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：
作为最著名的网络拥塞控制机制，随机早期检测（RandonlEarlyDetection，简称RED）算法由于其参数敏感性，无法在复杂多变的网络环境下保障良好的控制性能。为了改善RED敏感于参数的缺陷，增强算法的自适应性，文章将补偿模糊神经网络（compensatoryfuzzyneuralnetwork，简称CFNN）引入拥塞控制算法的设计中，结合RED和CFNN，得到了基于CFNN的RED变种算法（REDbasedonCFNN，简称CFNNRED）。与传统的RED相比，CFNNRED的改进在于：配置神经元一定的模糊逻辑规则，迅速得到丢包率，增强算法的可操作性和可实现性；通过神经网络的自学习，增强算法的自适应性和鲁棒性。最后通过仿真证明，CFNNRED算法的自适应性增强，对队列的控制能力得到加强，队列更加平稳，网络能够提供更加稳定的服务质量保障。
英文摘要：
As the most well-known congestion control mechanism, RED couldn＇ t keep good performance in the dynamical and changeable network circumstances because of its parameters＇ sensitivity. In order to overcome RED＇ s sensitivity and promote its seff-adaptation, a novel AQM algorithm called CFNNRED was presented by combining RED with CFNN. Different to RED, CFNNRED got packet drop probability according to fuzzy logic rules preconfigured in neurons and adjusting by self-learning of the neural network. As a result, CFNNRED got better performance than RED with higher self-adaptation and robustness. Simulations demonstrated that CFNNRED could exert powerful control on queues and provide better QoS.
同期刊论文项目
同项目期刊论文
H-infinity control for networked control systems with data packet dropouts and partly unknown transi
Mean square consensus control for second-order multi-agent systems under fixed topologies and measur
Building-Environment Control With Wireless Sensor and Actuator Networks: Centralized Versus Distribu
Empty glass bottle inspection method based on fuzzy support vector machine neural network and machin
Stabilizing switching design for switched linear systems: A state-feedback path-wise switching appro
Scheduling strategy for logarithmic quantization under bandwidth constraints based on controller are
Dual decomposition method for optimal and fair congestion control in Ad Hoc networks: Algorithm, imp
位置：