基于球不变过程的信道衰落模型性能分析及误差研究

发布时间：2018-05-19 15:47

本文选题：SIRP + 信道衰落模型　；参考：《北京交通大学》2017年硕士论文

【摘要】：随着无线衰落信道模型的不断提出,如何在统一归类的基础上进行系统化的数学分析成为了难点,球不变随机过程(Spherically Invariant Random Process,SIRP)很好地为该分析提供了理论工具。SIRP是一般化的高斯过程,继承了高斯过程易于处理、线性闭合等诸多优良性质。且已有理论证明,现有的大多数衰落模型均可归类为SIRP。本文主要关注了基于SIRP的无线信道衰落模型,对其参数化特性和系统可靠性进行了讨论,并进一步研究了 SIRP下的系统误差。论文的主要工作包括以下几点:(1)阐述了如何通过梅林变换和H函数对SIRP的参数进行求解,并重点关注小尺度衰落情形,从广义Gamma分布入手,选取了 Nakagami-m和Weibull这两种具有代表性的小尺度衰落信道作为研究对象,分析了在非视距场景下,两种信道中特征变量V的相关特性;(2)依据无线系统性能的统一数学表示,选取误比特率和中断概率作为评价可靠性的主要指标,推导了两者在非视距(Not Line of Sight,NLOS)场景和视距(Line of Sight,LOS)场景中基于球不变衰落的一般公式,分析了 Nakagami-m和Weibull信道下的可靠性。并讨论了一种在SIRP条件的广义矩问题,给出了三种情形下系统性能的边界结果;(3)分析了在球不变衰落信道下,如何根据实测数据在规定误差范围内选择信道模型X这一核心问题。使用Kullback-Leibler(K-L)距离及其近似估计算法,对存在误差的两种小尺度信道模型进行了评价;(4)提出了球不变随机衰落模型下的两个概念:概率密度函数比率η和关于Gmax(d)的相对误差评判指标r。通过仿真,定量地分析了在所要求的精度内由模型选择不准确带来的系统误差的范围;同时比较了不同信噪比下,当K-L距离为不同量级时与理论模型系统性能的差别并分析了原因;(5)在小尺度衰落的基础上,简单给出了大尺度衰落对SIRP信道的影响。提出了大尺度衰落在SIRP信道下的统一表达方式以及几种受到阴影对数正态分布影响的混合衰落模型。
[Abstract]:With the development of wireless fading channel model, it is difficult to make systematic mathematical analysis on the basis of unified classification. Spherically Invariant Random process SIRP) provides a theoretical tool for this analysis. SILP is a generalized Gao Si process, which inherits many excellent properties such as easy processing of Gao Si processes, linear closure and so on. It has been proved that most of the existing fading models can be classified as SIRP. This paper focuses on the wireless channel fading model based on SIRP, discusses its parameterized characteristics and system reliability, and further studies the system error under SIRP. The main work of this paper includes the following points: 1) how to solve the parameters of SIRP by Merlin transform and H function, and focus on the case of small scale fading, starting with the generalized Gamma distribution. Two representative small scale fading channels, Nakagami-m and Weibull, are selected as the research objects. In the non-line-of-sight scenario, the correlation characteristics of the characteristic variable V in the two channels are analyzed. (2) the unified mathematical representation of wireless system performance is presented. The error bit rate (BER) and the outage probability (BER) are selected as the main indexes to evaluate the reliability. The general formulas based on sphere invariant fading in the non-line-of-sight (not Line of SightNLOSs) scene and the line-of-sight line of SightLos (SightLos) scenario are derived, and the reliability in Nakagami-m and Weibull channels is analyzed. In this paper, a generalized moment problem under SIRP condition is discussed, and the boundary result of system performance in three cases is given. How to select the channel model X according to the measured data within the specified error range is a key problem. Using the Kullback-Leibler-K-L) distance and its approximate estimation algorithm, two kinds of small scale channel models with errors are evaluated. Two concepts under the ball-invariant random fading model are proposed: probability density function ratio 畏 and relative error evaluation index r about Gmaxd. Through simulation, the range of system error caused by inaccurate model selection within the required accuracy is quantitatively analyzed, and the different signal-to-noise ratio (SNR) is compared. When the K-L distance is different from the theoretical model, the difference between the performance of the system and the theoretical model is analyzed. Based on the small scale fading, the influence of large-scale fading on SIRP channel is simply given. A unified representation of large-scale fading over SIRP channels and several mixed fading models affected by the shadow logarithmic normal distribution are proposed.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN92

【相似文献】