【作者】 王磊；

【导师】 杨余旺；

【作者基本信息】 南京理工大学 ， 计算机应用技术， 2014， 博士

【摘要】 近年来,技术的进步和制造成本的降低使得无线传感器网络(WSN, Wireless Sensor Network)得到了突飞猛进的发展。由于传感器节点的易部署性和多功能性,WSN被广泛应用到各种领域,如医疗卫生、目标跟踪、环境监控等。传感器节点的主要任务是收集、保存数据,并将数据传输给汇聚节点做进一步的操作。传感器节点能量的有限性和无线链路的不可靠性给设计高效的通信、存储机制带来了巨大的挑战。路由技术是无线传感器网络中的一种重要的基础支撑技术,路由协议的好坏关系到组网的质量。由于无线传感器网络在网络特点、业务特点和应用需求等方面与传统网络存在显著差别,无线传感器网络路由的设计面临一系列新的挑战。尤其是近年来新技术的出现为传统路由技术的发展带来了新的活力。本文研究无线传感器网络中能量高效的数据传输机制、数据存储与修复技术以及能量高效的组播技术。(1)本文研究了无线传感器网络中基于网络编码的多径路由的能量效率。采用的多径网络模型是缠绕多路径模型,所用的网络编码方案为随机线性网络编码。研究了基于网络编码的缠绕多路径中成功交付率和能量消耗的分析模型。(2)分析了网络中不同参数对网络性能的影响,如跳数,数据包长度,子包大小,簇大小等,并通过对这些参数的分析,导出一种优化的传输方案,该方案使得网络能够以更少的资源开销完成网络传输。此外,通过大量的离散事件仿真发现,实验结果与分析结果是一致的,并测试了在特殊环境下(如网络拥塞,节点失效等)的网络性能。(3)本文提出了一种无线传感器网络中能量高效的数据存储和修复方案。通过构造具有稀疏特性的网络码字来实现能量高效的分布式存储。此外,利用网络编码在中间节点进行编码的思想,设计了一种迭代的修复策略,该方法使得在无线多跳网络中能够高效地进行数据恢复,节约了数据包传输的次数。并通过理论证明了该方案可以工作在非常小的有限域上。最后通过实验证明,该方法相比传统方法在降低能量消耗的同时,还能均衡网络中的能量负载,从而能够提升网络的生存时间。(4)提出将多速率网络编码方法应用于无线多跳网络的组播通信中,该方法能够显著提升组播网络的吞吐量。并且提出了一种多速率网络码字的构造方法,该方法在绝大多数的网络情况下,都能使得接收节点以各自的最大流速率接收数据。并且提出了一个“流对齐”的方法来分离网络中的独立链路,从而显著地降低网络的复杂度。(5)提出了一个充分条件,当该条件满足时,网络中的所有目的节点均能以最大流的速率接收数据。大量的实验证明,该方法所提出的多速率编码方法能够提升网络的吞吐量,降低数据传输的能量消耗,并保证网络中接收节点的解码成功率。更多还原

【Abstract】 In recent years, the advancement of new technology and decreament of the cost of manufactory accelate the development of Wireless Sensor Networks (WSNs). Since the sensor nodes are easy to be deployed and have various functions, they are widely used in different fields such as medicalcare, target tracking, environment mornitoring and so on. The main task of sensor nodes is to collect data, store data, and transmit the data to a sink node for further processing and operation. The limitation of sensor nodes and unreliability make the designment of efficient communication, storage mechanism become significant challenges.Routing technology is an important supporting technology of WSNs, the performance of routing protocol has a significant influence on the quality of networking. Because there are many differences in the feature of network, feature of business and application requirement between WSNs and traditional networks, the routing design of WSNs faces a series of new challenges. This dissertation addressed reliable data transmssion technology, data storage technology and multicast technology with high energy efficiency.(1) In this dissertation, we studied the energy efficiency of network-coding-based multipath routing technology for WSNs. The employed multipath network model is the braided routing model, and the employed network coding scheme is the random linear network coding. Moreover, we studied the Successful Delivery Ratio (SDR) and Energy Comsumption (EC) of the network-coding-based multipath routing technology.(2) We analyzed the influence of different parameters such as the number of hops, the length of packets, the size of sub-packet, the size of cluster and so on. Based on these results, an optimal scheme is achieved, which makes the network accomplish network transmission with less resource. Moreover, through extensive simulations of discrete events, we observed that the results of simulations are consistent to that of analysis. Some evaluations on extreme environment (such as congested networks and failed nodes) are carried, which shows that the proposed scheme is robust to these environment.(3) We proposed an energy-efficient distributed storage and repair scheme. Through constructing sparse generator matrix, we achieved a scheme to implement energy-efficient storage scheme. In addition, the idea that re-encoding at intermediate nodes from network coding theory is employed to design an iterative repair scheme, which significantly reduces the required transmission time, and hence making the repair in Wireless Multi-hop Networks much efficient. Through theorectical analysis, we proved that the employed scheme could work over a very small finite field, which implies that the computation overhead would be low. Finally, through extensive experiments, we showed the scheme could reduce the total energy consumption, and in the mean time, the energy consumption is highly balanced. Therefore, the lifetime of the network is increased.(4) We introduced the multi-rate network coding into the communication of Wireless Multi-hop Networks, which could increase the throughput of multicast networks. And we proposed a multi-rate network codes construction method. By using this method, the receivers could receive at their individual max-flows in many practical applications. Moreover, we provided a method, called Flow-Alignment, to separate independent routes from the topologies, which could ruduce the computation complexity of multi-rate network codes.(5) We provided a sufficient condition to judge whether the transmssion scheme that makes all the receivers receive at individual max-flow exists. When the condition is satisfied, the goal can be achieved. Through extensive experiments, we observed that the proposed coding scheme could increase the throughput of network, reduce the energy consumption of data transmission, and guarantee the decodabiltiy of receivers.更多还原