本文關(guān)鍵詞： 分布式融合 Kalman濾波 功耗分配 信道估計誤差 估計融合　出處：《西安電子科技大學(xué)》2014年碩士論文　論文類型：學(xué)位論文

【摘要】：無線傳感器網(wǎng)絡(luò)(Wireless Sensor Networks, WSNs)作為一門交叉學(xué)科,結(jié)合了傳感器技術(shù)、無線通信以及計算機網(wǎng)絡(luò)等技術(shù),近年來引起了廣泛關(guān)注,成為了國內(nèi)外的研究熱點,也越來越廣泛地應(yīng)用到眾多軍事和非軍事領(lǐng)域,如定位跟蹤、遠程遙控、軍事偵察等。但是由于傳感器自身的一些缺陷,比如能量、帶寬以及數(shù)據(jù)處理能力等,使得數(shù)據(jù)傳輸效率以及網(wǎng)絡(luò)生命周期等受到了限制。為了解決通信帶寬以及能量受限,提高信息收集的效率,數(shù)據(jù)融合(Data Fusion)技術(shù)起到了非常大的作用。但是,把數(shù)據(jù)融合應(yīng)用到實際中,仍然存在很多挑戰(zhàn)：傳感器系統(tǒng)的功耗分配問題,信道信息的不確定性等,本文針對目前存在的這些問題展開研究,利用壓縮策略,使得系統(tǒng)能在降低傳輸量的同時達到集中式融合的估計性能；研究了在存在同步誤差情況下,系統(tǒng)的最優(yōu)功耗分配,以及信道估計誤差對系統(tǒng)性能的影響。具體的工作包括以下內(nèi)容：1.針對傳感器之間量測噪聲相關(guān)的情況,我們提出了一種基于線性變換的分布式融合算法。與傳統(tǒng)的噪聲之間互不相關(guān)的假設(shè)相比,相關(guān)噪聲的假設(shè)更符合實際情況。我們針對線性融合系統(tǒng)提出了一種基于傳感器節(jié)點量測值的線性變換的無損準則,并給出了能夠達到集中式融合性能的必要條件。另外,我們也給出了當量測矩陣為行滿秩和列滿秩時的最優(yōu)壓縮矩陣；2.針對無線傳感器網(wǎng)絡(luò)中傳感器節(jié)點能量受限這一特點,我們分別針對標量信源和向量信源兩種情況,研究了網(wǎng)絡(luò)節(jié)點的功耗分配策略�；趥鞲衅骱腿诤现行拇嬖谕秸`差這一現(xiàn)象,我們提出了對應(yīng)的功耗分配策略,并得到了最優(yōu)分配系數(shù)的閉式解。同時,我們也做了大量的仿真來驗證該算法的性能,結(jié)果表明該算法優(yōu)于平均能量分配的算法；3.現(xiàn)有的大部分關(guān)于數(shù)據(jù)融合的算法都是在基于信道狀態(tài)信息在融合中心處完全已知的假設(shè)下進行研究的,但是在實際應(yīng)用中這一假設(shè)很難達到,估計過程就意味著信道狀態(tài)信息的不確定。針對這一現(xiàn)象,我們重點研究信道估計誤差對估計融合的影響,以及訓(xùn)練數(shù)據(jù)和傳輸數(shù)據(jù)之間的能量最優(yōu)分配,并對線性最小均方誤差(Linea minimum mean square error, LMMSE)和最優(yōu)的加權(quán)最小二乘(Optimal Weighted Least squares, OWLS)兩種融合算法進行了大量的分析和仿真。
[Abstract]:Wireless Sensor Networks (WSNs) as an interdisciplinary subject, which combines sensor technology, wireless communication and computer network, has attracted wide attention in recent years and has become a research hotspot at home and abroad. It has also been widely used in many military and non-military fields, such as positioning and tracking, remote control, military reconnaissance, etc. However, due to some shortcomings of the sensor itself, such as energy, bandwidth and data processing capability, In order to solve the problem of bandwidth and energy constraints and improve the efficiency of information collection, data fusion data fusion technology plays a very important role. There are still many challenges in applying data fusion to practice: the power allocation problem of sensor system, the uncertainty of channel information and so on. It makes the system achieve the estimation performance of centralized fusion while reducing the amount of transmission, and studies the optimal power allocation of the system under the condition of synchronization error. And the influence of channel estimation error on the performance of the system. The specific work includes the following contents: 1. For the measurement of noise between sensors, We propose a distributed fusion algorithm based on linear transformation. The assumption of correlated noise is more in line with the actual situation. We propose a lossless criterion of linear transformation based on sensor node measurements for linear fusion systems, and give the necessary conditions to achieve centralized fusion performance. We also give the optimal compression matrix when the equivalent measurement matrix is row full rank and column full rank. In view of the energy limitation of sensor nodes in wireless sensor networks, we consider the scalar source and vector source, respectively. In this paper, the power allocation strategy of network nodes is studied. Based on the synchronization error between sensor and fusion center, we propose a corresponding power allocation strategy and obtain the closed solution of the optimal allocation coefficient. We also do a lot of simulations to verify the performance of the algorithm. The results show that this algorithm is superior to the average energy allocation algorithm. 3. Most of the existing algorithms for data fusion are based on the assumption that the channel state information is completely known at the fusion center. However, this assumption is difficult to achieve in practical application, and the estimation process means that the channel state information is uncertain. In view of this phenomenon, we focus on the influence of channel estimation error on estimation fusion. And the optimal allocation of energy between the training data and the transmitted data. Two fusion algorithms, Linea minimum mean square error (LMMSE) and optimal weighted least squares optimal Weighted Least squares (owl LSs), are analyzed and simulated.
【學(xué)位授予單位】：西安電子科技大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：TP212.9;TN929.5

【相似文獻】

相關(guān)期刊論文 前10條

1 董霄劍,蔣良成,趙夙,尤肖虎;信道估計誤差與估計長度關(guān)系的頻域分析[J];電子學(xué)報;2000年S1期

2 王中鵬;張少中;;平坦信道下的MIMO信道估計誤差分析[J];電訊技術(shù);2008年02期

3 曾二林;朱世華;廖學(xué)文;;存在信道估計誤差的有限反饋多用戶多輸入多輸出傳輸性能分析[J];電子與信息學(xué)報;2009年02期

4 黃森華;邱玲;朱近康;;信道估計誤差對下行多用戶迫零波束成型系統(tǒng)吞吐量的影響[J];應(yīng)用科學(xué)學(xué)報;2006年05期

5 戰(zhàn)金龍;劉宏清;廖桂生;;V-BLAST OFDM系統(tǒng)中一種穩(wěn)健的檢測算法[J];西安電子科技大學(xué)學(xué)報;2007年05期

6 郭明喜;沈越泓;;非理想信道條件下的V-BLAST系統(tǒng)檢測算法[J];數(shù)據(jù)采集與處理;2011年03期

7 戰(zhàn)金龍;廖桂生;劉宏清;;V-BLAST系統(tǒng)中一種新的穩(wěn)健的檢測算法[J];系統(tǒng)工程與電子技術(shù);2007年05期

8 楊U，

本文編號：1525801