Adaptive service quality ensuring method in multi-playing environment

Abstract

Combining initiating mechanism of sending party with initiating mechanism of receiving party, the invented method possesses advantages of both parties. Based on network state, initiating mechanism of receiving party adjusts quality of video received by receiver in real time, meanwhile sends a request for adjusting QoS grade to sending party at proper time through control layer to start up initiating mechanism of sending party. With request for adjusting being received, the sending party analyzes and adjusts integral video quality through sampling and negotiating to control, optimize video quality as well as raise network availability, preventing feed back storm and maintaining records of information and state for group members. In the invention, number required can be adjusted by terminal ends without need of intermediate node and router so as to suit current network and obtain best service quality.

Description

technical field [0001] The present invention relates to the technical field of network protocol communication, in particular to an adaptive QoS (Quality-of-Service) guarantee method in a multicast environment, especially a mechanism initiated by a sender and a mechanism initiated by a receiver. Combined with the adaptive QoS guarantee method, the receiver initiates the mechanism to adjust the video quality received by the receiver in real time according to the network conditions, and at the same time activates the sender to initiate the mechanism at an appropriate time, and the sender uses sampling negotiation to dynamically adjust the overall video quality , not only can control and optimize video quality and network utilization as a whole, but also isolate the part affected by network congestion in the multicast group, and provide different video quality for different users. Background technique [0002] There are two ways of traditional IP (Internet Protocol) communicatio...

AI Technical Summary

Problems solved by technology

However, this mechanism also has its own disadvantages

First of all, this mechanism has high requirements for encoding, and the adjustment of the receiving quality by each receiver often affects each other, which is difficult to coordinate

In the layered multicast mode, the adjustment granularity of layered encoding is relatively coarse, and generally only a few layers can be supported, because too many layers will increase the difficulty of encoding and control the amount of information, and will also consume a large amount of state on the router surface

In addition, dynamically joining and leaving multicast groups will also increase overhead

And because each layer of streams travels on a completely different multicast tree, these multicast trees may not even share the same bottleneck link, in which case removing a layer of streams may not solve the congestion

Additionally, if network congestion is actually occurring on a link near the sender, the sender will still encode and send at the highest image quality, although the vast majority of receivers will not receive higher-level data

This is a great waste of the system resources of the sender and the network bandwidth in its vicinity

Images