本文選題：無(wú)線傳感器網(wǎng)絡(luò) + 節(jié)點(diǎn)定位��； 參考：《南京理工大學(xué)》2016年博士論文

【摘要】：無(wú)線傳感器網(wǎng)絡(luò)(Wireless Sensor Networks,WSNs)是隨著半導(dǎo)體技術(shù)、微系統(tǒng)技術(shù)、通信技術(shù)等技術(shù)的發(fā)展而產(chǎn)生和迅速發(fā)展起來(lái)的。無(wú)線傳感器網(wǎng)絡(luò)由各類集成化的微型傳感器節(jié)點(diǎn)協(xié)同感知、采集和處理網(wǎng)絡(luò)覆蓋的地理區(qū)域中感知對(duì)象的數(shù)據(jù),通過(guò)嵌入式系統(tǒng)對(duì)數(shù)據(jù)進(jìn)行處理,并通過(guò)隨機(jī)自組織無(wú)線通信網(wǎng)絡(luò)將這些數(shù)據(jù)傳送給基站,最后通過(guò)互連網(wǎng)或衛(wèi)星網(wǎng)絡(luò)到達(dá)管理節(jié)點(diǎn)。無(wú)線傳感器網(wǎng)絡(luò)目前廣泛地應(yīng)用于國(guó)防軍事、國(guó)家安全、環(huán)境監(jiān)測(cè)和醫(yī)療衛(wèi)生等領(lǐng)域。而在這些應(yīng)用研究中,節(jié)點(diǎn)定位和網(wǎng)絡(luò)覆蓋是無(wú)線傳感器網(wǎng)絡(luò)應(yīng)用的兩個(gè)研究熱點(diǎn)。節(jié)點(diǎn)的位置信息對(duì)于實(shí)現(xiàn)無(wú)線傳感器網(wǎng)絡(luò)眾多應(yīng)用起到至關(guān)重要的作用;而網(wǎng)絡(luò)覆蓋則決定了無(wú)線傳感器網(wǎng)絡(luò)所能提供的服務(wù)范圍,也在很大程度上影響了網(wǎng)絡(luò)的成本和各種具體應(yīng)用的性能。本文針對(duì)無(wú)線傳感器網(wǎng)絡(luò)中的節(jié)點(diǎn)定位與網(wǎng)絡(luò)覆蓋技術(shù)進(jìn)行了比較深入的研究與探討。在節(jié)點(diǎn)定位方面,首先分析改進(jìn)了傳統(tǒng)DV-Hop定位算法,然后以核方法為研究手段,結(jié)合主流學(xué)習(xí)算法,提出了兩種新型定位算法;在網(wǎng)絡(luò)覆蓋方面,主要針對(duì)靜態(tài)覆蓋中基于虛擬勢(shì)場(chǎng)的覆蓋算法做了研究,提出了基于靜電場(chǎng)理論的移動(dòng)傳感器網(wǎng)絡(luò)部署算法。本文的主要研究?jī)?nèi)容以及創(chuàng)新點(diǎn)如下:1.提煉出了 DV-Hop算法產(chǎn)生誤差的幾個(gè)原因,在此基礎(chǔ)上提出使用粗定位到精確定位的遞增式算法,并且在三邊測(cè)量法計(jì)算階段引入共線度概念,選擇定位質(zhì)量好的單元進(jìn)行計(jì)算,從而使定位算法得到了較高的定位精度;2.借助移動(dòng)信標(biāo)節(jié)點(diǎn),使其按照規(guī)劃的路徑移動(dòng),從而產(chǎn)生若干個(gè)虛擬信標(biāo)節(jié)點(diǎn),這樣可以有效減少真實(shí)信標(biāo)節(jié)點(diǎn)的數(shù)量。同時(shí)將由這些虛擬信標(biāo)節(jié)點(diǎn)與監(jiān)控區(qū)域未知節(jié)點(diǎn)交流獲得的信號(hào)向量作為直推支持向量機(jī)中訓(xùn)練的有標(biāo)簽數(shù)據(jù)樣本。根據(jù)訓(xùn)練樣本的特點(diǎn),提出一種多類對(duì)多類的分類法,據(jù)此推斷未知節(jié)點(diǎn)的位置。實(shí)驗(yàn)與仿真結(jié)果表明,該方法獲得了較高的定位精度;3.提出了一種基于節(jié)點(diǎn)跳數(shù)和核方法的無(wú)線傳感器網(wǎng)絡(luò)定位算法,該算法的基本思想是利用高斯核函數(shù)度量節(jié)點(diǎn)間的相似性。通過(guò)收集和利用實(shí)際距離和節(jié)點(diǎn)間跳數(shù)信息,將信標(biāo)節(jié)點(diǎn)間的跳數(shù)信息和距離信息作為訓(xùn)練數(shù)據(jù),使用偏最小二乘法學(xué)習(xí)并構(gòu)建其間的最優(yōu)模型,并用此模型預(yù)測(cè)未知節(jié)點(diǎn)到已知節(jié)點(diǎn)距離。實(shí)驗(yàn)結(jié)果表明,本定位算法定位精度較高,受信標(biāo)節(jié)點(diǎn)的數(shù)目影響較小,并且具有將強(qiáng)的環(huán)境適應(yīng)性,適合不同的部署環(huán)境等特點(diǎn);4.提出了一種基于虛擬勢(shì)場(chǎng),分布式、自適應(yīng)、可擴(kuò)展的移動(dòng)傳感器網(wǎng)絡(luò)部署算法,該算法把部署區(qū)域內(nèi)的障礙物、節(jié)點(diǎn)等看作是帶電荷的粒子,粒子受到其它障礙物和粒子的庫(kù)侖力作用而產(chǎn)生運(yùn)動(dòng),最終所有節(jié)點(diǎn)在力的相互作用下自動(dòng)擴(kuò)散到整個(gè)網(wǎng)絡(luò)而完成部署。仿真結(jié)果表明,該算法在各種場(chǎng)景下的性能指標(biāo)表現(xiàn)良好。
[Abstract]:Wireless Sensor Networks (WSNs) is produced and developed rapidly with the development of semiconductor technology, microsystem technology, communication technology and so on. Wireless sensor network is composed of all kinds of integrated micro-sensor nodes, collects and processes the data of sensing objects in the geographical area covered by the network, and processes the data through embedded system. The data is transmitted to the base station through random self-organizing wireless communication network, and finally to the management node through the Internet or satellite network. Wireless sensor networks are widely used in military defense, national security, environmental monitoring and health care. In these applications, node location and network coverage are two hot spots in wireless sensor networks. The location information of nodes plays an important role in the realization of many applications of wireless sensor networks, and network coverage determines the range of services that wireless sensor networks can provide. It also affects the cost of the network and the performance of various specific applications to a great extent. In this paper, the node location and network coverage technology in wireless sensor networks are deeply studied and discussed. In the aspect of node localization, this paper first analyzes and improves the traditional DV-Hop localization algorithm, then takes the kernel method as the research means, combines the mainstream learning algorithm, proposes two new localization algorithms, in the aspect of network coverage, In this paper, the virtual potential field based coverage algorithm in static coverage is studied, and a mobile sensor network deployment algorithm based on electrostatic field theory is proposed. The main contents and innovations of this paper are as follows: 1. In this paper, several reasons for errors in DV-Hop algorithm are extracted. On the basis of this, an incremental algorithm with coarse positioning to accurate location is proposed. The concept of collinearity is introduced into the calculation stage of trilateral measurement method, and the unit with good positioning quality is selected for calculation. Thus, the location algorithm can get a higher positioning accuracy. With the help of mobile beacon nodes, they can be moved according to the planned path, thus generating several virtual beacon nodes, which can effectively reduce the number of real beacon nodes. At the same time, the signal vectors obtained from the communication between these virtual beacon nodes and unknown nodes in the monitoring area are used as the labeled data samples trained in the direct push support vector machine. According to the characteristics of training samples, a multi-class to multi-class classification method is proposed to infer the location of unknown nodes. The experimental and simulation results show that the method has a high positioning accuracy. This paper presents a localization algorithm for wireless sensor networks based on node hops and kernels. The basic idea of the algorithm is to measure the similarity between nodes by using the Gao Si kernel function. By collecting and utilizing the information of actual distance and hops between nodes, the information of hops and distance between beacon nodes is taken as training data, and the optimal model is constructed by using partial least square method. The model is used to predict the distance between unknown nodes and known nodes. The experimental results show that the localization algorithm has the advantages of high accuracy, little influence by the number of beacon nodes, strong adaptability to the environment, and suitable for different deployment environments. This paper presents a deployment algorithm for mobile sensor networks based on virtual potential field, distributed, adaptive and extensible. The algorithm regards obstacles and nodes in the deployment area as charged particles. Particles are moved by other obstacles and particles, and eventually all nodes are automatically diffused to the entire network under the interaction of the forces. Simulation results show that the performance of the algorithm is good in various scenarios.
【學(xué)位授予單位】：南京理工大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2016
【分類號(hào)】：TP212.9;TN929.5

【參考文獻(xiàn)】

相關(guān)期刊論文 前7條

1 石欣;印愛民;張琦;;基于K最近鄰分類的無(wú)線傳感器網(wǎng)絡(luò)定位算法[J];儀器儀表學(xué)報(bào);2014年10期

2 喬鋼柱;李雪蓮;曾建潮;;基于TinyOS平臺(tái)的RSSI定位系統(tǒng)設(shè)計(jì)與實(shí)現(xiàn)[J];電子科技;2013年02期

3 錢志鴻;王義君;;面向物聯(lián)網(wǎng)的無(wú)線傳感器網(wǎng)絡(luò)綜述[J];電子與信息學(xué)報(bào);2013年01期

4 程偉;史浩山;王慶文;;基于差分修正的傳感器網(wǎng)絡(luò)加權(quán)質(zhì)心定位算法[J];系統(tǒng)仿真學(xué)報(bào);2012年02期

5 羅飛;杜明輝;;基于改進(jìn)圓模型的到達(dá)角度和時(shí)間估計(jì)[J];電波科學(xué)學(xué)報(bào);2007年05期

6 吳紹華;張乃通;;基于UWB的無(wú)線傳感器網(wǎng)絡(luò)中的兩步TOA估計(jì)法[J];軟件學(xué)報(bào);2007年05期

7 蔣杰;方力;張鶴穎;竇文華;;無(wú)線傳感器網(wǎng)絡(luò)最小連通覆蓋集問題求解算法[J];軟件學(xué)報(bào);2006年02期

，

本文編號(hào)：2002990