农田无线传感器网络仿蛛网建模及高抗毁关键技术研究

Since complexity of farmland environment and changeable characteristics of crops, higher invulnerability of special wireless sensor network is required. Creating a strong adaptability of network prototype system, as well as connecting the cross-layer optimization mechanism between link layer and network layer is the fundamental way to achieve this requirement. The multipath and self-healing characteristic of cobweb provides a useful reference for farmland wireless sensor network. This project which bionic object is based on the cobweb, has established a structural model of cobweb in order to analyze the association mechanism between structure characteristics and invulnerability, dissect the influence of cobweb configuration parameters on vibration information transmission, explore the effect law of network destruction to the information transmission and finally construct a bionic cobweb model characterized by the invulnerability of topology and information transmission. Based on the cobweb model, takes connectivity, cost, power consumption and optimal reliability as constraint condition to explore the static and dynamic topology management mechanism under the driving condition of invulnerability. Proposing a comprehensive evaluation algorithm of route invulnerability and using the multi-objective routing optimization method to find the best balance among hop count, energy consumption and balancing capability of network load, further more to find out the dynamic hierarchical routing control strategy with local reconfiguration ability. The research results will improve the invulnerability of farmland wireless sensor network, provide technical support for invulnerability research of other general computer networks and wireless communication network.

农田环境复杂、作物特征多变对专用无线传感器网络的抗毁性提出了更高要求，创建强适应性网络原型系统，突破链路层和网络层的跨层协同优化机制，是实现这一要求的根本途径。蜘蛛网的多径性、自愈性，为农田无线传感器网络构建与抗毁性提升提供有益借鉴。本项目对蜘蛛网进行结构和信息传输方式分析，研究抗毁特性机理，构建具有高抗毁性拓扑结构和路由传输特点的人工蜘蛛网模型；以连通度、成本、功耗和可靠性综合最优为约束条件，探寻抗毁性能驱动条件下的静动态拓扑管理机制；提出路由抗毁度综合评判算法，利用多目标优化方法寻求跳数、能耗和网络负载均衡能力之间的最佳平衡，探明局部重构容错的动态分层路由控制策略。研究成果将改善农田无线传感器网络的抗毁性能，也为其他计算机网络、无线通信网络的抗毁性研究提供技术支撑。

农田无线传感器网络（FWSNs，farmland wireless sensor networks）具有大规模、能量成本受限、节点众多、拓扑变化复杂等特点，对网络抗毁性能提出了严苛的要求。蜘蛛网与无线传感器网络在拓扑结构方面具有显著的相似之处，将自然界中蜘蛛网所具备的抗毁性独特优势与无线传感器网络通信技术相结合有较高研究价值。通过系统研究，本项目形成一套提升FWSNs抗毁性的建模与控制算法，提高FWSNs的可靠性，具体研究内容如下：.（1）开发一种基于3D打印的螺旋式人工蛛网振动测试试验装置，用于研究给定激励条件下蜘蛛网的振动信息传输规律，提炼人工蛛网的振动特征与有中心分层无线传感器网络的相似性。.（2）建立人工蛛网网络拓扑模型，采用中心节点、蛛网层数、单层节点总数等参数描述网络拓扑结构，采用端到端延时分析链路、节点破坏时对网络传输性能的影响；构建基于仿蛛网的负载容量模型及流量分配机制，以有效节点占比、网络效率比和仿真轮数为评价抑制网络抗级联失效性能的指标；建立基于层次分析的兴趣信息更新方法，通过消息内容的相似程度比较和阈值层次模型的计算动态调整阈值，根据更新触发机制更好地适应信息交换率的要求。. （3）建立一种基于节点平均路径数和节点、链路平均使用次数的人工蛛网模型抗毁性量化指标，用于评测失效网络组件的全网影响度和权重；构建基于仿蛛网分层分簇拓扑结构的旋转路由能量均衡协议，建立一种包含半径系数ε、簇头候选区分区可调参数Z、簇头候选区角度可调参数α、簇头候选区旋转角可调参数β、簇头选举因子调节系数χ的网络模型，利用仿真轮数和能量消耗速率来评价网络的抗毁性能；利用正态性检验，方差齐性检验和多因素方差分析等统计学方法，解析结构特征与网络抗毁性和全网平均剩余能量之间的关联机制，量化分析各网络模型参数的重要性程度，将影响程度作为网络模型参数优化计算的依据，网络抗毁性和全网平均剩余能量作为优化目标函数，利用多目标优化算法进行网络参数优化。