面向空间监测的三维可移动无线传感器网络理论与技术

The high-end application of space surveillance has put forward overall challenges to wireless sensor networks（WSNs）including three-dimensional (3D), mobile and multimodal signals. However, it is obvious that the existing techniques in WSNs can not face these challenges at the same time. Facing the three scientific problems-lack of topology theory in both 3D and mobile WSNs, moving algorithms rarely aim for target surveillance, and the difficulty to fuse the multimodal signals due to their different dimensions, in this project, systematic and extensive research will be carried out on novel 3D, mobile and multimodal WSNs based on preemptive research background of our team:(1) provide the nested-array theory in topology of 3D mobile WSNs;(2) propose coordinating movement algorithms for multi-sensor based on target learning;(3) perform multimodal information fusion and collaborative imaging via deep learning，etc. This project will not only offer frontier and theoretical foundation of intelligentized WSNs, but also provide important technical support for applications such as target surveillance, anti-terrorism and anti-riot, disaster assistance. Therefore, it has both great scientific and real-application values.

面向高端应用的空间监测对无线传感器网络提出了三维立体、可移动、信号多模态化这三个方面的全面挑战。但现有的无线传感器网络技术显然难以同时应对这些挑战。针对三维移动无线传感器网络拓扑设计理论稀缺、传感器移动算法甚少考虑目标监测性能、多模态信号维数不同难以融合的三个关键科学问题，本项目在研究团队已有先发研究背景的基础上，系统和深入地开展三维、移动、多模态无线传感器网络的创新性研究:(1)建立基于嵌套布置的三维可移动无线传感器网络拓扑设计理论;(2)提出基于目标学习的多传感器协调移动算法;(3)实现基于深度学习等方法的多模态信息融合与协同成像。本项目将为无线传感器网络的智能化发展奠定前沿理论基础，为实现空间目标监测、反恐防爆、灾害救援等应用提供重大技术支撑，因而兼具重要的科学意义和实际应用价值。

面向高端应用的空间监测对无线传感器网络提出了三维立体、可移动、信号多模态化等方面的全面挑战。本项目针对这些挑战，在多通道传感器的波形及发射波束设计方法、三维空间多传感器部署与协同移动模型、多目标跟踪数据处理架构及多模态信息融合的理论与技术等方面取得突破性进展，建立了集中式/分布式异构雷达传感器半实物仿真平台，车载毫米波雷达传感器和视频传感器多模态融合系统，以及包含显示界面的三维可移动无线传感器网络可视化空间监测软硬件应用示范系统。总共发表SCI收录国际期刊论文40篇，其中本领域Top期刊15篇，EI收录国际会议论文36篇，其中国际会议最佳论文奖3篇，Spinger专著一章；授权专利13项；培养博士生6人，硕士生28人；主办国际学术会议5次。本项目为我国开发具有自主知识产权的智能化无线传感器网络空间监测系统提供了理论指导与实测论证，为空间移动目标的反制提供了全新的技术途径。本项目所提理论与技术可广泛应用于安防监控、灾害救援、环境保护、自动驾驶、数字交通等领域。