刘远超, 郑金芳. 网络负载过高中的均衡技术研究与仿真[J]. 微电子学与计算机, 2016, 33(2): 115-117.
引用本文: 刘远超, 郑金芳. 网络负载过高中的均衡技术研究与仿真[J]. 微电子学与计算机, 2016, 33(2): 115-117.
LIU Yuan-chao, ZHENG Jin-fang. Under the Optimal Channel Selection of Network Load Balancing Technology Research[J]. Microelectronics & Computer, 2016, 33(2): 115-117.
Citation: LIU Yuan-chao, ZHENG Jin-fang. Under the Optimal Channel Selection of Network Load Balancing Technology Research[J]. Microelectronics & Computer, 2016, 33(2): 115-117.

网络负载过高中的均衡技术研究与仿真

Under the Optimal Channel Selection of Network Load Balancing Technology Research

  • 摘要: 随着网络传输技术的发展, 信道传输中的负载均术受到了各国学者的重视.当前的网络最优信道均衡技术异构网络复杂, 易造成呼叫阻塞率并且网络的吞吐量也随之降低.对此提出一种采用基于门限值的负载均衡指数的两阶段的负载均衡算法, 针对负载均衡的目标, 在基本通用分配问题的基础上, 采用均衡GAP算法和纵向效用GAP算法, 在网络通信中将传输信息均衡地分配到接入网中.根据不同数据需求偏向性, 通过多维度分析计算出不同传输对网络属性权重, 通过加权计算对不同业务选择最大参数进行网络传输.仿真结果表明, 基于层次分析法和多属性决策的负载均衡方法能够有效均衡网络负载, 考虑到包括负载在内的多个网络属性, 降低了系统的呼叫阻塞率, 同时提高了系统吞吐量.

     

    Abstract: With the development of the network transmission technology, the channel transmission load are attention by international scholars. The current network optimal heterogeneous network of channel equalization technology is complex, easy cause call model.next and network throughput decreases. Put forward a kind of load balancing based on threshold value index of the two stages of the load balancing algorithm, aiming at the goal of load balance, on the basis of the basic general assignment problem, and longitudinal utility GAP algorithm, using the algorithm of balanced GAP will transmit information in network communication balanced allocation in the access network. Demand preference according to different data, through dimensional analysis to calculate the attribute weights of different transmission network, by weighted calculation to choose a different business biggest parameters in network transmission. The simulation results show that based on the analytic hierarchy process (ahp) and multiple attribute decision making method of load balance can effectively balance the network load, considering the multiple network attributes, including load, reduces the system call model.next, at the same time improve the system throughput.

     

/

返回文章
返回