Comparative study of metrics for IPTV transport in the access network

Abstract: Last few years, services such as Internet Protocol Television (IPTV) and Voice over Internet Protocol (ValP) have been in focus. Network operators see a huge potential in streaming real-time media over IP-networks and this lays the ground for so called paid services such as Video on Demand (VOD) as well as free services like YouThbe and SVT Play. IPTV delivers content over a closed infrastructure that receives media from a particular provider using a set-top-box, or STB for short, in the home of customers. Services such as Internet TV (SVT Play, TV3 Play and YouTube for example), have the same look and feel as IPTV but are delivered over the open infrastructure of the internet, relying on the best-effort channel. The latter is called over-the-top or OTT , referring to the media going around the STB rather than through it. The benefit of such a service can be put into one word, flexibility, watch what you want, when you want and where you want. But these type of services have same drawbacks. Problems such as packet loss, delay and jitter in the network greatly affects the quality for the end-user. A significant problem for OTT-traffic in the access network, when dealing with DSL copper transmission, are the old telephone cables used for the last mile to customers, this part of the network was not built for this kind of traffic and suffers from external disturbances in form of 50Hz power lines, cross talk and radio frequency interferences. This paper aims to evaluate and compare different methods for classifying network traffic and how Quality-of-Service (QoS) parameters (packet loss, delay, jitter, out-of-sequence) on the lP level affect Quality-of-Experience (QoE) from a head-end to a receiver in a home network connected to an access network. A simulated IPTV transmission is done in a simulated network with and without an externai disturbance signal injected to a pair in the same cable. . "; i