论文部分内容阅读
针对主动队列管理(AQM)机制面对动态突变的网络存在参数配置难问题,提出一种将模糊AQM和活动流参数估计策略相结合的自适应AQM算法(NFL).在综合权衡各性能指标的基础上,设计了一组能适应一定网络变化的模糊规则,并对算法进行了运算优化.为捕获网络突发流,引入了一种基于Bloom滤波器的无状态维护活动流参数估计策略,并依此提出一个模糊AQM输出增益补偿器.实验结果表明,NFL能较好地适应网络变化,相对其他算法,具有更快的收敛速度和稳定的稳态队列控制性能.
Aiming at the difficulty in parameter configuration of AQM with dynamic mutation, a new adaptive AQM algorithm (NFL) based on fuzzy AQM and activity flow parameter estimation is proposed. After comprehensive consideration of various performance indexes Based on the above, a set of fuzzy rules that can adapt to certain network changes is designed and the algorithm is optimized.A strategy based on Bloom filter for stateless maintenance of active flows is proposed A fuzzy AQM output gain compensator is proposed in this paper.Experimental results show that NFL can adapt to network changes well and has faster convergence speed and stable steady state queue control performance than other algorithms.