卫星网络混合负载均衡策略下的多径流量分配算法
Multipath traffic allocation algorithm based on hybrid load balancing strategy in satellite networks
魏德宾 1王英杰 2梁超2
作者信息
- 1. 大连大学信息工程学院,辽宁大连 116622;大连大学通信与网络重点实验室,辽宁大连 116622;南京理工大学自动化学院,江苏南京 210094
- 2. 大连大学信息工程学院,辽宁大连 116622;大连大学通信与网络重点实验室,辽宁大连 116622
- 折叠
摘要
为均衡卫星网络流量分配,满足用户QoS需求,提出一种全局和局部结合的混合负载均衡策略下的多径流量分配算法(HLB-MTD).根据QoS业务需求进行部分路径筛选,基于表的散列算法进行流量的初次分配;在初次分配的基础上针对突发拥塞进行局部调整,建立重映射目标函数,通过改进交叉和变异概率的遗传算法求出最优解;求得局部优化的,流到路径的映射策略.仿真结果表明,该算法可有效缓解卫星网络拥塞,实现流量均衡分配,在丢包率、平均排队时延、等指标上有更好的提升.
Abstract
To balance the satellite network traffic distribution and meet the user's QoS requirements,a multi-path traffic distri-bution algorithm(HLB-MTD)based on a hybrid global-local load balancing was proposed.The partial path screening was per-formed according to the QoS service requirements,and the initial allocation of traffic was performed based on the table-based hash algorithm.Based on the initial allocation,local adjustments were made for the burst congestion,and the remapping objec-tive function was established.The optimal solution was obtained by improving the crossover and mutation probability inheritance algorithm.The flow to path mapping strategy for local optimization was obtained.Simulation results show that the algorithm can effectively alleviate the congestion of satellite network,achieve the balanced distribution of traffic,and improve the packet loss rate,average queuing delay,and other indicators.
关键词
卫星网络/流量分配/负载均衡/服务质量需求/目标优化/散列重分配/遗传算法Key words
satellite network/traffic distribution/load balancing/quality of service(QoS)requirement/objective optimization/hash redistribution/genetic algorithm引用本文复制引用
出版年
2024
NETL
NSTL
万方数据