Multi-dimensional behavior analysis and optimizationof traffic in campus network

被引:0
|
作者
Gong X. [1 ]
Shang Q. [1 ]
Fu Z. [1 ]
Qian J. [1 ]
机构
[1] Computer Center, Peking University, Beijing
关键词
HTTP Streaming; Multiple dimensions; Protocol preference; Traffic behavior analysis; Traffic optimization;
D O I
10.13245/j.hust.16S127
中图分类号
学科分类号
摘要
Aimed at increasing congestion in campus network export and the coarse-grain flow control strategy adapting to the changing campus traffic environment difficultly, traffic features of several high bandwidth occupied network protocols were analyzed, general traffic behavior in campus network was analyzed deeply from time, regional, access, and chain multiple dimensions, more fine-grained traffic optimization strategies were proposed according to the analysis results. The results show that, HTTP Streaming traffic takes the highest percentage of campus inbound traffic and grows quickly, bandwidth of users inbound traffic appears power law distribution, different regions, access ways, and ISP chains have different preferences on network application protocols. More fine-grained traffic optimization strategies can ease campus network export congestion and optimize network performance effectively. © 2016, Editorial Board of Journal of Huazhong University of Science and Technology. All right reserved.
引用
收藏
页码:131 / 137
页数:6
相关论文
共 3 条
  • [1] Chan K.M., Lee J.Y.B., Improving adaptive HTTP streaming performance with predictive transmission and cross-layer client buffer estimation, Multimedia Tools and Applications, 75, 10, pp. 5917-5937, (2016)
  • [2] Brice A., Abdelhamid M., On traffic patterns of HTTP applications, Global Telecommunications Conference, pp. 1-6, (2011)
  • [3] Aaron G., Ashok A., Aditya A., A comparative study of handheld and non-handheld traffic in campus Wi-Fi networks, Passive and Active Measurement, pp. 173-183, (2011)