独立分量分析法在高速通信系统抗干扰中的研究

发布时间：2018-06-25 09:29

本文选题：独立分量分析 + OFDMA　；参考：《杭州电子科技大学》2017年硕士论文

【摘要】：目前广泛应用的OFDMA技术以及作未来5G移动通信系统的多小区多用户大规模MIMO技术,以其高速传输数据的优点成为现代高速通信系统领域的研究热点。对于采用OFDMA技术的系统,相对于传统的调制技术能够极大的提高信息传输速率和频带利用率。但是在复杂的电磁环境中,往往存在各种人为的恶意干扰或者是同频干扰,那么有效的消除干扰对OFDMA系统的影响具有重要的研究意义。对于大规模MIMO技术,在有限的频谱资源下,采用空间复用技术来提高信息传输速率,从而提高系统容量。但是传统的MIMO系统采用导频复用进行信道估计和解码,会引入导频污染,从而使系统性能恶化。因此对于多小区多用户大规模MIMO系统,如何在系统正确解码的同时有效的避免导频污染问题,已经引起了广泛的关注。为了克服上述问题,本文基于独立分量分析法提出了OFDMA以及大规模MIMO等现代高速通信系统的抗干扰研究算法。首先,针对OFDMA系统中普遍存在的干扰使通信系统性能变差的问题,本文提出了基于独立分量分析算法的OFDMA抗干扰研究算法。该算法在单通道下将接收信号建模成基于子载波矩阵盲源分离的模型,将单音干扰建模成OFDMA信号的虚拟用户,利用独立分量分析法完成子载波矩阵解调,并对单音干扰和OFDMA用户信息进行分离,从而完成抗干扰。所提出的盲源分离抗干扰算法在分离过程中已完成了用户信息的解调,这种盲解调过程并不需要去除循环前缀(Cyclic Prefix,CP)部分。理论分析和仿真结果表明基于ICA的OFDMA系统盲解码抗干扰算法相比于传统的去除CP的FFT解码方法,具有更高的信道增益因而抗干扰性能更好。其次,针对传统的MIMO系统中采用导频复用进行信道估计和解码,而引入导频污染的问题,本文提出了基于独立分量分析算法的多小区多用户大规模天线MIMO抗导频污染研究算法。该算法首先采用了盲源分离的思想将期望信号和干扰信号分离,然后充分利用信道能量的区分度,从所有小区用户中识别出期望小区用户,然后采用少量的参考序列从期望小区的所有用户中识别出某一个期望用户。该算法不需要发送导频进行信道估计,因而不存在导频污染,相较于传统的基于导频进行信道估计与解码的抗干扰算法,具有较好的性能。理论分析和仿真结果表明了多小区多用户大规模MIMO系统下本文提出的基于独立分量分析的抗导频污染的盲解码算法性能明显优于在信道未知的条件下的基于导频辅助的ZF、MMSE的信道估计与解码算法以及基于SVD的信道估计与解码算法具有更好的性能。
[Abstract]:The OFDMA technology and the multi-cell multi-user large-scale MIMO technology which is widely used in the future 5G mobile communication system has become the research hotspot in the field of modern high-speed communication system because of its advantages of high-speed data transmission. For the system using OFDMA technology, compared with the traditional modulation technology, the information transmission rate and frequency band efficiency can be greatly improved. However, in the complex electromagnetic environment, there are many kinds of malicious interference or co-frequency interference, so the effective elimination of interference on the OFDMA system has an important research significance. For large-scale MIMO technology, spatial multiplexing technology is used to improve the transmission rate of information under the limited spectrum resources, thus increasing the capacity of the system. However, the traditional MIMO system uses pilot multiplexing to estimate and decode the channel, which will lead to pilot pollution, which makes the system performance deteriorate. Therefore, for multi-cell multi-user large-scale MIMO systems, how to decode the system correctly and effectively avoid pilot pollution has attracted much attention. In order to overcome the above problems, an anti-jamming algorithm for modern high-speed communication systems such as OFDMA and large-scale MIMO is proposed based on independent component analysis (ICA). Firstly, in order to solve the problem that the interference in OFDMA system makes the communication system performance worse, this paper proposes an OFDMA anti-jamming algorithm based on independent component analysis (ICA) algorithm. In this algorithm, the received signal is modeled as a blind source separation model based on the subcarrier matrix in a single channel, and the single tone interference is modeled as a virtual user of the OFDMA signal, and the subcarrier matrix demodulation is completed by independent component analysis (ICA). And the single tone interference and OFDMA user information are separated to achieve anti-interference. The proposed anti-jamming algorithm for blind source separation has completed the demodulation of user information in the process of separation. This blind demodulation process does not need to remove the cyclic prefix (CP) part. Theoretical analysis and simulation results show that the blind decoding anti-jamming algorithm of OFDMA system based on ICA has higher channel gain and better anti-jamming performance than the traditional FFT decoding method without CP. Secondly, aiming at the problem of channel estimation and decoding using pilot multiplexing in traditional MIMO systems, and introducing pilot pollution, this paper proposes a multi-cell multi-user MIMO anti-pilot pollution research algorithm based on independent component analysis (ICA) algorithm. The algorithm firstly uses the idea of blind source separation to separate the desired signal from the interference signal, and then makes full use of the discriminating degree of channel energy to identify the desired cell user from all the cell users. Then a small number of reference sequences are used to identify a desired user from all users of the desired cell. This algorithm does not need to transmit pilot for channel estimation, so there is no pilot pollution. Compared with the traditional anti-jamming algorithm based on pilot estimation and decoding, this algorithm has better performance. Theoretical analysis and simulation results show that the proposed blind decoding algorithm based on independent component analysis (ICA) is superior to that based on pilot when the channel is unknown. The auxiliary channel estimation and decoding algorithm of ZFN MMSE and the channel estimation and decoding algorithm based on SVD have better performance.
【学位授予单位】：杭州电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.53

【参考文献】