中文摘要：

在多小区场景下,若要实现网络辅助干扰抵消（NAIC）,需要首先获取邻小区信息（例如邻小区使用的调制方式等）。由于通过信令传输的方式会导致较大的系统开销,因此邻区信息的盲识别成为了第三代合作伙伴计划（3GPP）工作组及学术界研究的方向之一。目前调制方式的识别方法大多以高阶累积量作为特征统计量来进行识别。这些方法由于复杂度高、效率不佳且工程上较难实现,因此很少应用于实际系统。考虑到长期演进系统采用了3种正交振幅调制方式（QAM）进行调制,分别是正交相移键控（QPSK）、16QAM和64QAM,提出了一种低复杂度的基于星座图功率阚值的MQAM调制方式盲识别算法。所提算法具有较低的复杂度和较高的识别率。

英文摘要：

Under multi-cell scenario, the target cell should obtain the information of the neighbor cells including the modulation classification before the process of network-assisted interference cancellation. Because achieving the neighbor cells＇ information through signaling methods would result in a larger system overhead, the blind identification has become one of the most con- cerning problem to the third generation partnership project （3GPP） and the communication companies. Most of the research in modulation classification based on higher order cumulant as a statistic feature to do the identifying job, while this method is rarely applied to practical system because of its high complexity, inefficiency and difficulty in engineering. Considering the LTE system adopts QPSK, 16QAM and 64QAM as the modulation methods, this paper proposes an MQAM classification algorithm based on the power of the data points in constellation graphic, which is effective even in low SNR scenario through the simula- tion result.