摘要
技术的发展推动着传统网络向智能网络转变,智能网络中类型繁多的流量业务所产生的巨大、多样的数据流量易导致网络拥塞等问题,这给网络拥塞控制体系提出了更高的要求。主动队列管理机制是拥塞控制体系中不可或缺的一环,随机早期检测(random early detection,RED)算法及其改进算法都是主动队列管理机制中缓解网络拥塞的常见方法,但都存在无法适应网络动态变化所带来的队列长度不稳定的问题。针对该问题,提出一种新型主动队列管理算法SP-ARED(square power-ARED),SP-ARED算法通过优化丢弃概率函数以及引入新的阈值上限,来更好地缓解网络拥塞。NS2的仿真结果表明,在复杂、动态的网络环境下,除时延以外,SP-ARED算法在平均队列长度稳定性、吞吐量、丢包率和时延抖动方面的表现都有了提高。
The development of technology promotes the transformation from the traditional network to the intelligent network.The hefty and diverse data traffic generated by various traffic services in in-telligent network is easy to cause network congestion and other problems,which puts forward higher re-quirements for network congestion control system.The active queue management mechanism is an in-dispensable part of the congestion control system.Random Early Detection(RED)algorithm and its im-proved algorithm are common methods to alleviate network congestion in the active queue management mechanism,but they all have the problem that the queue length is not stable due to the dynamic changes of the network.To solve this problem,a new active queue management algorithm SP-ARED(Square Power-ARED)is proposed,which can better alleviate network congestion by optimizing discard probability function and introducing new threshold upper limit.The simulation results of NS2 show that the performance of SP-ARED algorithm is improved in terms of average queue length stability,throughput,packet loss rate and delay jitter except delay under the complex and dynamic network envi-ronment.
作者
潘成胜
张松
赵晨
石怀峰
PAN Chengsheng;ZHANG Song;ZHAO Chen;SHI Huaifeng(School of Electronics and Information Engineering,Nanjing Information Science&Technology University,Nanjing 210044,China;School of Automation,Nanjing Science and Technology University,Nanjing 210094,China)
出处
《火力与指挥控制》
CSCD
北大核心
2023年第1期1-7,共7页
Fire Control & Command Control
基金
国家自然科学基金(61931004,61801073)
军委科技委基础加强计划资助项目(2022-JCJQ-JJ-0488)。
