization thresholds and codebook that are adaptive to the transmitting power and the number of users. Then based on the adaptive thresholds, the channel estimation and signal detection matrices are designed for the final uplink multiuser signal detection using the quantized uplink reference and desired signals. The method combines high-related antennas and jointly sets different thresholds for 1-bit ADCs to improve quantization precision. Simulation results and comparisons with the conventional quantization method with fixed threshold show that the proposed method makes full use of the channelcorrelation, improves the accuracy of channel estimation by about 50%, and reduces the system symbol error rate by more than 5 dB when the antenna spacing is 0. 5λ.