论文部分内容阅读
针对在服务承载网中,传统的服务器放置方法会出现随着服务器个数的增加降低网络整体性能的问题,提出基于最大流效应的服务器放置算法。该算法不仅考虑到了流量因素,还将带宽、延迟、丢包率等网络节点性能因素简化为“距离因子”进行了考虑,解决了传统服务器放置引起的全局低效和网络拥堵问题,保证最大限度地满足用户的流量请求。经过实验验证,基于最大流效应的服务器放置算法相对于传统的服务器放置算法,在平均情况下,平均相应时间减低10%到25%,传输开销降低25%到50%,显著提高了网络性能。
In the service bearer network, the traditional server placement method will reduce the overall performance of the network as the number of servers increases, and propose a server placement algorithm based on the maximum stream effect. The algorithm not only considers the traffic factor, but also considers the network node performance factors such as bandwidth, delay and packet loss rate as the “distance factor”, and solves the problem of global inefficiency and network congestion caused by traditional server placement and guarantees Maximize the user’s traffic request. The experimental results show that the algorithm based on the maximum flow effect reduces the average time by 10% to 25% and reduces the transmission overhead by 25% to 50% compared with the traditional server placement algorithm, which improves the network performance remarkably.