On Some Live IPTV QoS Key Techniques

Live IPTV QoS is the non-functional attributes of live IPTV service, e.g. channel start-up time, channel change time, play-out progress synchronization of different terminals. QoS is critical to users who experience IPTV services. In complex Internet environment, a large amount of users might in different local networks. How to utilize resource in local content distribute network and from other neighbor terminals, and how to provide nessesary QoS assurance to provide a fluent, fast respond and interactive environment, becomes a crucial problem in live IPTV QoS optimization research.According to analysis of current live IPTV QoS optimization methods, we address that there are lots of problems have not been resolved or even concerned. Especially QoS metrics of different content distribution network and enhancement of IPTV service interactivity. Based on current status live IPTV QoS optimization methods, we address several critical problems. The contribution of this dissertation includes:1. Propose a fast channel start-up and channel change approach based on IP multicast environment. We propose a fast channel start-up and channel change approach in IP multicast environment combined with P2P method. Based on the principles of highly utilization of upload bandwidth of peers in P2P networks, terminals can request other peers whtching the same target channel to send terminals short-time and fast speed data, to fulfill play-out buffer, shorten channel start-up and change time and improve channel start-up and change QoS metrics. This method can keep low utilization rate of back-bone bandwidth, effectively reduce channel start-up and channel change time. New data generated by this approach is mainly used to provide multimedia programmes for users, effectively reduce users wait time.2. Proposed the issues of nodes'playback progress synchronization which nodes locate in small communities,and proposed an QoS optimization method to improve playback progress synchronization. For services close related with live IPTV applications, the playback synchronization of nodes that play a key problem that must be addressed. This paper presents a playback synchronization optimization method. This method mainly includes the establishment of synchronized small communities at the preparation phase, the establishment of small communities algorithms, and the adaptation algorithm when new nodes join the system by adding a small community. This algorithm can detect neighbor nodes information, select the appropriate parent or adjust the application layer structure topology. The nodes playback synchronization in small communities can be enhanced and maintained at less within the scope.3. Proposed an optimized method of nodes playback synchronization in large-scale networks. In order to solve large-scale P2P network nodes playback synchronization problem and to improve interaction between end-users and ISPs, we use protential-based algorithm inspired by electronical field theory.We first proposed large-scale nodes in the P2P network playback synchronization issues. Then to use electronical potential concepts in field theory, the issues involved on this mathematical model. We proposed the cocept of nodes potential in live IPTV services. In the distributed environment, differences between nodes potential is calculated. We proposed an application topology establishment and optimization algorithm, and a bandwidth distribution algorithm. These method can improve nodes playback synchronization, real-time and nodes hops.4. Proposed a node architecture supporting live IPTV QoS optimization in Internet environment. This architecture can utilize topology structure in different distribution networks, to provide diversity QoS assurance mechanism in distributed environment for providing live IPTV services. Nodes in the system have equal positions, interact with each other according to protocols. Nodes could exist in different distribution networks simultaneously. According to cooperation between nodes in data exchange operation, help IPTV node to find more nodes which have higher QoS metrics. Share playout buffer in several overlay networks, to satisfy complicated QoS requirements.