无线多媒体传感器网络服务质量保障问题的研究

发布时间：2018-03-04 06:24

本文选题：无线多媒传感器网络　切入点：服务质量保障　出处：《河北工业大学》2014年博士论文　论文类型：学位论文

【摘要】：低成本硬件如CMOS摄像机、麦克风在无线传感器网络（WSN）的可用性，推进了无线多媒体传感器网络（WMSNs）的发展。资源受限的传感器节点需检测包括音/视频、图像等多媒体数据及标量传感数据并为这些数据提供服务质量保障，面临巨大挑战。利用服务区分方法按应用需求进行业务分类，保障优先级更高的业务拥有更多的网络资源进行服务；利用跨层联合优化高效利用网络资源，充分利用各层可利用的参数及技术更好的提供WMSNs服务质量保障。本文利用服务区分方法和跨层优化方法对WMSNs进行服务质量保障，所做的主要工作如下：基于扁平网络结构，根据期限要求进行服务区分，结合物理层、MAC层和网络层联合优化，提出能效、实时保障的通信机制。按接收信号强度进行分层，发送端按每跳延迟期限结合节点能量进行优先级排队，采用轮询争用的实时MAC转发数据，接收端根据转发速度、能效竞争接收数据，转发过程中进行简单的许可控制、尽早丢弃逾期数据，有效控制网络负载避免拥塞。仿真结果表明能提供更高的投递率和吞吐量，有效的降低了节点能耗，延长了网络生存周期。基于分层网络结构，联合优化物理层、MAC层和网络层，建立了有效跨层参数的WMSNs服务质量保障路由模型，提出多蚁群多路径路由算法，利用多蚁群算法寻找满足度量的最优路由。利用分簇蚂蚁处理网络的可扩展性，处理节点能量持续问题；前向蚂蚁、反馈蚂蚁以及视频蚂蚁分别用于最优路径及视频传输；维护蚂蚁处理拥塞问题和链路失效问题，多种群蚂蚁分工合作。满足应用需求的路径利用增强因子积累更多信息素吸引更多同类蚂蚁，相反挥发信息素避免蚂蚁到达。仿真实验表明，新的路由算法有效的保障了WMSNs的多个QoS度量。基于树形网络结构，对传感器节点提供的业务划分优先级，结合MAC层、传输层联合优化，提出新的拥塞控制方法。拥塞指示结合节点剩余缓存检测拥塞，MAC层中信道访问时为不同业务分配不同的控制报文，保证高优先级业务快速访问信道、拒绝低优先级业务访问，，有效降低局部拥塞的同时保障高优先级业务的QoS需求；测量Sink及父节点的速率误差按业务优先级调节业务速率进行全局拥塞的控制。仿真结果表明，新方法能在随机服务时间、节点失效的情况下获得比CCF、PCCP更优的吞吐量、延迟，保障了优先级高的业务优先享受网络资源获得更好性能。
[Abstract]:Low-cost hardware such as CMOS cameras, the availability of microphones in wireless sensor networks (WSNs) has pushed forward the development of wireless multimedia sensor networks (WMSNs). Sensor nodes with limited resources need to be detected, including audio / video. Multimedia data and scalar sensing data, such as images, and providing quality of service (QoS) guarantee for these data face great challenges. To ensure that higher priority services have more network resources for service, cross-layer joint optimization and efficient use of network resources, Make full use of the parameters and technologies available in each layer to provide better WMSNs quality of service assurance. In this paper, we use the method of service differentiation and cross-layer optimization to guarantee the quality of service of WMSNs. The main work is as follows:. Based on flat network structure, service differentiation according to deadline requirement, combined with physical layer MAC layer and network layer, the communication mechanism of energy efficiency and real-time guarantee is proposed. The sender queue according to the delay duration of each hop combined with the node energy, adopts the real-time MAC to transmit the data, and the receiver competes to receive the data according to the forwarding speed, energy efficiency, and carries on the simple permission control in the process of forwarding. The simulation results show that it can provide higher delivery rate and throughput, reduce node energy consumption effectively, and prolong the network lifetime. Based on the hierarchical network structure, the physical layer MAC layer and the network layer are jointly optimized, the WMSNs QoS routing model with effective cross-layer parameters is established, and the multi-ant colony multi-path routing algorithm is proposed. Multi-ant colony algorithm is used to find the optimal route to meet the measurement. Cluster ant is used to deal with the scalability of the network and to deal with the problem of node energy persistence. The forward ant, feedback ant and video ant are used for the optimal path and video transmission, respectively. Maintenance ants deal with congestion and link failure, multi-colony ants work together. Paths that meet application requirements use enhancement factors to accumulate more pheromones to attract more ants of the same kind. The simulation results show that the new routing algorithm can effectively guarantee the multiple QoS metrics of WMSNs. Based on the tree network structure, the service provided by sensor nodes is prioritized, and combined with MAC layer, the transport layer is optimized. A new congestion control method is proposed, in which congestion indication combines with node residual buffer to detect channel access in congested MAC layer, which assigns different control packets to different services to ensure fast access to the channel for high-priority services and deny access to low-priority services. It can effectively reduce the local congestion and guarantee the QoS requirement of high-priority traffic, and measure the rate error of Sink and its parent node to control the global congestion according to the traffic priority. The simulation results show that the new method can control the global congestion at random service time. In the case of node failure, the throughput and delay are better than that of CCFN PCCP, which ensures that the high priority services enjoy better network resource performance.
【学位授予单位】：河北工业大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TP212.9;TN919.8

【参考文献】