无线信道建模及其在错误估计编码中的应用

发布时间：2018-06-03 00:38

本文选题：无线信道建模 + Trace　；参考：《国防科学技术大学》2014年博士论文

【摘要】：无线信道建模是无线通信领域的基础性技术,它试图以数学模型的方法从纷繁复杂且看似随机出现的无线信道传输错误中,挖掘和认知无线通信过程中错误出现的内在规律,并以模型结构和模型参数的形式定性或者定量地表达这些内在规律,从而为上层协议的设计和性能评估提供依据。本文基于真实无线信道中的测量Trace建立信道模型,并使用信道模型研究目标无线信道中具有指导意义的统计规律,然后将这些规律应用于一种基于编码的报文比特错误率估计方法。本文的主要研究内容包括以下四个方面:(1)研究具有重尾特性的无线信道建模问题无线信道的重尾特性符合人们对于无线信道中传输错误出现规律的客观观察,并且对于上层协议性能具有一定的影响。因此,建立一个能够准确反映无线信道重尾特性的信道模型具有十分重要的理论和实践意义。现有的专门针对重尾信道的模型存在模型结构复杂,运行时成本高、参数化过程复杂等不足。本文从测量Trace的run长度和burst长度互补累积分布函数入手,提出了一种基于独立非同分布的随机过程的重尾信道模型inid。inid模型结构简单灵活、运行时成本低(一个贝努利试验),仅包含两个参数且参数化过程简单。(2)研究通用无线信道建模问题无线信道内在的时变特性以及传输错误出现的随机性和不可预测性,导致了无线信道的测量Trace具有多样化的建模特性。时变的无线信道及其具有不同建模特性的测量Trace加大了无线信道建模的难度。本文从测量Trace的共性入手定义了外部run、内部run、错误过程等概念,并基于错误过程和外部run提出了一种新的Trace表达方法。利用这种新的Trace表达方法,本文提出了一种精确、通用的无线信道模型GEM。GEM模型结构简单、适用性好、参数具有明确的物理意义且建模过程与具体测量Trace的建模特性无关。试验结果表明GEM模型能够精确地刻画具有不同建模特性的Trace。(3)研究模型驱动错误估计编码算法错误估计编码(Error Estimating Coding,EEC)是一种基于编码的报文比特错误率估计方法。它的基本思想是发送端在发送数据中加入少量的冗余信息,接收端基于这些冗余信息估计报文的比特错误率。报文的比特错误率与无线信道的错误特性紧密相关,因此在错误估计编码算法的设计过程中,考虑无线信道的错误特性具有重要的理论和实践意义。本文结合基于模型的信道解码后无线信道错误特性分析,提出了模型驱动错误估计编码算法MEEC。试验结果表明MEEC算法能有效克服信道突发特性和短期记忆特性对于算法估计精度的影响,能以更少的空间开销获得更高的估计精度。(4)研究基于信道模型的Smart错误估计编码现有错误估计编码算法的设计和验证都基于这样一个应用模式:发送端在信道编码前,计算和编码错误估计位;接收端在信道解码后,重新计算错误估计位并估计报文比特错误率。基于这种应用模式的错误估计编码称为信道解码后的错误估计编码。本文首次提出信道解码前的错误估计编码,即发送端在信道编码后、调制前,计算和编码错误估计位;接收端在解调后、信道解码之前,重新计算错误估计位并估计报文比特错误率。相比信道解码后的错误估计编码,信道解码前的错误估计编码的优点是接收端不需要信道解码就可以估计报文比特错误率。因为信道解码比较耗时,所以第二种应用模式的实时性更高。本文使用信道模型建模信道解码前的无线信道并分析目标信道的错误特性,结合基于多进制正交幅度调制(M-QAM)技术的物理层符号错误结构,提出了一种针对信道解码前无线信道的错误估计编码Smart-EEC。试验结果表明基于Smart-EEC的EEC算法和MEEC算法很好地实现了估计精度与空间开销的折中。综上所述,本文主要针对无线信道建模、基于模型的目标信道错误特性认知及其在错误估计编码技术中的应用问题展开深入研究。本文的研究丰富了无线信道建模的理论,推动了信道模型与错误估计编码的融合,具有一定的理论意义和应用价值。
[Abstract]:Wireless channel modeling is a basic technology in the field of wireless communication. It tries to exploit and recognize the inherent laws of errors in the process of wireless communication in the process of complex and seemingly random wireless channel transmission by means of mathematical models, which are expressed qualitatively or quantitatively in the form of model structure and model parameters. The inherent law provides the basis for the design and performance evaluation of the upper layer protocol. This paper builds a channel model based on the measurement Trace in the real wireless channel, and uses the channel model to study the guiding statistical rules in the target wireless channel, and then applies these rules to a coded bit error rate estimation. The main research contents of this paper include the following four aspects: (1) to study the wireless channel modeling with heavy tail characteristics, the heavy tailing characteristics of the wireless channel conform to the objective observation of the law of transmission error in the wireless channel, and have a certain influence on the performance of the upper layer protocol. The channel model reflecting the heavy tailed characteristics of the wireless channel has very important theoretical and practical significance. The existing model for heavy tailed channel is complex, high cost and complex parameterized process. In this paper, a basis for measuring the run length and burst length of Trace to complement the cumulative distribution function is proposed. The heavy tailed channel model inid.inid model in an independent and non identically distributed random process is simple and flexible, and the cost is low (a Bernoulli test). It contains only two parameters and the process of parameterization is simple. (2) the study of the inherent time-varying characteristics of the wireless channel of universal wireless channel modeling and the randomness and inability of the transmission error. Measurability results in a variety of modeling characteristics of the wireless channel measurement Trace. The time-varying wireless channel and the measurement Trace with different modeling characteristics increase the difficulty of modeling the wireless channel. This paper defines the concepts of external run, internal run, error path and so on from the generality of measuring Trace, and based on the error process and external run extraction. A new method of Trace expression is presented. Using this new Trace expression method, this paper proposes an accurate, universal wireless channel model GEM.GEM model with simple structure, good applicability, a clear physical meaning and the modeling process is independent of the modeling of the specific Trace. The experimental results show that the GEM model can be accurate. The Trace. (3) study of model driven error estimation coding algorithm error estimation coding (Error Estimating Coding, EEC) is a coding based error rate estimation method based on encoded bits. Its basic idea is that the transmitter adds a small amount of redundant information to the transmitted data and the receiver is based on these redundant information. The bit error rate of the message is estimated. The bit error rate of the message is closely related to the error characteristics of the wireless channel. Therefore, it is of great theoretical and practical significance to consider the error characteristics of the wireless channel during the design of the error estimation coding algorithm. The MEEC. test results of the model driven error estimation coding algorithm show that the MEEC algorithm can effectively overcome the influence of the burst characteristics and short-term memory characteristics on the estimation accuracy of the algorithm, and can obtain higher estimation accuracy with less space overhead. (4) the existing error estimation coding algorithm for Smart error estimation coding based on channel model is studied. The design and verification are based on such an application pattern: the transmitter calculates and encodes error estimation bits before channel coding; the receiver recalculate error estimates and estimate the error rate of message bits after channel decoding. Error estimation coding based on this application mode is called error estimation coding after channel decoding. This paper is the first time in this paper. The error estimation code before the channel decoding is proposed, that is, before the channel coding, the error estimation bit is calculated and coded before the channel coding, and the receiver is recalculating the error estimation bit and estimating the error rate of the message before the demodulation, the channel decoding, and the error estimation coding before the channel decoding, and the error estimation before the channel decoding is superior. The point is that the receiver can estimate the bit error rate without the need of channel decoding. Because the channel decoding is more time-consuming, the second application modes are more real-time. This paper uses the channel model to model the wireless channel before the channel decoding and analyzes the error characteristics of the target channel, which is based on the multilevel orthogonal amplitude modulation (M-QAM) technology. The error structure of the physical layer symbol is proposed, and a error estimation coding Smart-EEC. test results for the channel decoding before the channel decoding shows that the Smart-EEC based EEC algorithm and the MEEC algorithm achieve the tradeoff between the estimation precision and the space overhead. In summary, this paper mainly deals with the wireless channel modeling and the model based target letter. This research enriches the theory of wireless channel modeling and promotes the fusion of channel model and error estimation coding, which has some theoretical significance and application value.
【学位授予单位】：国防科学技术大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN92

【相似文献】