A new approach to performance-oriented flow control
作者： K. Bharath-Kumar , J. Jaffe
DOI: 10.1109/TCOM.1981.1095007
关键词: Distributed computing 、 Network delay 、 Network simulation 、 Network traffic control 、 Throughput 、 Network performance 、 Flow control (data) 、 Computer science 、 Electrical and Electronic Engineering
摘要: Flow control is proposed as a means of obtaining an "optimal tradeoff" between low delay and high throughput in computer networks. Several versions are defined based on network power. A class algorithms which attempt to optimize performance investigated. These operate the design principles dynamic, distributed execution use local information. force be suboptimal, we thus investigate relative each different configurations. properties power objective function examined. In certain configurations, two variations unfair users by not permitting them send any messages. version ("product powers") corrects this deficiency. Other discussed include nonconvexity generalized function.
sci-hub.se 参考文章(3)
A. Giessler, J. Hänle, A. König, E. Pade, Free buffer allocation — An investigation by simulation Computer Networks. ,vol. 2, pp. 191- 208 ,(1978) , 10.1016/0376-5075(78)90028-4
J. P. Gray, T. B. McNeill, SNA multiple-system networking IBM Systems Journal. ,vol. 18, pp. 263- 297 ,(1979) , 10.1147/SJ.182.0263
L. Kleinrock, P. Kermani, Dynamic Flow Control in Store-and-Forward Computer Networks IEEE Transactions on Communications. ,vol. 28, pp. 271- 279 ,(1980) , 10.1109/TCOM.1980.1094661