针对5G网络优化工作中出现的新空口承载语音(Voice over New Radio,VoNR)用户感知差的问题,探讨平均意见评分(Mean Opinion Score,MOS)值影响因素与分析流程,对后续出现的此类问题提供分析思路及具体解决方案。通过对网络现有参数和物...针对5G网络优化工作中出现的新空口承载语音(Voice over New Radio,VoNR)用户感知差的问题,探讨平均意见评分(Mean Opinion Score,MOS)值影响因素与分析流程,对后续出现的此类问题提供分析思路及具体解决方案。通过对网络现有参数和物理参数的核查,修改网络异常参数,以射频优化为基础、以参数优化为辅助,达到快速提升5G VoNR语音用户使用感知的目的。通过采用相应的网络优化手段,有效提升5G网络VoNR接通率,降低5G网络VoNR掉话率。展开更多
Quality of experience ( QoE ) based scheduling algorithm of long term evalution ( LTE ) network with various traffics is studied. Utility functions are adopted to estimate mean opinion score (MOS) for different ...Quality of experience ( QoE ) based scheduling algorithm of long term evalution ( LTE ) network with various traffics is studied. Utility functions are adopted to estimate mean opinion score (MOS) for different traffics and a new MOS metric called normalized MOS is defined. A scheduling algorithm based on normalized MOS and greedy algorithm is proposed, aiming at maximizing the entirety MOS level of the whole users in the cell. We compare the performance of the proposed algorithm with other typical scheduling algorithms and the simulation results show that the algorithm pro- posed outperform other ones in term of QoE and fairness.展开更多
会话类流存在实时性高、时延抖动低、吞吐量大的特点,为了更好地对会话类混合流进行分类和提高用户的QoE(Quality of Experience),文中从网络实时会话流入手,分析混合流的QoS(Quality of Service)并建立与QoE的模型关系,从而得出不同Qo...会话类流存在实时性高、时延抖动低、吞吐量大的特点,为了更好地对会话类混合流进行分类和提高用户的QoE(Quality of Experience),文中从网络实时会话流入手,分析混合流的QoS(Quality of Service)并建立与QoE的模型关系,从而得出不同QoS特征下的QoE的概率分布,并把其中最明显的特征概率分布以网络流特征的形式加入到原有流分类特征集当中,然后对QQ视频、2D游戏、3D游戏、Skype语音四种会话混合流进行分类。通过C4.5决策树机器学习算法进行分类。实验结果表明:在流分类的正确率、召回率和F-测度上,使用QoE概率分布特征的分类算法相比于现有方案,性能均有显著提高。展开更多
文摘提出了一种基于加权巴克谱失真(W BSD)的语音质量客观评价算法。鉴于不同巴克带谱失真对语音质量的影响程度不同,该算法提出在巴克域中对不同巴克带谱失真进行加权来求失真语音的客观失真测度。并且通过分析改进型巴克谱失真测度(M BSD)中15个巴克带失真与平均意见分(M O S)间的相关度,提出了一组以相关系数幂次方为权重的有效巴克谱失真权重矢量。实验结果表明:与M BSD算法相比,W BSD在算法复杂度没有明显增加的情况下与M O S的相关度提高了3%。
文摘针对5G网络优化工作中出现的新空口承载语音(Voice over New Radio,VoNR)用户感知差的问题,探讨平均意见评分(Mean Opinion Score,MOS)值影响因素与分析流程,对后续出现的此类问题提供分析思路及具体解决方案。通过对网络现有参数和物理参数的核查,修改网络异常参数,以射频优化为基础、以参数优化为辅助,达到快速提升5G VoNR语音用户使用感知的目的。通过采用相应的网络优化手段,有效提升5G网络VoNR接通率,降低5G网络VoNR掉话率。
基金Supported by China National S&T Major Project(2013ZX03003002-003)Beijing Natural Science Foundation(4152047)National High Technology Research and Development Program of China(863Program)(2014AA01A701)
文摘Quality of experience ( QoE ) based scheduling algorithm of long term evalution ( LTE ) network with various traffics is studied. Utility functions are adopted to estimate mean opinion score (MOS) for different traffics and a new MOS metric called normalized MOS is defined. A scheduling algorithm based on normalized MOS and greedy algorithm is proposed, aiming at maximizing the entirety MOS level of the whole users in the cell. We compare the performance of the proposed algorithm with other typical scheduling algorithms and the simulation results show that the algorithm pro- posed outperform other ones in term of QoE and fairness.
文摘会话类流存在实时性高、时延抖动低、吞吐量大的特点,为了更好地对会话类混合流进行分类和提高用户的QoE(Quality of Experience),文中从网络实时会话流入手,分析混合流的QoS(Quality of Service)并建立与QoE的模型关系,从而得出不同QoS特征下的QoE的概率分布,并把其中最明显的特征概率分布以网络流特征的形式加入到原有流分类特征集当中,然后对QQ视频、2D游戏、3D游戏、Skype语音四种会话混合流进行分类。通过C4.5决策树机器学习算法进行分类。实验结果表明:在流分类的正确率、召回率和F-测度上,使用QoE概率分布特征的分类算法相比于现有方案,性能均有显著提高。