认知无线网络中高效频谱检测关键技术的研究

发布时间：2018-05-01 20:13

本文选题：认知无线网络 + 频谱检测　；参考：《北京邮电大学》2014年博士论文

【摘要】：随着无线通信业务的飞速增长,无线用户数量急剧上升,异构网络林立,不仅导致了频谱资源日益紧缺,也造成了网络端到端性能的下降。为了提高频谱资源利用效率,提升网络端到端效能,认知无线网络这一致力于对频谱资源进行动态共享、促进异构网络融合的智能网络应运而生。认知技术是认知无线网络实现的关键技术,而其中的频谱检测则是认知技术的核心,也是实现频谱资源高效利用的前提。因此,频谱检测成为了业内研究的热点。尽管学者们针对频谱检测进行了大量的研究,提出了多种频谱检测算法,但是频谱检测仍然面临许多挑战。本文主要针对频谱检测面临的难点展开细致深入的研究。首先在前两章中介绍了认知无线电及认知无线网络的基本原理,并对相关的频谱检测技术、研究现状及频谱检测所面临的难点与挑战进行了分析与讨论。随后,我们提出了一种准确高效的循环平稳特征频谱检测算法,深入的分析了算法在不同衰落信道中的检测性能,并全面分析了基于多天线检测的算法性能,主要取得了以下几个方面的研究成果： 1)准确高效的循环平稳特征检测算法：现有的循环平稳特征检测虽然检测准确性高但存在着计算复杂度大、检测时间长的问题。针对这个问题,本文提出了一种准确高效的循环平稳特征频谱检测算法。通过简化现有循环平稳特征检测的检验统计量,有效的降低了算法的计算复杂度,并根据该检验统计量在不同条件下的概率密度函数,推导了改进算法检测性能的闭合表达式。大量仿真结果验证了改进算法的有效性和准确性：首先,通过对计算复杂度的分析,本文所提出的改进算法与现有算法相比,计算复杂度得到显著降低；其次,检测性能的理论值与仿真值能够很好的吻合,验证了简化后的检验统计量的有效性及检测性能闭合表达式的准确性；虽然,简化检验统计量导致改进算法的检测性能与原有算法相比有0.36dB的性能损失,但却以较小的性能损失换取了计算复杂度的大大降低,并且改进算法仍保持了较高的检测性能,为频谱检测的优化设计提供了基础。 2)改进算法在衰落信道中的检测性能研究：针对主用户与次用户之间的信道为多径衰落和阴影衰落的实际场景,本文对不同的多径衰落信道和复合多径-阴影衰落信道中的改进算法检测性能进行了深入的分析。为了准确量化衰落信道对改进算法检测性能的影响,本文针对Nakagami衰落信道、瑞利衰落信道、莱斯衰落信道和复合瑞利-对数正态阴影衰落信道、复合Nakagami-对数正态阴影衰落信道中的检测性能进行定量分析,并分别利用概率密度函数(PDF, Probability Density Function)方法和矩母函数(MGF, Moment Generation Function)方法分别推导了改进算法在相应衰落信道中检测性能的闭合表达式。通过比较,MGF方法简单易行但有局限性,而PDF方法复杂但更具普适性,虽然这两种检测性能的分析方法各有利弊,但均为分析衰落信道下的检测性能提供了有力的手段,并且这些分析方法不仅适用于分析循环平稳特征检测的性能,同样也适用于分析其他检测方法的性能。 3)基于多天线的改进算法检测性能研究：现有的针对基于多天线的循环平稳特征检测研究较少,且仅有的研究均针对独立多天线,而没有考虑更实际的相关多天线的情况。由于多天线的相关性会削弱多天线的分集增益,造成检测性能的下降。因此,为了对多天线的检测性能进行全面的研究,本文全面分析了基于独立多天线和相关多天线的改进算法的检测性能。对于独立多天线检测,首先考虑到认知无线网络中完整的信道状态信息不易获得的实际情况,本文采用了无需完整信道状态信息的平方律合并方式,随后研究了改进算法在衰落信道中的检测性能；对于相关多天线检测,重点研究了常数相关、指数相关、线性阵列任意相关多天线的检测性能。最后,利用MGF方法,推导了基于平方律合并的独立多天线和相关多天线的检测性能闭合表达式。仿真结果表明,尽管信道衰落与多天线的相关性会削弱检测性能,但是多天线技术依然显著的提高了检测的准确性,该研究有助于量化基于多天线的循环平稳特征检测的可达性能,并为该方法的实际应用起到积极的促进作用。本文最后,我们对所做的工作进行了总结,并就下一步的研究方向及重点进行了展望。
[Abstract]:With the rapid increase of wireless communication service , the number of wireless users has increased sharply , and heterogeneous network forest stands , which not only leads to the increasing frequency spectrum resources , but also reduces the end - to - end performance of the network . In order to improve the utilization efficiency of spectrum resources , improve the end - to - end efficiency of the network , the cognitive wireless network is devoted to dynamic sharing of the spectrum resources . The cognitive technology is the key technology of the cognitive radio network , and the spectrum detection is the premise of the efficient utilization of the spectrum resources . Therefore , the spectrum detection becomes the hot spot in the industry research .

In the first two chapters , we introduce the basic principle of cognitive radio and cognitive wireless network , and analyze and discuss the difficulties and challenges faced by the spectrum detection technology , research situation and spectrum detection .

1 ) An accurate and efficient algorithm for detecting cyclostationary features : The detection accuracy is high , and the detection time is long . In view of this problem , this paper proposes a kind of accurate and efficient algorithm for detecting circular stationary feature spectrum . By simplifying the test statistic of the existing circular stationary feature detection , the closed expression of the improved algorithm detection performance is deduced .
Secondly , the theoretical value and the simulation value of the detection performance can be well matched , and the validity of the simplified test statistic and the accuracy of the closed expression of the detection performance are verified ;
Although the simplified test statistic results in the performance loss of 0.36 dB compared with the original algorithm , the performance loss is reduced in exchange for the lower computational complexity , and the improved algorithm still keeps higher detection performance , which provides the basis for the optimization design of spectrum detection .

In order to accurately quantify the effect of fading channel on detection performance of fading channel , this paper makes a quantitative analysis of the performance of the improved algorithm in the fading channel of multi - path fading channel , Rayleigh fading channel , Rice fading channel and composite Rayleigh - log fading channel .

3 ) Based on multi - antenna improved algorithm detection performance research : The existing research on multi - antenna based loop smooth feature detection is few , and only the research is aimed at independent multi - antenna , and the detection performance of multi - antenna is not taken into account . Therefore , in order to study the detection performance of multi - antenna , the paper comprehensively analyzes the performance of the improved algorithm based on independent multi - antenna and related multi - antenna .
Based on the MGF method , the closed expression of the detection performance of independent multi - antenna and related multi - antenna based on square law is studied . The simulation results show that , although the correlation of channel fading and multi - antenna can weaken the detection performance , the simulation results show that the multi - antenna technology can improve the accuracy of detection , which can help quantify the accessibility of the multi - antenna based circular stationary feature detection and play a positive role in the practical application of the method .

In the end , we summarize the work done , and look forward to the research direction and focus of the next step .

【学位授予单位】：北京邮电大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN92

【参考文献】