基于随机几何理论的认知无线Ad Hoc网络容量与时延研究

发布时间：2018-05-05 10:49

本文选题：认知无线Ad + Hoc网络　；参考：《北京邮电大学》2014年博士论文

【摘要】：认知无线电技术是解决频谱短缺与频谱利用率低下这一矛盾的有效方法,它通过允许非授权用户在不影响授权用户工作前提下动态使用频谱来实现频谱的高效利用。认知无线Ad Hoc网络(Cognitive Radio Ad Hoc Networks, CRAN)兼具了认知无线电技术和Ad Hoc网络的特点,不但可动态利用频谱,还具有Ad Hoc网络的自组织、灵活性等特点。对于CRAN性能的分析必须考虑到分布式网络的多跳结构、动态网络拓扑以及随着时间和位置变化的可用频谱信息等特点。合理高效的性能度量对提高网络性能至关重要。本论文基于无线Ad Hoc网络性能度量已有的研究成果,采用随机几何理论对CRAN建模,研究了其容量和时延。具体的研究成果和创新点如下： 1、分析了基于Underlay频谱共享模式的CRAN的容量和时延。当主、次网络分别建模为相互独立的泊松点过程并且次网络采用Underlay的方式与主网络共享频谱时,次网络通过限制其节点密度满足主网络中断约束来实现频谱的动态接入。分析了参考传输距离外与方向角度共同确定的选择区域内最近接收节点策略下的期望前进密度(Expected Density of Progress)。这里期望前进密度定义为网络单位面积内成功完成一跳的节点的个数和有效传输距离的乘积。以期望前进密度为目标,分别讨论了参考传输距离与发送概率的单独优化和同时优化,推导了单独优化时最优的参考传输距离的上界和发送概率的表达式,以及同时优化时两个参数满足的等式。给出了Underlay频谱共享情况下CRAN的局部时延和端到端时延的闭式结果。数值结果显示,在次网络节点密度允许的范围内存在一个最优的参考传输距离,使得期望前进密度最大。 2、分析了基于Overlay频谱共享模式的CRAN容量和时延。当主网络使用频谱的行为建模为连续时间两个状态的马尔可夫过程,次网络建模为泊松点过程并且采用Overlay的方式与主网络共享频谱时,在主网络处于空闲状态时以时隙ALOHA MAC (Medium Access Control)协议接入信道。研究了选择区域内三种接收节点(最远、最近和随机)选择策略下的期望前进密度。数值结果显示,采用最近接收节点选择策略时期望前进密度要优于其他两种策略的期望前进密度。推导了三种接收节点选择策略下的局部时延和端到端时延的解析式。数值和仿真结果说明,当发送概率较小时,采用最近接收节点策略得到的端到端时延是最小的。 3、分析了基于不同业务模型的CRAN的端到端时延。当主网络建模为均匀泊松点过程时,次网络多跳路由模型建模为线型多跳路径,并采用时分复用(TDMA)与ALOHA相结合的MAC协议以Underly方式接入信道。分析了源节点分别采用两种业务到达模型(Backlogged和Geometric)时次网络数据传输进入稳态的条件,稳态时的成功传输概率和主网络中断约束下次网络源节点密度的取值范围。推导了基于成功传输概率的局部时延和端到端时延的解析式。在此基础上对端到端时延进行优化,推导了Backlogged业务模型下最小端到端时延对应的最优的跳数和源节点发送概率满足的表达式,以及Geometric业务模型下最小端到端时延对应的最优的跳数满足的关系式。
[Abstract]:Cognitive radio technology is an effective method to solve the contradiction between spectrum shortage and spectrum utilization . It can realize the efficient utilization of spectrum by allowing unauthorized users to dynamically use spectrum without affecting authorized users . Cognitive wireless Ad Hoc Networks ( CRAN ) has the characteristics of cognitive radio technology and Ad Hoc network .

Reasonable and efficient performance measures are essential to improve the performance of the network . Based on the research results of the performance metrics of wireless Ad Hoc networks , the paper studies the capacity and time delay of CRAN based on the stochastic geometry theory . The research results and innovation points are as follows :

1 . The capacity and time delay of CRAN based on Underlay spectrum sharing mode are analyzed . When the primary and secondary networks are modeled as mutually independent Poisson point processes and the sub - networks share the spectrum with the main network in the manner of Underlay , the secondary network realizes the dynamic access of the frequency spectrum by limiting the node density of the sub - network to satisfy the main network interruption constraint . The expected density of progress in the selection area determined jointly with the direction angle outside the reference transmission distance is analyzed . In this paper , it is expected that the forward density is defined as the product of the number of nodes successfully completing a hop within the network unit area and the effective transmission distance . In order to achieve the desired forward density , the expression of the upper bound and the transmission probability of the optimal reference transmission distance and the closed result of the end - to - end delay are discussed . The numerical results show that there is an optimal reference transmission distance in the range allowed by the density of the sub - network node , so that the maximum forward density is expected .

This paper analyzes the CRAN capacity and time delay based on Overlay spectrum sharing mode . When the main network uses the behavior of frequency spectrum as the Markov process of two states of continuous time , the sub - network is modeled as the Poisson point process and the channel is shared with the time slot ALOHA MAC ( Medium Access Control ) protocol when the main network is in the idle state . The numerical results show that the local time delay and the end - to - end delay of the three receiving nodes in the selection area are analyzed . Numerical and simulation results show that when the transmission probability is small , the end - to - end delay obtained by adopting the most recent receiving node strategy is minimum .

3 . The end - to - end delay of CRAN based on different business models is analyzed . When the main network is modeled as a homogeneous Poisson point process , the sub - network multi - hop routing model is modeled as a linear multi - hop path , and a MAC protocol combining time division multiplexing ( TDMA ) and ALOHA is used to access the channel .

【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】：TN929.5

【共引文献】