半封闭狭长空间无线数据传输可靠性关键技术研究

Wireless sensor network plays important role for the safety monitoring of a kind of semi-confined long-narrow spaces, such as the coal mine laneway, the ultra-long tunnel and so on, where the reliable data transmission is one of the difficult scientific problems to be solved. Aimed at the special features of the semi-confined long-narrow space, such as the harsh communication environment, the large multi-path fading, and so on , the key technologies on reliable data transmission are explored from the time synchronization, the data aggregation, and the topology control aspects. With the introduction of consensus concept, the time synchronization over fading channels gets researched. The event-triggered communication mechanism is introduced to improve the energy efficency. Moreover the consensus of the time synchronization algorithm is analyzed with the aid of matrix transform and stability theory. Furthermore, the design of event-trigger function and the synchronization parameters are also researched in a distributed way. With the constructed data utility and privacy model, the data privacy preserving problem is then researched based on the differential privacy technology. Hence, the optimization mechanism of injection noise intensity considering the fusion precision is explored, and the data aggregation mechanism considering the the privacy presering gets further clarified.The data aggregation energy efficiency is increased from the network topology control point of view. Moreover with the use of power control and coloring thought, the construction of the planar topology and the chose of its child sink nodes are solved respectively. The network adjacency weight optimization mechanism is further explored based on the weighted Schatten-p norm.The project’s research results can play an important role in environmental monitoring, equipment personnel monitoring and post-disaster relief for a kind of semi-confined long-narrow space, such as the coalmine laneway, the ultra-long tunnel and so on.

无线传感器网络为煤矿巷道、超长隧道等半封闭狭长空间的安全监控提供重要支撑，面向狭长拓扑的可靠数据传输是亟待解决的科学难题。针对半封闭狭长空间无线通信环境差、多径衰落大等特性，项目从时间同步、数据融合和拓扑控制等方面探索数据可靠传输关键机制。基于趋同思想研究衰落信道下的时间同步问题，引入事件触发机制提高同步能效性，基于矩阵变换和稳定性理论研究同步算法收敛性，阐明衰落信道下触发方程、同步参数分布式设计机理；建立数据可用性和隐私性度量模型，基于差分隐私技术研究数据隐私保护方法，阐明考虑融合精度的注入噪声强度优化机理，探索考虑隐私保护的数据融合机制；从网络拓扑角度探索数据融合能效性提高，基于功率控制和着色思想构建平面拓扑以及进行子网关节点选择，阐明基于加权Schatten-p范数的网络邻接权重优化机制。项目研究成果对煤矿巷道、超长隧道等狭长空间的环境监控、设备人员监控及灾后救援等起重要支撑作用。

随着物联网与移动计算的快速发展，无线传感器网络所承载的应用类型与日俱增，在矿井巷道、超长隧道等半封闭狭长空间得到越来越多的应用，无线数据传输性能成为该类应用系统能否成功部署的关键。本项目针对半封闭狭长空间WSNs，开展无线通信环境差、多径衰落大、节点能量受限下的无线数据可靠传输关键技术研究，主要包括：1）多径衰落信道下时间同步研究。开展了多径衰落信道下的时间同步研究，分析了半封闭狭长空间无线衰落信道特性并建立其统计模型，从一致性理论角度研究网络节点的时间同步，分析了衰落信道对同步精度的影响；探索了一种事件触发通信机制，给出事件触发方程和触发参数的完全分布式设计方法，阐明衰落信道下事件触发时间同步算法收敛的充分必要条件；构建了有线和无线融合的时间敏感网络架构，基于三层网络模型研究一类时间敏感网络中分布式精确时间同步问题，提高了时间同步数据传输的实时性。2）考虑隐私保护的数据融合研究。研究了一种基于噪声添加的隐私保护方法，引入隐私保护度指标以衡量用户隐私的泄露程度，基于概率统计方法分析所提方法的数据保护隐私度；阐明了差分噪声信号强度的分布式确定方法计算方法，形成综合考虑数据隐私性和可用性的数据融合机制；研究了一种基于对称加密和双层真值发现的激励机制，提出了基于双层真值发现模型进行数据可靠性评估的激励机制。3）基于图论的拓扑控制优化研究。基于群智慧对GPA算法构建了一种动态易维护的类簇网络拓扑；研究了一种分簇路由方法，以相对剩余能量、能量开销指标为参数，构建了基于最小生成树的最优传输路径；提出了一种长跳步可靠性保障机会路由算法，基于应用期望可靠性以及节点无线通信链路质量，优化了候选转发节点集的数量。经过四年的研究，本项目取得了系列研究成果，对半封闭狭长空间的环境监控、设备人员监控等具有重要理论价值和实际意义。