基于Wishart随机矩阵的发射天线数目估计算法研究
发布时间：2018-10-19 17:45
【摘要】：在非合作MIMO通信场景中,通信信号参数盲识别作为信号检测和信号空时解码解调的中间步骤,具有着十分关键的作用,已经在频谱管理与监测、软件无线电和军事无线通信对抗等民用和军事领域得到了极为广泛的应用。而准确的MIMO发射天线数目估计则是MIMO通信信号参数盲识别的关键问题之一,能够为MIMO信道盲估计、MIMO空时码盲识别和MIMO信号调制方式盲识别等其他关键问题的高效解决提供先验知识,具有着十分重要的研究意义和研究价值,已经开始得到无线通信领域中众多学者的广泛关注和研究。本文在前人工作的基础上对MIMO发射天线数目估计问题进行了探索和研究,具体主要工作内容包括:(1)简略介绍了MIMO系统以及随机矩阵理论等基础理论知识;详细阐述了AIC、MDL、PET以及IWME这四种经典MIMO发射天线数目估计算法的数学理论基础和推导过程,同时,根据上述四种算法的计算机仿真结果,分析了算法正确估计概率与信噪比、采样数据长度以及接收天线数目等仿真参数的关系。(2)提出了一种基于随机矩阵理论的预测特征值上限算法(RPET),与基于经典多元统计理论的PET算法相比,RPET算法能够明显提升低信噪比和采样数据长度较小条件下的正确估计概率,获得略优于IWME算法的综合估计性能。(3)由于RPET算法利用相邻随机变量置信区间的比值预测噪声特征值上限,导致该上限值偏高,算法在低信噪比条件下容易产生欠估。针对这一问题,提出了一种基于Wishart随机矩阵特征值平方均值分布特性的假设检验算法(WSE)。WSE算法利用Wishart随机矩阵特征值平方均值的分布函数和精确的噪声功率估计值求解检测统计量的判决门限。与RPET算法相比,WSE算法能够将低信噪比和采样数据长度较小条件下的正确估计概率再次明显提高,同时具有更加优异的综合估计性能。
[Abstract]:In the non-cooperative MIMO communication scene, blind identification of communication signal parameters, as an intermediate step of signal detection and signal space-time decoding and demodulation, plays a very important role in spectrum management and monitoring. Software radio and military wireless communication countermeasures and other civil and military fields have been widely used. Accurate estimation of the number of MIMO transmit antennas is one of the key problems in blind identification of MIMO communication signal parameters. It can provide a priori knowledge for MIMO channel blind estimation, MIMO space-time code blind identification and MIMO modulation blind identification and so on. It has very important research significance and research value. It has been widely concerned and studied by many scholars in the field of wireless communication. In this paper, based on the previous work, the problem of MIMO antenna number estimation is explored and studied. The main contents are as follows: (1) the basic theoretical knowledge of MIMO system and stochastic matrix theory is briefly introduced; The mathematical theory foundation and derivation process of four classical MIMO antenna number estimation algorithms, AIC,MDL,PET and IWME, are described in detail. At the same time, according to the computer simulation results of the above four algorithms, the correct estimation probability and signal-to-noise ratio are analyzed. (2) A prediction eigenvalue upper bound algorithm based on stochastic matrix theory (RPET),) is proposed. Compared with PET algorithm based on classical multivariate statistical theory, RPET algorithm is able to show that Improve the probability of correct estimation under the condition of low signal-to-noise ratio (SNR) and small sampling data length, The synthetic estimation performance is slightly better than that of the IWME algorithm. (3) because the RPET algorithm uses the ratio of the confidence interval of adjacent random variables to predict the upper limit of the noise eigenvalue, the upper bound value is too high, and the algorithm is prone to underestimate under the condition of low signal-to-noise ratio (SNR). In response to this problem, A hypothesis checking algorithm based on the distribution of eigenvalue square mean of Wishart random matrix is proposed. (WSE). WSE algorithm uses the distribution function of square mean of eigenvalue of Wishart random matrix and accurate noise power estimate to solve the decision threshold of detection statistics. Compared with the RPET algorithm, the WSE algorithm can significantly improve the probability of correct estimation under the condition of low SNR and smaller sampling data length, and has better comprehensive estimation performance.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN827.1
本文编号：2281877
[Abstract]:In the non-cooperative MIMO communication scene, blind identification of communication signal parameters, as an intermediate step of signal detection and signal space-time decoding and demodulation, plays a very important role in spectrum management and monitoring. Software radio and military wireless communication countermeasures and other civil and military fields have been widely used. Accurate estimation of the number of MIMO transmit antennas is one of the key problems in blind identification of MIMO communication signal parameters. It can provide a priori knowledge for MIMO channel blind estimation, MIMO space-time code blind identification and MIMO modulation blind identification and so on. It has very important research significance and research value. It has been widely concerned and studied by many scholars in the field of wireless communication. In this paper, based on the previous work, the problem of MIMO antenna number estimation is explored and studied. The main contents are as follows: (1) the basic theoretical knowledge of MIMO system and stochastic matrix theory is briefly introduced; The mathematical theory foundation and derivation process of four classical MIMO antenna number estimation algorithms, AIC,MDL,PET and IWME, are described in detail. At the same time, according to the computer simulation results of the above four algorithms, the correct estimation probability and signal-to-noise ratio are analyzed. (2) A prediction eigenvalue upper bound algorithm based on stochastic matrix theory (RPET),) is proposed. Compared with PET algorithm based on classical multivariate statistical theory, RPET algorithm is able to show that Improve the probability of correct estimation under the condition of low signal-to-noise ratio (SNR) and small sampling data length, The synthetic estimation performance is slightly better than that of the IWME algorithm. (3) because the RPET algorithm uses the ratio of the confidence interval of adjacent random variables to predict the upper limit of the noise eigenvalue, the upper bound value is too high, and the algorithm is prone to underestimate under the condition of low signal-to-noise ratio (SNR). In response to this problem, A hypothesis checking algorithm based on the distribution of eigenvalue square mean of Wishart random matrix is proposed. (WSE). WSE algorithm uses the distribution function of square mean of eigenvalue of Wishart random matrix and accurate noise power estimate to solve the decision threshold of detection statistics. Compared with the RPET algorithm, the WSE algorithm can significantly improve the probability of correct estimation under the condition of low SNR and smaller sampling data length, and has better comprehensive estimation performance.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN827.1
【相似文献】
相关期刊论文 前1条
1 顾菁华;梁广;龚文斌;余金培;;基于FPGA的DBF发射天线数字波束形成网络的设计与实现[J];实验技术与管理;2009年01期
相关硕士学位论文 前1条
1 朱德明;基于Wishart随机矩阵的发射天线数目估计算法研究[D];西安电子科技大学;2014年
，本文编号：2281877
本文链接：https://www.wllwen.com/kejilunwen/wltx/2281877.html
