雷达传感网部署与融合关键技术
发布时间：2019-06-21 03:33
【摘要】：分布式雷达传感器(Radar Sensor,RS)是一个独立小型系统,它发送已知波形,接收、分析目标和障碍物的回波以确定与目标相关的信息。雷达传感网(Radar Sensor Network,RSN)就是由分布在广阔的地理区域内、具有协同工作模式的RS节点有机地组成的。它们可以用于监测大面积区域,并从多个角度观测目标。RSN在反射物体(例如飞机、船舶、车辆以及人等)的检测、定位、跟踪等领域越来越重要。然而,RSN的目标检测性能受其关键技术(如部署方式、分簇算法、融合方式等)的影响很大。本文以提高监测区域内目标检测概率为目的,针对雷达传感器网络中的几个关键技术进行了研究。主要研究内容包括:1.搭建了适用于雷达传感器网络的数据传输模型和目标检测系统,确立了本文RSN关键技术研究的基础。2.本文提出了两种适用于RSN的节点部署方法:Hexagonal Deployment Strategy(HDS)和Diamond Deployment Strategy(DDS)。该两种方法是在使RS节点均匀分布于监测区域的思想下被提出的。在保证一定目标虚警概率的前提下,通过这两种部署方式提高目标检测概率。仿真实验证明,无论从目标检测概率还是消耗的平均能量的表现来看,DDS部署策略都优于HDS部署策略,而这两种部署方式都优于Random Deployment Strategy(DDS)部署策略。所以对于RSN网络,DDS和HDS部署策略都是有效的部署策略。3.本文研究了无线传感器网络中的两种经典分簇算法:LEACH算法和HEED算法,并将其应用于RSN中,研究它们在RSN系统中的适用性。本文还将这两种分簇算法应用到经由DDS部署方式确定了RS位置的RSN中,从而提高系统性能。蒙特卡罗仿真表明,同等条件下,按HEED算法分簇的RSN网络比按LEACH算法分簇的网络具有更长的网络生存时间。在采用HEED算法分簇时,由DDS部署的RSN网络的目标检测概率大于由RDS部署的RSN网络的检测概率。4.本文在Path-Loss信道衰落模型下,提出了两种适用于RSN的检测级数据融合方法:Decision Fusion Rules with Binary Transmission(BT)和Decision Fusion Rules without Binary Transmission(NBT)。针对多跳的RSN,将这两种融合方法与两种部署方式相结合,进一步提高RSN目标检测性能。仿真结果表明,相比于NBT融合算法,BT融合算法下RSN的检测概率高,消耗的能量少。
[Abstract]:Distributed Radar Sensor (Radar Sensor,RS) is an independent small system, which transmits known waveforms, receives and analyzes the echoes of targets and obstacles to determine the information related to the targets. Radar sensor network (Radar Sensor Network,RSN) is composed of RS nodes which are distributed in a wide geographical area and have cooperative working mode. They can be used to monitor large areas and observe targets from many angles. RSN is becoming more and more important in the fields of detection, positioning and tracking of reflected objects (such as aircraft, ships, vehicles and people, etc.). However, the target detection performance of RSN is greatly affected by its key technologies (such as deployment mode, clustering algorithm, fusion mode, etc.). In order to improve the detection probability of targets in the monitoring area, several key technologies in radar sensor networks are studied in this paper. The main research contents are as follows: 1. The data transmission model and target detection system suitable for radar sensor network are built, and the foundation of RSN key technology research in this paper is established. 2. In this paper, two node deployment methods: Hexagonal Deployment Strategy (HDS) and Diamond Deployment Strategy (DDS). For RSN are proposed. The two methods are proposed under the idea that RS nodes are evenly distributed in the monitoring area. On the premise of ensuring the false alarm probability of a certain target, the target detection probability is improved by these two deployment methods. The simulation results show that the DDS deployment strategy is superior to the HDS deployment strategy in terms of the target detection probability and the average energy consumed, and both of them are superior to the Random Deployment Strategy (DDS) deployment strategy. So for RSN networks, DDS and HDS deployment policies are effective deployment strategies. In this paper, two classical clustering algorithms in wireless sensor networks, LEACH algorithm and HEED algorithm, are studied and applied to RSN, and their applicability in RSN system is studied. In this paper, the two clustering algorithms are also applied to the RSN where the location of RS is determined by DDS deployment, so as to improve the performance of the system. Monte Carlo simulation shows that under the same conditions, the RSN network clustering according to HEED algorithm has longer network survival time than the network grouped according to LEACH algorithm. When HEED algorithm is used for clustering, the target detection probability of RSN network deployed by DDS is higher than that of RSN network deployed by RDS. 4. In this paper, two detection level data fusion methods: Decision Fusion Rules with Binary Transmission (BT) and Decision Fusion Rules without Binary Transmission (NBT). For Path-Loss channel fading model are proposed, which are suitable for RSN. For multi-hop RSN, these two fusion methods are combined with the two deployment methods to further improve the performance of RSN target detection. The simulation results show that compared with NBT fusion algorithm, BT fusion algorithm has higher detection probability and less energy consumption.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN958;TP212.9
本文编号：2503754
[Abstract]:Distributed Radar Sensor (Radar Sensor,RS) is an independent small system, which transmits known waveforms, receives and analyzes the echoes of targets and obstacles to determine the information related to the targets. Radar sensor network (Radar Sensor Network,RSN) is composed of RS nodes which are distributed in a wide geographical area and have cooperative working mode. They can be used to monitor large areas and observe targets from many angles. RSN is becoming more and more important in the fields of detection, positioning and tracking of reflected objects (such as aircraft, ships, vehicles and people, etc.). However, the target detection performance of RSN is greatly affected by its key technologies (such as deployment mode, clustering algorithm, fusion mode, etc.). In order to improve the detection probability of targets in the monitoring area, several key technologies in radar sensor networks are studied in this paper. The main research contents are as follows: 1. The data transmission model and target detection system suitable for radar sensor network are built, and the foundation of RSN key technology research in this paper is established. 2. In this paper, two node deployment methods: Hexagonal Deployment Strategy (HDS) and Diamond Deployment Strategy (DDS). For RSN are proposed. The two methods are proposed under the idea that RS nodes are evenly distributed in the monitoring area. On the premise of ensuring the false alarm probability of a certain target, the target detection probability is improved by these two deployment methods. The simulation results show that the DDS deployment strategy is superior to the HDS deployment strategy in terms of the target detection probability and the average energy consumed, and both of them are superior to the Random Deployment Strategy (DDS) deployment strategy. So for RSN networks, DDS and HDS deployment policies are effective deployment strategies. In this paper, two classical clustering algorithms in wireless sensor networks, LEACH algorithm and HEED algorithm, are studied and applied to RSN, and their applicability in RSN system is studied. In this paper, the two clustering algorithms are also applied to the RSN where the location of RS is determined by DDS deployment, so as to improve the performance of the system. Monte Carlo simulation shows that under the same conditions, the RSN network clustering according to HEED algorithm has longer network survival time than the network grouped according to LEACH algorithm. When HEED algorithm is used for clustering, the target detection probability of RSN network deployed by DDS is higher than that of RSN network deployed by RDS. 4. In this paper, two detection level data fusion methods: Decision Fusion Rules with Binary Transmission (BT) and Decision Fusion Rules without Binary Transmission (NBT). For Path-Loss channel fading model are proposed, which are suitable for RSN. For multi-hop RSN, these two fusion methods are combined with the two deployment methods to further improve the performance of RSN target detection. The simulation results show that compared with NBT fusion algorithm, BT fusion algorithm has higher detection probability and less energy consumption.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN958;TP212.9
【参考文献】
相关博士学位论文 前2条
1 王蕊;无线传感器网络部署与拓扑控制研究[D];哈尔滨工程大学;2009年
2 谭振江;多传感器光测系统数据融合技术的应用研究[D];中国科学院研究生院（长春光学精密机械与物理研究所）;2003年
相关硕士学位论文 前1条
1 王宁;雷达传感器与红外传感器信息融合技术在无人机上应用研究[D];沈阳航空航天大学;2011年
，本文编号：2503754
本文链接：https://www.wllwen.com/kejilunwen/wltx/2503754.html
