综合式拥塞控制机制的研究

针对无线传感器网络中的SFB拥塞控制算法在网络流量大时,其分类机制受影响而导致TCP流传输性能下降的问题,介绍了一种综合式拥塞控制机制的算法实现,并在NS-2平台上对该机制的性能进行了仿真研究。综合式拥塞控制机制通过建立新路径对数据包分流来解决拥塞问题;当新路径建立失败时,采用公平汇聚算法,按比例减少源节点的发送率,保证每个源节点公平发送数据;对于轻度拥塞情况,采用一种节能算法,即设置拥塞时间阈值,只有当拥塞时间超过阀值时才启动新路径查找机制。仿真结果表明,综合式拥塞控制机制能够减少丢包率,有效缓解拥塞,并能够根据数据权重的不同可靠发送重要数据。

Research of Integrated Congestion Control Mechanism

In order to solve the problem that classifying mechanism of SFB congestion control algorithm for wireless sensor network could be influenced to cause transmission performance degradation of TCP flow under large network flow,the paper introduced algorithm implementation of an integrated congestion control mechanism,and did simulation research of performance of the mechanism by use of NS-2 platform.The mechanism solves congestion by establishing a new path to distributing data packets.When the new path is not established successfully,the mechanism uses fair aggregate algorithm to limit sending rate of source node proportionably so as to ensure each source node to send data fairly.For light congestion,the mechanism uses an energy-saving algorithm,namely setting a threshold value of congestion time and starting the new path searching mechanism when congestion time is over the threshold value.The simulation result showed that the mechanism can reduce packet loss rate,relieve congestion effectively,and send importance data according to data weight.