發(fā)布時(shí)間：2018-10-20 09:24

【摘要】：隨著無線通信技術(shù)的高速發(fā)展,無線視頻流業(yè)務(wù)應(yīng)用越來越廣泛,人們對(duì)無線視頻流的服務(wù)質(zhì)量期望也逐漸提高。為了獲得用戶對(duì)視頻流服務(wù)的認(rèn)可,服務(wù)提供商迫切需要建立一種以用戶認(rèn)可程度為標(biāo)準(zhǔn)的質(zhì)量評(píng)價(jià)體系。傳統(tǒng)的服務(wù)質(zhì)量(Quality of Service, QoS)是一種被廣泛采用的服務(wù)度量方法,但是QoS強(qiáng)調(diào)技術(shù)層面的客觀評(píng)價(jià)指標(biāo),不能直接體現(xiàn)用戶對(duì)視頻流質(zhì)量的真實(shí)感受。用戶體驗(yàn)質(zhì)量(Quality of Experience, QoE)是一種以用戶認(rèn)可程度為標(biāo)準(zhǔn)的服務(wù)度量方法,它以用戶對(duì)業(yè)務(wù)的使用感受為研究重點(diǎn),能評(píng)價(jià)業(yè)務(wù)中多種QOS指標(biāo)對(duì)用戶體驗(yàn)的影響。QoE直接反映了用戶對(duì)服務(wù)的認(rèn)可程度,是決定無線視頻流業(yè)務(wù)能否取得成功的關(guān)鍵因素。因此,用戶體驗(yàn)質(zhì)量不僅是學(xué)術(shù)界的重點(diǎn)研究課題,也是工業(yè)界實(shí)現(xiàn)業(yè)務(wù)發(fā)展關(guān)注的焦點(diǎn)。為了保證無線網(wǎng)絡(luò)視頻流業(yè)務(wù)的用戶體驗(yàn)質(zhì)量,對(duì)無線視頻流業(yè)務(wù)進(jìn)行質(zhì)量評(píng)估及其應(yīng)用優(yōu)化有重要的研究意義和實(shí)用價(jià)值。本文圍繞無線視頻流業(yè)務(wù)的QoE,研究無線網(wǎng)絡(luò)中視頻流業(yè)務(wù)的用戶體驗(yàn)質(zhì)量估計(jì)模型及其應(yīng)用。 在無線視頻流業(yè)務(wù)用戶體驗(yàn)質(zhì)量估計(jì)方面,針對(duì)現(xiàn)有QoE估計(jì)方法存在的評(píng)估指標(biāo)不全面、評(píng)估準(zhǔn)確度不理想等問題,本文提出了一種基于徑向基函數(shù)神經(jīng)網(wǎng)絡(luò)(Radial Basis Function Neural Networks, RBFN)的無參考質(zhì)量評(píng)估模型。首先,我們分析了端到端跨層參數(shù)對(duì)視頻流業(yè)務(wù)用戶體驗(yàn)質(zhì)量產(chǎn)生的影響。然后,在無需原視頻作比較的前提下,建立了基于RBFN的QoE估計(jì)模型,詳細(xì)闡述了基于RBFN的QoE估計(jì)模型的評(píng)估原理與流程。最后,我們仿真驗(yàn)證所提出的QoE估計(jì)模型,并與其它四種典型的無參考估計(jì)模型進(jìn)行比較分析,結(jié)果表明我們所提出的基于RBFN的QoE估計(jì)模型不僅評(píng)估準(zhǔn)確度最高,而且具有低的時(shí)間復(fù)雜度。 針對(duì)QoE估計(jì)模型在無線視頻流業(yè)務(wù)優(yōu)化的應(yīng)用方面,本文提出了一種基于QoE的視頻流業(yè)務(wù)傳輸控制優(yōu)化機(jī)制,該優(yōu)化機(jī)制聯(lián)合了丟包率與端到端單向時(shí)延增減趨勢信息對(duì)網(wǎng)絡(luò)狀態(tài)進(jìn)行細(xì)分并判斷網(wǎng)絡(luò)擁塞程度,視頻流業(yè)務(wù)發(fā)送端根據(jù)監(jiān)測的網(wǎng)絡(luò)狀態(tài)與由基于RBFN的QoE估計(jì)模型計(jì)算的用戶體驗(yàn)質(zhì)量,采取相應(yīng)的策略動(dòng)態(tài)地調(diào)整發(fā)送端的傳輸速率,即編碼比特率,以達(dá)到提升用戶體驗(yàn)質(zhì)量的目的。實(shí)驗(yàn)結(jié)果表明,我們提出的基于RBFN的QoE估計(jì)模型的業(yè)務(wù)傳輸控制策略在網(wǎng)絡(luò)狀態(tài)波動(dòng)的情況下,能夠?qū)崿F(xiàn)視頻流業(yè)務(wù)QoE的提升,并且具有良好的視頻播放穩(wěn)定性。
[Abstract]:With the rapid development of wireless communication technology, wireless video streaming services are more and more widely used. In order to obtain users' recognition of video streaming service, service providers urgently need to establish a quality evaluation system based on the degree of user acceptance. Traditional quality of service (Quality of Service, QoS) is a widely used method of service measurement, but QoS emphasizes the objective evaluation index at the technical level, which can not directly reflect the users' true feeling about the quality of video stream. User experience quality (Quality of Experience, QoE) is a service measurement method based on the degree of user acceptance. It can evaluate the influence of various QOS indexes on the user experience. QoE directly reflects the user's approval of the service and is the key factor to determine the success of wireless video streaming service. Therefore, the quality of user experience is not only the focus of academic research, but also the focus of business development in industry. In order to guarantee the user experience quality of wireless network video stream service, it is of great significance and practical value to evaluate the quality of wireless video stream service and its application optimization. This paper focuses on the QoE, of wireless video streaming services; the user experience quality estimation model of video streaming services in wireless networks and its application. In the aspect of user experience quality estimation of wireless video stream service, the existing QoE estimation methods have some problems, such as the evaluation index is not comprehensive, the evaluation accuracy is not ideal, and so on. In this paper, a non-reference quality evaluation model based on radial basis function neural network (Radial Basis Function Neural Networks, RBFN) is proposed. Firstly, we analyze the effect of end-to-end cross layer parameters on the quality of video stream service user experience. Then, the QoE estimation model based on RBFN is established, and the evaluation principle and flow of QoE estimation model based on RBFN are described in detail. Finally, we simulate and verify the proposed QoE estimation model and compare it with the other four typical non-reference estimation models. The results show that the proposed QoE estimation model based on RBFN is not only the most accurate. And it has low time complexity. Aiming at the application of QoE estimation model in wireless video stream traffic optimization, this paper proposes an optimization mechanism of video stream traffic transmission control based on QoE. The optimization mechanism combines packet loss rate and end-to-end one-way delay trend information to subdivide the network state and judge the degree of network congestion. According to the monitored network status and the user experience quality calculated by the QoE estimation model based on RBFN, the video stream service sender dynamically adjusts the transmission rate of the sender, that is, the coded bit rate, by adopting the corresponding strategy. In order to improve the quality of user experience. The experimental results show that the proposed service transmission control strategy based on RBFN QoE estimation model can achieve the enhancement of video stream service QoE under the condition of network state fluctuation and has good video playback stability.
【學(xué)位授予單位】：浙江大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：TN919.8

【參考文獻(xiàn)】

相關(guān)期刊論文 前2條

1 張瑞;張泉;;TD-LTE技術(shù)在互聯(lián)網(wǎng)中的應(yīng)用研究[J];電腦知識(shí)與技術(shù);2012年08期

2 邱錦波;朱光喜;;一種無線視頻傳輸?shù)目鐚幼赃m應(yīng)不平等保護(hù)方法[J];電子與信息學(xué)報(bào);2007年01期

，

本文編號(hào)：2282718