超声波传感器认知体域网络的构建

Most Body area networks are based on wireless sensors with the lack of cross-layer awareness, self-autonomy. This project is for wearable body area network by using ultrasonic sensors, whether it is in the reduction of the radiation hazards of the sensing equipment on the human body, or ultrasonic sensors cognitive wireless network on key issues facing the cognitive, dynamic adaptation, adaptive topology control, network communications, It has an great scientific significance and promotional value both in the reduction of the radiation hazards on the human body and in the key issues of cognition, dynamic adaptation, adaptive topology control and network communications. Through the establishment of the cognitive radio body area network architecture, cognitive loop design methods under cross-layer mechanism can be studied to solve the problem of network awareness. The problem of dynamic adaptation of the network can be discussed with the introduction of wireless body area network node motion model, user context-aware model and Gaussian process machine learning methods. "Game of Life" cellular automata is studied to adjust nodes' status in order to solve the adaptive topology control problems. As to network communication problem, the intelligent Agent system, roaming human data collection network to effect the packets' transmission, is introduced. Simulation platform about networks and cognition is built to test the availability and reliability of the theoretical system to improve its impact on the relevant theories and methods. The availability and reliability of the theoretical system are tested with simulation platform about networks and cognition in order to improve the system's relevant theories and methods.

目前体域网大都是架构在无线电传感器上，并且缺少跨层认知、自主决策等方面的特性。本项目将超声波传感器应用于穿戴式体域网，并构建基于超声波传感器的认知无线体域网络，无论是在消减其相关感知设备对人体的辐射危害，还是在解决超声波传感器认知无线网络所面临的认知、动态适配、自适应拓扑控制、网络通信等关键问题，都有着重要的科学意义和推广价值。通过建立相应的体域网认知无线网络架构模型，研究跨层机制作用下的认知环机理，解决网络认识性问题；引入无线体域网节点运动模型，通过用户环境感知和高斯过程机器学习的方法，实现节点连通性的预测，解决网络的动态适配问题；研究"生命游戏"元胞自动机指导节点状态调整，解决自适应拓扑控制问题；针对网络通信问题，引入智能Agent系统，漫游人体数据采集网络，指导报文传输。通过搭建网络与认知仿真实验环境，检验该理论体系的可用性与可靠性，以其对有关理论和方法进行改进。

可穿戴式或植入式的体域网通常使用无线电波作为通信介质，虽然通信效果较好，但长时间近距离的电磁辐射会危害人体健康。本项目将超声波传感器应用于体域网，并构建基于超声波传感器的认知无线体域网络，不仅消减相关感知设备对人体的辐射危害，同时解决了超声波传感器认知无线网络的认知、动态适配、自适应拓扑和网络通信等关键问题。从超声波传感器的特殊性出发，建立超声波体域网的体系架构和认知模型；通过探讨人体运动规律，研究节点运动模型，解决超声波体域网与环境的动态适配问题；通过研究节点的机会连通性，达到网络拓扑的自适应控制；提出并制定体域网无线节点在容忍延迟网络环境下的路由策略和缓存控制方案，提高网络生命周期和路由性能；通过探索链路连通性变化和超声波传输特性，建立自适应的超声波通信模型，解决数据传输问题。本项目较为完整地构建了基于超声波的认知无线体域网络，完成了项目预期的研究目标、研究内容以及成果要求。项目组共培养5位博士研究生、14位硕士研究生，已有4位博士、13位硕士毕业。已发表学术论文35篇， 完成授权专利5项，其中发明专利3项、实用新型专利2项。