This paper discusses traffic engineering with multi protocol label switching (MPLS) in a backbone of Internet data center (IDC) network. The traditional SPF routing limitations are analyzed and the traffic engineerin...This paper discusses traffic engineering with multi protocol label switching (MPLS) in a backbone of Internet data center (IDC) network. The traditional SPF routing limitations are analyzed and the traffic engineering concept is given. MPLS, resource reservation protocol (RSVP) and enhanced link status Protocols intermediate system to intermediate system (IS IS) are reviewed to provide a background for traffic engineering, the general issues of designing an MPLS system of IDC network for traffic engineering are then discussed. Finally a practical example with MPLS traffic engineering is shown.展开更多
Segment Routing(SR)is a new routing paradigm based on source routing and provide traffic engineering(TE)capabilities in IP network.By extending interior gateway protocol(IGP),SR can be easily applied to IP network.How...Segment Routing(SR)is a new routing paradigm based on source routing and provide traffic engineering(TE)capabilities in IP network.By extending interior gateway protocol(IGP),SR can be easily applied to IP network.However,upgrading current IP network to a full SR one can be costly and difficult.Hybrid IP/SR network will last for some time.Aiming at the low flexibility problem of static TE policies in the current SR networks,this paper proposes a Deep Reinforcement Learning(DRL)based TE scheme.The proposed scheme employs multi-path transmission and use DRL to dynamically adjust the traffic splitting ratio among different paths based on the network traffic distribution.As a result,the network congestion can be mitigated and the performance of the network is improved.Simulation results show that our proposed scheme can improve the throughput of the network by up to 9%than existing schemes.展开更多
The Software Defined Networking(SDN) paradigm separates the control plane from the packet forwarding plane, and provides applications with a centralized view of the distributed network state. Thanks to the flexibility...The Software Defined Networking(SDN) paradigm separates the control plane from the packet forwarding plane, and provides applications with a centralized view of the distributed network state. Thanks to the flexibility and efficiency of the traffic flow management, SDN based traffic engineering increases network utilization and improves Quality of Service(QoS). In this paper, an SDN based traffic scheduling algorithm called CATS is proposed to detect and control congestions in real time. In particular, a new concept of aggregated elephant flow is presented. And then a traffic scheduling optimization model is formulated with the goal of minimizing the variance of link utilization and improving QoS. We develop a chaos genetic algorithm to solve this NP-hard problem. At the end of this paper, we use Mininet, Floodlight and video traces to simulate the SDN enabled video networking. We simulate both the case of live video streaming in the wide area backbone network and the case of video file transferring among data centers. Simulation results show that the proposed algorithm CATS effectively eliminates network congestions in subsecond. In consequence, CATS improves the QoS with lower packet loss rate and balanced link utilization.展开更多
针对红外船舶图像目标特征模糊、背景复杂以及小目标漏检等问题,基于YOLOv8提出一种面向海上交通中船舶目标的检测算法YOLO-IST(YOLO for infrared ship target)。在基线模型的骨干网络中引入C2f_DBB模块和CPCA注意力机制,通过增加特征...针对红外船舶图像目标特征模糊、背景复杂以及小目标漏检等问题,基于YOLOv8提出一种面向海上交通中船舶目标的检测算法YOLO-IST(YOLO for infrared ship target)。在基线模型的骨干网络中引入C2f_DBB模块和CPCA注意力机制,通过增加特征提取层来提升模型对目标的识别能力;利用C2f_Faster_EMA模块替换颈部网络中的C2f模块,以提升模型检测精度和速度;采用多重注意力的动态检测头Dynamic Head优化模型框架,增强模型对小船舶目标的检测效果。研究结果表明:YOLO-IST的召回率R_(ecall)、精确率P_(recision)、平均精度M_(ap@50)、平均精度M_(ap@50-95)和F_(1score)分别达到89.7%、90.5%、94.7%、66.6%、90.1%,较基线模型YOLOv8分别提升了4.5%、3.8%、4.4%、4.7%、4.2%。该模型的提出在海上智能交通中具有较广泛的应用前景。展开更多
文摘This paper discusses traffic engineering with multi protocol label switching (MPLS) in a backbone of Internet data center (IDC) network. The traditional SPF routing limitations are analyzed and the traffic engineering concept is given. MPLS, resource reservation protocol (RSVP) and enhanced link status Protocols intermediate system to intermediate system (IS IS) are reviewed to provide a background for traffic engineering, the general issues of designing an MPLS system of IDC network for traffic engineering are then discussed. Finally a practical example with MPLS traffic engineering is shown.
基金supported by the National Key R&D Project(No.2020YFB1804803)the Research and Development Program in Key Areas of Guangdong Province(No.2018B010113001)。
文摘Segment Routing(SR)is a new routing paradigm based on source routing and provide traffic engineering(TE)capabilities in IP network.By extending interior gateway protocol(IGP),SR can be easily applied to IP network.However,upgrading current IP network to a full SR one can be costly and difficult.Hybrid IP/SR network will last for some time.Aiming at the low flexibility problem of static TE policies in the current SR networks,this paper proposes a Deep Reinforcement Learning(DRL)based TE scheme.The proposed scheme employs multi-path transmission and use DRL to dynamically adjust the traffic splitting ratio among different paths based on the network traffic distribution.As a result,the network congestion can be mitigated and the performance of the network is improved.Simulation results show that our proposed scheme can improve the throughput of the network by up to 9%than existing schemes.
基金partly supported by NSFC under grant No.61371191 and No.61472389
文摘The Software Defined Networking(SDN) paradigm separates the control plane from the packet forwarding plane, and provides applications with a centralized view of the distributed network state. Thanks to the flexibility and efficiency of the traffic flow management, SDN based traffic engineering increases network utilization and improves Quality of Service(QoS). In this paper, an SDN based traffic scheduling algorithm called CATS is proposed to detect and control congestions in real time. In particular, a new concept of aggregated elephant flow is presented. And then a traffic scheduling optimization model is formulated with the goal of minimizing the variance of link utilization and improving QoS. We develop a chaos genetic algorithm to solve this NP-hard problem. At the end of this paper, we use Mininet, Floodlight and video traces to simulate the SDN enabled video networking. We simulate both the case of live video streaming in the wide area backbone network and the case of video file transferring among data centers. Simulation results show that the proposed algorithm CATS effectively eliminates network congestions in subsecond. In consequence, CATS improves the QoS with lower packet loss rate and balanced link utilization.
