【摘要】：近年來,位置服務產業(yè)發(fā)展迅猛,并在物聯網、智慧城市建設、救災減災等諸多領域有著廣闊的市場空間。當前,全球定位系統(tǒng)(Global Positioning System, GPS)已解決了室外定位問題,而對于GPS無法覆蓋的室內環(huán)境,實現高精度定位依然面臨著極大挑戰(zhàn)。超寬帶(Ultra-Wideband, UWB)技術具有穿透能力強、測距精度高、抗多徑能力強等優(yōu)點,是實現高精度定位的關鍵技術。本文研究了基于超寬帶的定位方法,討論了超寬帶定位技術的研究現狀及仍需解決的問題,針對室內多徑信道提出了兩種改進的時延估計(Time of Arrival,TOA)算法,并給出了仿真分析,從理論上證明了算法的有效性。另外,本論文針對無線傳感網中參考節(jié)點覆蓋率不足的場景,提出了兩種高效的協(xié)作定位方法,所提方法改善定位性能的同時實現了網絡中能量、信息的有效管理。其主要貢獻如下： 首先,本論文針對適于UWB測距的TOA估計算法,分析了現有TOA估計算法的不足以及制約TOA估計精度的關鍵因素,探討了接收信號的特征,提出了基于校驗窗的TOA估計算法和基于誤差特征的TOA估計算法。不同于傳統(tǒng)TOA估計算法,各個能量樣本獨立判決,基于校驗窗的TOA估計算法充分利用了信號樣本在時間維度上的相關性,通過多個能量樣本聯合檢測提高了估計精度,因此該算法可以用于改善其它TOA估計算法,是一種通用的TOA改進技術。基于誤差特征的TOA估計算法則從另外一個角度考慮,分析了閾值和檢測誤差之間的關系,建立了誤差類型函數,通過誤差類型函數自適應調整閾值,與傳統(tǒng)TOA估計算法不同的是,該算法不需要先驗信道狀態(tài)信息(Channel State Information, CSI),采用了基于判決反饋的自適應閾值,有良好的抗干擾性能。 此外,本論文針對網絡定位的實施問題做了進一步討論,重點研究了協(xié)作定位技術。眾所周知,協(xié)作定位很好的解決了參考節(jié)點覆蓋率不足的問題,但當前協(xié)作定位由于復雜度過高、節(jié)點能耗大等問題受到應用限制。本論文將協(xié)作定位分解為定位估計和信息控制兩個子模塊,提出了基于引導滲流控制信息傳播的協(xié)作定位方案,在引導滲流的控制作用下,只有滿足激活條件的節(jié)點可以實施位置估計和位置信息廣播,因此,所提方案實現了能量管理和信息管理。同時,由于所提方案減少了誤差信息的傳播,提高了定位性能。 最后,本論文將引導滲流的思路擴展至貝葉斯網絡定位,提出了基于分層的非參數置信度傳播(Nonparametric Belief Propagation, NBP)定位算法。論文重新構架了基于引導滲流定位問題的圖模型,形成了分層的圖模型,設計了基于分層圖模型的消息傳播準則,給出了相應的用于實現的基于分層的NBP定位算法。仿真結果驗證了所提算法的有效性。傳統(tǒng)的協(xié)作定位算法中,由于各個節(jié)點都廣播位置信息,且消息不受控制的傳播,導致各節(jié)點位置估計收斂速度很慢甚至無法收斂,所提算法通過控制節(jié)點狀態(tài)以及消息傳播,形成了有序的流動狀態(tài),使得消息總是從置信度高的節(jié)點流向置信度低的節(jié)點,因此,有效實現了能量管理,抑制了誤差傳播現象。
[Abstract]:In recent years, the position service industry has developed rapidly, and has broad market space in many fields, such as the Internet of things, intelligent city construction, disaster relief, disaster relief and so on. At present, the Global Positioning System (GPS) has solved the problem of outdoor location, and still faces the extremely high precision positioning for the indoor environment which can not be covered by GPS. Great challenge. Ultra-Wideband (UWB) technology has the advantages of strong penetration ability, high precision and strong anti multipath ability. It is the key technology to realize high precision positioning. This paper studies the localization method based on ultra wideband, discusses the research status of UWB positioning technology and the problems still need to be solved, and proposes the indoor multipath channel. Two improved Time of Arrival (TOA) algorithm and simulation analysis are given, and the effectiveness of the algorithm is proved theoretically. In addition, two efficient cooperative positioning methods are proposed for the scenario of insufficient coverage of reference nodes in wireless sensor networks. The proposed method improves the location performance and realizes the network at the same time. The main contributions are as follows:
First, this paper, aiming at the TOA estimation algorithm suitable for UWB ranging, analyzes the shortcomings of the existing TOA estimation algorithms and the key factors that restrict the accuracy of the TOA estimation. The characteristics of the received signals are discussed, and the TOA estimation algorithm based on the checkwindow and the TOA estimation method based on the error characteristics are proposed. Different from the traditional TOA estimation algorithms, each energy sample is different. In this independent decision, the TOA estimation algorithm based on the checkwindow makes full use of the correlation of the signal samples in the time dimension and improves the estimation accuracy through joint detection of multiple energy samples. Therefore, the algorithm can be used to improve the other TOA estimation algorithms and is a general TOA improvement technique. The TOA estimation algorithm based on the error features is from the other. On the other hand, the relationship between threshold and detection error is analyzed, and an error type function is established. The error type function adaptively adjusts the threshold value. Unlike the traditional TOA estimation algorithm, the algorithm does not need a priori channel state information (Channel State Information, CSI), and the adaptive threshold based on decision feedback is adopted. It has good anti-jamming performance.
In addition, this paper further discusses the implementation of network positioning, focusing on Collaborative positioning technology. It is well known that collaborative positioning is a good solution to the problem of insufficient coverage of reference nodes. However, the current collaborative positioning is restricted by the application because of high complexity and large energy consumption. It is solved as two sub modules of location estimation and information control. A collaborative positioning scheme based on guided percolation control information transmission is proposed. Under the control of guided percolation, only the nodes that satisfy the activation condition can implement position estimation and position information broadcasting. Therefore, the proposed scheme realizes energy management and information management. The proposed scheme reduces the propagation of error information and improves the positioning performance.
Finally, this paper extends the thought of percolation to Bayesian network positioning, and proposes a hierarchical non parametric Nonparametric Belief Propagation (NBP) location algorithm. The paper reconstructs a graph model based on the guidance percolation location problem, forms a hierarchical graph model, and designs a hierarchical graph model. The simulation results verify the effectiveness of the proposed algorithm. In the traditional cooperative location algorithm, because the location information of each node is broadcast and the message is not transmitted by the control, the convergence rate of the location estimation of each node is slow or even convergent, which is proposed in the traditional cooperative positioning algorithm. By controlling the state of the node and the message propagation, the algorithm forms an orderly flow state, which makes the message always flow from the high confidence node to the low confidence node. Therefore, the energy management is effectively realized and the error propagation is suppressed.
【學位授予單位】：北京郵電大學
【學位級別】：碩士
【學位授予年份】：2015
【分類號】：TN925

【共引文獻】

相關博士學位論文 前1條

1 曹曉倩;面向病態(tài)場景圖像對的立體匹配算法研究[D];中國科學院研究生院（西安光學精密機械研究所）;2014年

相關碩士學位論文 前2條

1 陳明;基于混合部件模型的姿態(tài)估計方法研究[D];南京航空航天大學;2014年

2 李彬;無線網絡中的分布式定位算法研究[D];北京理工大學;2015年

，

本文編號：2124241