摘要
针对λ网络中的高性能传输问题,提出了一种分布式速率分配算法.该算法基于离散时间系统模型对λ网络进行建模,得到其速率分配问题的数学依赖关系,根据最大-最小公平准则,利用本地信息来分配每个源端和目的端之间的容量,以确保较低期望速率会话的速率最大化,并把剩余容量均匀地分配给全部会话,直至网络被充分利用.结果表明,所提出的分布式速率分配算法能快速收敛到一个最大-最小公平速率分配方式,而且还能实现自适应期望速率的快速变化.
Aiming at the problem of high performance transmission inλ-networks,a distributed rate allocation algorithm was proposed.The modeling forλ-networks was carried out with the proposed algorithm based on the discrete time system model,and the mathematical dependence relationship of rate allocation problem was obtained.Based on the max-min fairness criterion,the capacity between each source end and destination end was allocated with the local information to ensure the maximization of session rates for the sessions with the lower desired rates,and the remaining capacity was evenly assigned to all the sessions until the network was fully utilized.The results show that the proposed distributed rate allocation algorithm can quickly converge to a max-min fairness rate allocation mode,but also can adapt to the fast change of desired rate.
作者
孙士国
SUN Shi-guo(Department of Basic Curriculum,North China Institute of Science and Technology,Sanhe 065201,China)
出处
《沈阳工业大学学报》
EI
CAS
北大核心
2018年第6期687-691,共5页
Journal of Shenyang University of Technology
基金
河北省科技厅科技攻关项目(112135107)
关键词
λ网络
分布式算法
容量
速率分配
最大最小公平准则
自适应
收敛
平衡点
λ-network
distributed algorithm
capacity
rate allocation
max-min fairness criterion
adaptivity
convergence
equilibrium point
