Lehrstuhl für Informatik III

    QoE Monitoring with VNFs (Virtual Network Functions) in the Cloud


    QoE Monitoring VNF is now available at https://go.uniwue.de/qoevnf. The VNF QoE monitoring is a plain software that exploits a C++ library, namely libtins to capture the un-encrypted video flows at network interface. The captured packets are then parsed to feed all necessary information for the video buffer estimation algorithm, such as IP address, TCP header and the payload of application layer protocols.

      In order to consider Quality of Experience (QoE) for orchestration and consolidation in a cloud management architecture, proper QoE monitoring functions are required. They can be applied either in long term or short term management components and provide the platform with information about the service quality perceived by the end users.

      The component is able to monitor all the video flows passing through from the streaming server to the client. We use a proxy program to route video traffic to the VNF. The video traffic from the streaming server is captured by our component. It is a software which utilizes a Python or C++ library, namely Scapy and libpcap. These open source software libraries provide real time packet sniffing and decoding. To analyze the video flows, we designed an algorithm, which eventually provides us with an estimated video buffer, stalling frequency and length based on download timestamps of video segments. The QoE for video streaming is calculated based on a reference QoE model. To validate the estimated values, we simultaneously sample the video buffer, stalling frequency and length at the client browser by using a Javascript-based web API. The discrepancy between the estimated and actual values shows the level of accuracy of the monitoring. More details of the estimation algorithm can be found in the publications.

    Dinh-Xuan, L., Seufert, M., Wamser, F., Tran-Gia, P.: Study on the Accuracy of QoE Monitoring for HTTP Adaptive Video Streaming Using VNF. 1st IFIP/IEEE International Workshop on Quality of Experience Management (QoE-Management). , Lisbon, Portugal (2017).


    Von Lam Dinh-Xuan