基于认知的移动Ad Hoc网络重构推理决策技术

发布时间：2018-09-11 06:15

【摘要】：认知移动自组织网络是由多个地位相等的移动节点所构成的认知网络,是移动自组织网络技术与认知网络技术的结合。认知网络的智能性主要体现在其重构能力上,网络重构技术使网络能够根据内外部环境条件变化导致的网络状态的变化,动态优化配置网络结构和网络协议,提高网络通信性能。随着无线通信网络的不断发展,用户对认知网络的要求越来越高,作为认知网络的关键技术之一,网络重构技术以其重要性和不可或缺性逐渐得到了大家的重视,已成为认知网络的一大研究重点。目前的网络重构技术研究并没有形成体系,还没有完全覆盖到重构实现的方方面面,基本上分布在几个细节的研究点上面。认知移动自组织网络的重构技术的研究尚处于初始阶段,研究内容主要集中在基于认知信息的各层协议的优化设计上。本文研究网络重构技术在认知移动自组织网络中的应用和操作,在已有相关文献的研究基础上,主要研究网络参数重构和路由协议重构。本文的研究内容主要分为以下几个部分:首先,基于网络参数的多样性,对移动Ad Hoc网络的参数进行了分类,并详细介绍了不同场景下的网络参数的选取方法。其次,由于网络重构本质上是一个多目标优化问题,详细介绍了优化目标函数的构建方法,包括各性能参数效用函数的构建和多指标合成模型的选取。然后,提出了一种基于贝叶斯网络、遗传算法和案例推理的网络参数重构算法,包括在某些代表性网络场景下基于贝叶斯网络和遗传算法的策略知识库的构建方法和在新的网络场景下基于案例推理的重构推理决策方法,参数重构通过协议参数的动态调整实现网络对动态环境的适应性,仿真结果证明参数重构算法的运用能在动态环境中提升网络性能,包括端到端时延、网络吞吐量和分组递交率。最后,提出了一种基于自适应神经模糊推理系统的路由协议重构算法,通过建立路由协议性能行为模型实现不同网络场景下选择不同路由协议的目标,保证网络在动态环境中的通信性能,仿真结果表明路由协议重构算法使网络能够基于网络场景选择能提供最优通信性能的路由协议,从而验证了算法的有效性。
[Abstract]:Cognitive mobile ad hoc network is a cognitive network composed of multiple mobile nodes of equal status. It is the combination of mobile ad hoc network technology and cognitive network technology. The intelligence of cognitive network is mainly reflected in its reconfiguration ability. Network reconfiguration technology enables the network to dynamically optimize network structure and network protocol according to the changes of network state caused by the changes of internal and external environment conditions. Improve network communication performance. With the development of wireless communication network, users are demanding more and more cognitive network. As one of the key technologies of cognitive network, network reconfiguration technology has been paid more and more attention for its importance and necessity. Has become a major research focus of cognitive networks. The current network reconfiguration technology research has not formed the system, has not completely covered all aspects of the implementation of the reconfiguration, basically distributed in several details of the research points. The research on the reconstruction of cognitive mobile ad hoc networks is still in the initial stage, and the research focuses on the optimization design of the protocols based on cognitive information. In this paper, the application and operation of network reconfiguration in cognitive mobile ad hoc networks are studied. Based on the existing literatures, the network parameter reconfiguration and routing protocol reconfiguration are studied. The main research contents of this paper are as follows: firstly, based on the diversity of network parameters, the parameters of mobile Ad Hoc network are classified, and the selection methods of network parameters under different scenarios are introduced in detail. Secondly, because network reconfiguration is essentially a multi-objective optimization problem, the construction method of optimization objective function is introduced in detail, including the construction of utility function of each performance parameter and the selection of multi-index composite model. Then, a network parameter reconstruction algorithm based on Bayesian network, genetic algorithm and case-based reasoning is proposed. It includes the method of constructing the strategy knowledge base based on Bayesian network and genetic algorithm in some representative network scenarios and the method of reconstructing reasoning decision based on case-based reasoning in the new network scenario. Parameter reconstruction realizes the adaptability of network to dynamic environment through dynamic adjustment of protocol parameters. Simulation results show that the application of parameter reconstruction algorithm can improve network performance in dynamic environment, including end-to-end delay. Network throughput and packet delivery rate. Finally, a routing protocol reconfiguration algorithm based on adaptive neural fuzzy inference system is proposed. The routing protocol performance behavior model is established to achieve the goal of selecting different routing protocols in different network scenarios. The simulation results show that the routing protocol reconfiguration algorithm enables the network to select the routing protocol which can provide the optimal communication performance based on the network scenario, thus validating the effectiveness of the algorithm.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN929.5

【共引文献】