摘要
分析了常用EJB集群系统的设计方法与负载均衡算法,设计了一种软件Proxy系统,并针对常用算法的不足提出了一种基于BP神经网络的负载均衡算法。该算法具有较强学习能力、容错性强和自适应能力强的特点。仿真实验证明,基于该算法的集群系统具有较好的适应能力和伸缩性。
The design methods and load balancing algorithm of EJB clustering system was analyzed, and a new approach based on the BP neural networks was presented to solve the load balancing algorithm problem. How to design the BP Neural Network model was researched. The example shows that the system based on this algorithm is proved to reach better effects.
出处
《计算机应用》
CSCD
北大核心
2006年第4期850-852,共3页
journal of Computer Applications