X. Peng, Yiping Duan, Bingrui Geng, Xiwen Liu, Xiaoming Tao, N. Ge
{"title":"基于qos的终端视频质量报警模型","authors":"X. Peng, Yiping Duan, Bingrui Geng, Xiwen Liu, Xiaoming Tao, N. Ge","doi":"10.1109/GlobalSIP45357.2019.8969366","DOIUrl":null,"url":null,"abstract":"With the advent of the 5G era, quality of experience (QoE) has become one of the most important indicators to measure mobile video quality. If QoE degradation events can be accurately predicted, it will be beneficial for operators to provide better services. However, many factors could affect QoE, including users’ subjective factors, network and users’ terminal parameters. This paper proposes a model for QoE degradation alarm in the mobile video stream. Specifically, a large-scale dataset is collected in the practical environment by a self-developed application. The dataset is cleaned to reduce the influence of redundancy and outliers firstly. Then, a QoE degradation alarm model is proposed based on data-driven approaches. The random forest is chosen as the core of this model which integrates the advantages of several classification tree and can deal with the high-dimensional data. In order to have a better prediction performance, Synthetic Minority Oversampling Technique (SMOTE) is used for solving the imbalance problem of the train set. As is shown by the experimental results, this alarm model can strike a balance between precision and false alarm rate and performance outperforms the state-of-the-art methods. This result means the proposed model can save the operational cost while increasing users’ satisfaction with the mobile video quality.","PeriodicalId":221378,"journal":{"name":"2019 IEEE Global Conference on Signal and Information Processing (GlobalSIP)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":"{\"title\":\"A QoE-Based Alarm Model for Terminal Video Quality\",\"authors\":\"X. Peng, Yiping Duan, Bingrui Geng, Xiwen Liu, Xiaoming Tao, N. Ge\",\"doi\":\"10.1109/GlobalSIP45357.2019.8969366\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"With the advent of the 5G era, quality of experience (QoE) has become one of the most important indicators to measure mobile video quality. If QoE degradation events can be accurately predicted, it will be beneficial for operators to provide better services. However, many factors could affect QoE, including users’ subjective factors, network and users’ terminal parameters. This paper proposes a model for QoE degradation alarm in the mobile video stream. Specifically, a large-scale dataset is collected in the practical environment by a self-developed application. The dataset is cleaned to reduce the influence of redundancy and outliers firstly. Then, a QoE degradation alarm model is proposed based on data-driven approaches. The random forest is chosen as the core of this model which integrates the advantages of several classification tree and can deal with the high-dimensional data. In order to have a better prediction performance, Synthetic Minority Oversampling Technique (SMOTE) is used for solving the imbalance problem of the train set. As is shown by the experimental results, this alarm model can strike a balance between precision and false alarm rate and performance outperforms the state-of-the-art methods. This result means the proposed model can save the operational cost while increasing users’ satisfaction with the mobile video quality.\",\"PeriodicalId\":221378,\"journal\":{\"name\":\"2019 IEEE Global Conference on Signal and Information Processing (GlobalSIP)\",\"volume\":\"31 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2019-11-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"2\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2019 IEEE Global Conference on Signal and Information Processing (GlobalSIP)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/GlobalSIP45357.2019.8969366\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 IEEE Global Conference on Signal and Information Processing (GlobalSIP)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/GlobalSIP45357.2019.8969366","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
摘要
随着5G时代的到来,体验质量(quality of experience, QoE)已成为衡量移动视频质量的最重要指标之一。如果能准确预测QoE退化事件,将有利于运营商提供更好的服务。然而,影响QoE的因素很多,包括用户主观因素、网络和用户终端参数。提出了一种移动视频流中QoE退化报警模型。具体而言,通过自行开发的应用程序在实际环境中收集大规模数据集。首先对数据集进行清理,减少冗余和异常值的影响。然后,提出了基于数据驱动方法的QoE退化报警模型。该模型以随机森林为核心,综合了几种分类树的优点,能够处理高维数据。为了获得更好的预测效果,采用合成少数派过采样技术(SMOTE)来解决训练集的不平衡问题。实验结果表明,该报警模型能够在准确率和虚警率之间取得平衡,性能优于现有方法。该结果表明,该模型在节省运营成本的同时,提高了用户对移动视频质量的满意度。
A QoE-Based Alarm Model for Terminal Video Quality
With the advent of the 5G era, quality of experience (QoE) has become one of the most important indicators to measure mobile video quality. If QoE degradation events can be accurately predicted, it will be beneficial for operators to provide better services. However, many factors could affect QoE, including users’ subjective factors, network and users’ terminal parameters. This paper proposes a model for QoE degradation alarm in the mobile video stream. Specifically, a large-scale dataset is collected in the practical environment by a self-developed application. The dataset is cleaned to reduce the influence of redundancy and outliers firstly. Then, a QoE degradation alarm model is proposed based on data-driven approaches. The random forest is chosen as the core of this model which integrates the advantages of several classification tree and can deal with the high-dimensional data. In order to have a better prediction performance, Synthetic Minority Oversampling Technique (SMOTE) is used for solving the imbalance problem of the train set. As is shown by the experimental results, this alarm model can strike a balance between precision and false alarm rate and performance outperforms the state-of-the-art methods. This result means the proposed model can save the operational cost while increasing users’ satisfaction with the mobile video quality.