Strict increase loose decrease equal step congestion control algorithm based on mobile communication network

Abstract

The invention relates to a strict increase loose decrease equal step congestion control algorithm based on a mobile communication network. According to the algorithm, packet loss ratio of the mobile communication network is calculated by adopting a linear prediction algorithm at an application layer. Code rate is regulated by adopting a mode of comparison of the number of times of cumulative abnormal cycles of packet loss ratio and different set values according to the size of packet loss ratio. Sending code rate is up-regulated when the network state is great, and the sending code rate is down-regulated when the network state is poor; and frame quality is influenced to some extent when amplitude of the code rate is down-regulated, while consequences of rising of packet loss ratio and increasing of end-to-end time delay due to network congestion caused by the fact that output code rate surpasses available bandwidth when the code rate is up-regulated are generated so that the condition of code rate up regulation is stricter than that of down regulation (i.e. strict increase loose decrease) Meanwhile, amplitude (step) of code rate regulation each time is fixed, and the code rate is rapidly regulated to a reasonable range so that self-adaptive sending code rate change is realized, video time delay and jittering are reduced and stability of video quality is guaranteed.

Description

technical field [0001] The invention relates to a QoS (Quality of Service, QoS) method based on a mobile communication network in the technical field of mobile communication. In particular, it relates to a step-size congestion control algorithm such as Strict Increase Loose Decrease (SILD) for real-time video transmission based on a mobile communication network application layer. Background technique [0002] With the rapid development of mobile communication technology, QoS in network video transmission has become an important research topic. Compared with the traditional IP network, the available bandwidth of 3G (3rd-generation, third-generation mobile communication) and 4G (4rd-generation, fourth-generation mobile communication) networks is limited, the frequency band used is high, and it is affected by multipath fading, shadow fading and Influenced by interference noise, the same-frequency adjacent-channel interference is serious, and the mobile channel is greatly affec...

AI Technical Summary

Problems solved by technology

That is, at the beginning of data transmission, the network situation cannot be judged, so the sending end cannot calculate the sending rate within a period of time. For real-time multimedia applications, there will be a large time delay and time delay at the beginning of the receiving end. jitter

Therefore, the congestion control of the traditional IP network mainly has the following problems: the control is at the transport layer rather than the application layer; it cannot meet the requirements of real-time, rapidity and video quality stability at the same time

However, the algorithm does not propose a specific and effective implementation, nor does it consider the impact of rate changes on video quality and the difference between increasing and decreasing the sending rate

Because the violent fluctuation of the video output bit rate will cause the parameters such as the time delay and jitter of the receiving end to change accordingly, which will affect the stability of the video picture and may cause significant changes in the picture quality

Images