{"title":"Quality of Experience estimation using frame loss pattern and video encoding characteristics in DVB-H networks","authors":"K. Singh, G. Rubino","doi":"10.1109/PV.2010.5706832","DOIUrl":null,"url":null,"abstract":"Automatic estimation of Quality of Experience (QoE) is of key importance for mobile television networks such as DVB-H. These networks can install network probes in order to monitor QoE. The QoE feedback can be used to take some corrective measures, in case the quality drops, to bring back QoE to satisfactory level. In this paper, we extend a previously proposed noreference QoE monitoring module for H.264 video over DVB-H networks. We consider an additional parameter called quantisation parameter (QP) and consider frame loss pattern, instead of packet loss pattern, apart from the parameters used in the earlier work such as motion activity and loss rank in a Group of Pictures (GOP). The earlier work is restricted to a fixed encoding bitrate. By considering QP this restriction is removed because QP determines the bitrate as well as the resulting video quality. The results show that our estimation module based on Random Neural Networks (RNN) captures the non-linear relationship between these parameters and QoE. Moreover, our consideration of additional parameters leads to significant improvement in QoE estimation accuracy.","PeriodicalId":339319,"journal":{"name":"2010 18th International Packet Video Workshop","volume":"44 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2010-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"9","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2010 18th International Packet Video Workshop","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/PV.2010.5706832","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 9
Abstract
Automatic estimation of Quality of Experience (QoE) is of key importance for mobile television networks such as DVB-H. These networks can install network probes in order to monitor QoE. The QoE feedback can be used to take some corrective measures, in case the quality drops, to bring back QoE to satisfactory level. In this paper, we extend a previously proposed noreference QoE monitoring module for H.264 video over DVB-H networks. We consider an additional parameter called quantisation parameter (QP) and consider frame loss pattern, instead of packet loss pattern, apart from the parameters used in the earlier work such as motion activity and loss rank in a Group of Pictures (GOP). The earlier work is restricted to a fixed encoding bitrate. By considering QP this restriction is removed because QP determines the bitrate as well as the resulting video quality. The results show that our estimation module based on Random Neural Networks (RNN) captures the non-linear relationship between these parameters and QoE. Moreover, our consideration of additional parameters leads to significant improvement in QoE estimation accuracy.