实时视频系统的数据编码速率智能控制

该文探讨在CDMA无线网络环境下实现更高QoS的实时视频传输，包括了对实时视频编码速率调整的智能控制方法，以及如何对编码速率调整的内容。采用网络特征反馈的样本，将其通过动态神经网络异步输入，经过合适的训练，得到一个确定神经网络参考模型，用该模型对未来的网络环境作出预测，并由预测出来的结果控制数据编码调整方案的一种智能控制手段。通过神经网络来实现对网络特征的捕捉，再作出合适的编码速率调整，这种调整更接近真实需要，更好地满足了实时QoS的实时视频系统要求，使得实时性能得到显著的提高。

Smart Control of Data Coding Rate for Real-time Video System

This article discusses higher QoS problem for real-time video transmission under CDMA wireless environment,including the smart control method of the coding rate adjustment for real-time video,and the content how to adjust the coding rate. According to the feedback samples got from network's characteristics,and be properly trained through dynamic asynchronous neural network,we can get a reference model of a certain neural network.By using the model,we can make a predication for the network environment in the near future,and control the data coding rate according to the predicated results.We realize the trap for network's characteristics,and then adjust the coding rate.This scheme is more suitable for the real needs,and meets the QoS requirement of real-time video system.and the performance can be significantly improved.