基于簇的异构无线传感器网络能量有效路由算法研究

【作者】 周海波；

【导师】 吴援明；

【作者基本信息】 电子科技大学 ， 信号与信息处理， 2010， 硕士

【摘要】 无线传感器网络(Wireless Sensor Networks,WSNs)作为一种新型的融合计算机、通信、传感器等多学科的信息获取和处理技术,其具有广泛的应用前景,受到了来自军界、学术界和工业界的普遍推崇与青睐。目前,WSNs的研究主要集中于同构WSNs领域。相对于同构网,异构无线传感器网络(Heterogeneous WSNs,HWSNs)由具有不同感测能力、计算处理能力、通信能力和初始能量配置的传感器节点组成,是同构WSNs在理论及应用领域研究的推广与深入,与实际应用场景更接近,更能满足广泛的异构世界对不同WSNs网络模型的需求。同时,由于传感器节点的电池能量有限,而且部署区域环境复杂,对其进行能量补给十分困难,如何实现网络的能量有效也是HWSNs首要研究的问题之一。目前,HWSNs的能量有效性研究主要建立在分簇模型上。本文针对HWSNs的异构特性,对异构分簇能量有效路由算法进行了探索性研究,提出了一种新的异构分簇能量有效路由算法。本文所作的主要工作如下:第一、构建异构网络模型。通过对当前同构WSNs网络模型的研究,并结合实际应用场景的需求提出了完整的HWSNs网络模型;论述了HWSNs的四种异构表现形式、异构网的特点与挑战、关键技术与性能评价指标;介绍了HWSNs的体系结构,为设计新的异构分簇路由算法奠定基础。第二、规划异构网络的能量消耗模型。在异构网络模型提出的基础上,对异构分簇网络模型的能量消耗进行了数学规划。以异构WSNs的生存周期最长为目标函数进行数学建模,推导出不同异构模型下求取优化簇头数目的函数,建立了同构模型和异构模型之间的统一,为HWSNs能量有效分簇路由算法的设计提供指导。第三、设计与验证EDFCM算法。通过综合考虑节点的初始能量、能量消耗速率及剩余能量等因素提出了基于能量消耗预测的分簇路由算法EDFCM。同时,为了保证稳定的簇头数目及传感任务的有效执行,EDFCM算法引入了簇头管理机制。为了对该算法进行综合的性能评价,本文在Matlab 2007上对EDFCM算法进行了模拟仿真。仿真结果表明:EDFCM算法改进了传统无线传感器网络分簇路由算法中簇头数目不稳定的缺陷并有效延长了网络的生存周期。在异构网络的稳定周期指标上,EDFCM算法较经典的LEACH算法与DEEC算法分别提高了200%与18%;此外,在数据的有效传送指标上,较前两种算法EDFCM分别提高了50%与14%。上述充分验证了EDFCM算法是适用于HWSNs的。更多还原

【Abstract】 As a new paradigm of technology in the range of information acquisition and processing that combines multi-subject such as computer science, communications, sensor technology and etc., wireless sensor networks (WSNs) are in a wide range of applications and very popular in the military fields, academic community and industry. Recently, the main researches on WSNs are focused on the homogeneous WSNs. Different from the homogeneous WSNs, heterogeneous WSNs (HWSNs ) are composed of varied types of sensor nodes with different sensing abilities, computation abilities, communications abilities and initial energy settings. HWSNs models are more suitable to the actual applications fields than those of the homogeneous WSNs, and can meet the demands of various applications for the technologies of sensor networks. In a word, HWSNs generalize the homogeneous WSNs.However, due to the limited energy setting in sensor nodes and complicated environments in application fields, it is hard to re-deploy energy for the running sensor networks. Thus, how to enhance the energy efficiency in HWSNs is a key technique. Presently, the energy efficiency researches are mainly focused on the clustering model. Based on the heterogeneity in HWSNs, the energy efficient clustered routing algorithms have been explored and a novel protocol for HWSNs is designed in this dissertation, especially. The main research works are listed in the following:1. The modeling about the HWSNs. Through the researches on the homogeneous WSNs models and the considerations about the actual applications, an integrated heterogeneous network model is proposed. Combining the heterogeneity, the characteristics, challenges, key techniques and the performance evaluations in HWSNs are discussed. In addition, the architecture of HWSNs is introduced.2. The programming of energy consumption in HWSNs. Based on the definition about the heterogeneous network, the mathematical programming of the energy consumption in clustered heterogeneous model is presented. Through an analysis of energy dissipation model, a function of the optimum number of clusters for HWSNs is obtained and the relationship between homogeneous WSNs and HWSNs is discussed, which indicate that homogeneous WSNs is just a special form in HWSNs.3. The design and evaluation of EDFCM algorithm. Based on the energy dissipation forecast scheme, EDFCM is proposed. The nodes’initial energy, energy consumption rate and the residual energy in the nodes are considered in this algorithm. Moreover, a cluster heads management scheme is introduced in EDFCM to guarantee the reliable transmission in HWSNs. We evaluate the performance of EDFCM via MATLAB 2007 simulations. Simulation results indicate that the number of cluster heads each round is stable and the dead time of the first dead node in our EDFCM can be prolonged nearly by 18% and 200% when compared with DEEC and LEACH, respectively. In addition, the amount of messages received by BS will increase by 14% and 50%, respectively. All the results above show EDFCM to be of advantage to HWSNs.更多还原