欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
提出一种有效的GSM／GPRS网络分组数据QoS的动态规划方案。针对分组数据在传送过程中的数据延迟大、动态QoS控制困难，提出一种基于分布式智能代理实现的流量控制结构，改进了QoS性能；采用基于模糊神经网络的流量预测技术，其优点是训练样本少、精确度高；为保证铁路突发数据业务传递的实时性，提出一种运算结构简单的业务优先级判决机制。仿真结果表明，通过分布式并行计算，可有效平衡网络负载，降低包延迟和路由开销，网络传输性能获得明显改善。
英文摘要：
The dynamic Quality of Service （QoS） scheme is presented for the GSM/GPRS wireless network. GPRS packet data transmission has the disadvantages of long delay and difficult QoS control. In order to solve these key problems, a load balancing architecture is constructed and it supports real time and burst data services by distributed intelligent agents. The fuzzy neural network is employed to predict GPRS traffic. Good results have been achieved via training samples. A traffic estimation algorithm and a simple priority decision mechanism are formed to deal with special applications such as burst data transmission for railways. From the view point of distributed computing, we have built an agent model to balance network services. The simulation shows that the distributed intelligent agent architecture can significantly reduce packet delay and route cost and relieve GPRS bottleneck.
同期刊论文项目
同项目期刊论文
期刊信息
位置：