基于网络编码的可用带宽预测算法的研究与实现
首发时间:2020-04-16
摘要:互联网20年的发展,使得网络传输越来越受到学术界和工业界的广泛关注。同时近些年数据规模的爆炸式增长、动态异质网络的个性化需求,使得保障网络传输质量成为重大挑战。而网络传输的优化目标其实是动态匹配到瓶颈链路的可用带宽。本文总结了经典的两类带宽预测算法,包括经典的调和平均法、Pathload、TOPP和UKF等算法的优缺点。本文设计了一种基于网络编码的带宽预测算法,阐述了其设计原理,论证了其重要的参数意义,并完成起代码的实现。其次,本文又设计了测试平台。该平台主要有发送端、瓶颈链路控制器、中间链路、接收端等组成。最后,本文针对五种算法,分别在单跳固定带宽、单跳混合流量、多跳混合流量和Internet中进行测试评定。对其评测结果进行对比和分析,并论证了Pace code在各个场景下的鲁棒性。
For information in English, please click here
Research and Implementation of Available Bandwidth Estimation Algorithm Based on Network Coding
Abstract:In recent years, the ever-changing demand for network applications has made network transmission more and more widely concerned by academia and industry. At the same time, the explosive growth of data scale and the personalized needs of dynamic heterogeneous networks have made it a major challenge to ensure network transmission quality. The optimization goal of network transmission is actually to dynamically match the available bandwidth of the bottleneck link. This article summarizes two classic types of bandwidth prediction algorithms, and selects its classic harmonic averaging, Pathload, TOPP, and UKF algorithms. First, this paper designs a bandwidth prediction algorithm based on network coding, expounds its design principles, demonstrates the significance of its important parameters, and completes the code implementation. Secondly, this article has designed a test platform. The platform is mainly composed of a sending end, a bottleneck link controller, an intermediate link, and a receiving end. Finally, this paper tests and evaluates the five algorithms in single-hop fixed bandwidth, single-hop mixed traffic, multi-hop mixed traffic, and the Internet. Compare and analyze its evaluation results, and demonstrate the robustness of Pace code in various scenarios.
Keywords: Network coding Network transmission Bandwidth estimation
基金:
引用
No.****
动态公开评议
共计0人参与
勘误表
基于网络编码的可用带宽预测算法的研究与实现
评论
全部评论