摘要:
为多媒体网络定义一个合理有效的性能分析模型,以描述基于神经网络的计算机网络性能评价过程,其中包括评价指标测量值的预处理、神经网络结构设计、评价仿真以及仿真结果分析4个部分。仿真实验结果表明,基于神经网络的评价方法无须建立复杂的数学模型,能够对多个网络性能指标进行综合考虑,从而实现在给定输入负载下的网络性能判定。
关键词:
多媒体网络,
网络性能,
神经网络,
评价方法
Abstract:
A rational and effective performance analysis model is defined for multimedia network, which is used to describe the computer network performance evaluation process based on neural network. It includes measuring value pretreatment of evaluation metrics, neural network architecture design, evaluation simulation and simulation result analysis. Simulation experimental results show that the evaluation method based on neural network without math model can consider many metrics, and realize the judgment of network performance under the given input overload.
Key words:
multimedia network,
network performance,
neural network,
evaluation method
中图分类号:
梁 靓;黄玉清;张晓琴. 基于神经网络的计算机网络性能评价[J]. 计算机工程, 2010, 36(4): 105-106.
LIANG Liang; HUANG Yu-qing; ZHANG Xiao-qin. Computer Network Performance Evaluation Based on Neural Network[J]. Computer Engineering, 2010, 36(4): 105-106.