【作者】 黄联芬；

【导师】 陈辉煌； 姚彦；

【作者基本信息】 厦门大学 ， 环境科学， 2008， 博士

【摘要】 当前,宽带无线通信是通信领域研究的热点。本文研究一种基于环境认知的宽带无线网络——认知无线Mesh网,它将认知无线电技术应用于无线Mesh网中,是具备环境认知、高频谱效率、高度灵活性、宽覆盖范围的新型宽带无线网络。本文在研究认知无线Mesh网络多信道MAC协议的基础上,重点针对基于频谱感知、基于信道质量感知及基于路由感知的MAC协议进行了相关的研究。首先在介绍IEEE802.11无线Mesh网MAC协议的基础上,对单网卡多信道MMAC协议与双网卡多信道DCA协议进行详细介绍和性能仿真分析,提出了无线Mesh网多信道MAC协议的数学模型,其包含Bernoulli模型和M/M/K排队模型,针对DCA协议进行了建模性能仿真验证,仿真结果和数学分析结果很接近,建立的模型比较准确。在研究认知无线电频谱感知技术的基础上,将其引入到无线Mesh网的MMAC和DCA多信道协议中,提出基于频谱感知的CWMN-MMAC和CWMN-DCA多信道MAC协议。在协议中增加一个专门控制窗口进行频谱探测的信息发布,建立了主用户信道利用率模型;采用基于随机感知策略的动态检测信道模型,得到网络饱和吞吐量数学公式;且针对主用户活动情况,对信道利用率及次用户对主网络性能影响等方面进行验证分析;针对同频干扰,在信道协商前增加频点识别功能,提出了基于频点感知的多信道CC-MMAC协议,仿真验证协议有效地减轻了同频干扰。受衰落和色散的影响,信道质量是随时间变化的,从而对网络的性能造成一定的影响,因此建立了高斯信道和衰落信道的MAC协议模型;在研究速率自适应的基础上,对速率自适应的MAC协议进行性能分析,提出基于机会速率自适应的OAR—DCA多信道协议,有效地提高了网络的吞吐量。在研究多径路由AODV协议的基础上,将路由层的下一跳思想引入到MAC层上,提出建立多个下一跳的MNH—MMAC协议:并对AODV路由协议进行改进,建立无连接限制的NDL-AODV路由协议,将NDL-AODV+MNH-MMAC协议与原有AODV+MMAC协议进行仿真分析和比较,它可在频率资源比较紧张条件下,改善网络的性能。更多还原

【Abstract】 Recently,broadband wireless communication becomes the hot research of communication area.This thesis studies the broadband wireless network based on the environment cognition,known as cognitive wireless mesh network,which applies the cognitive technology into wireless mesh network. It is a new broadband wireless network having environment cognition,high spectral efficiency,high flexibility and wide coverage.In this thesis, we conduct the research of wireless mesh network mutli-channel protocol,and then mainly study the MAC protocol based on the frequency sensing,channel quality sensing and route sensing.Firstly,we introduce the Multi-channel MAC protocol and Dynamic Channel Assignment in detail and then analyze their performances through simulation,based on introduction of the MAC protocol of wireless local area network and wireless mesh network.Then,we present the mathematical model of Multi-channel MAC protocol of wireless mesh network,including the Bernoulli model and M/M/K queuing model.We take a DCA protocol for example to validate our model.The experiments show that simulation results match the analytical results very well,which indicates that our model is accurate.Based on the research of wieless cognitive radio technology,we introduce it into the MMAC and DCA protocol in wireless mesh network and propose the CWMN-MMAC and CWMN-DCA Multi-channel MAC protocol based on the frequency cognition,in which we add a dedicated control window to broadcast the frequency information.We build the channel use ratio model of prime user,the dynamical channel detection model based on random cognitive method,and then obtain the mathematical expression of network saturation throughput.With regard to the activity of prime user,we analyze the channel use ratio and the interference introduced by the secondary user. For the co-frequency interference,we add the channel coginition before the channel negotiation and propose the Channel Cognitive Multi-channel MAC protocol.The simulation results show our method is effective lighten co-frequency interfence. Due to the effect of fading and dispersion,channel quality varies as the time change,which has the significant impact on the network performance. Thus,we build the MAC protocol model based on the Gaussian channel and fading channel.Based on the study of rate-adaptation mechanism,we study the performance of rate-adaptation MAC protocol and propose the Oportunistic DCA protocol,which effectively improve the network throughput.Based on the research of Multi-path Route AODV protocol,we build the multiple next-hop MNH-MMAC protocol based on the introduction of the idea of next-hop in route into MAC layer and NDL-AODL route protocol based on the improvement of AODV protocol.Then we compare the NDL-AODV+MNH-MMAC protocol and AODV+MMAC protocol through simulation.The simulation results show that NDL-AODV+MNH-MMAC can improve the network performance under the relatively scare frequency environment更多还原