| Title | A Scientific Data Traffic Scheduling Algorithm Based on Software-Defined Networking |
| Publication Type | Conference Paper |
| Year of Publication | 2019 |
| Authors | Zhou, Yaqiu, Ren, Yongmao, Zhou, Xu, Yang, Wanghong, Qin, Yifang |
| Conference Name | 2019 IEEE 21st International Conference on High Performance Computing and Communications; IEEE 17th International Conference on Smart City; IEEE 5th International Conference on Data Science and Systems (HPCC/SmartCity/DSS) |
| Keywords | Bandwidth, common data flow, compositionality, computer network security, data flow computing, data plane, delays, Firewalls (computing), network security devices, performance evaluation, Predictive Metrics, private network, pubcrawl, resilience, Scheduling algorithms, Science DMZ model, scientific big data transfer, Scientific Computing Security, scientific data flow transfer, scientific data traffic scheduling algorithm, scientific data transfer, scientific information systems, SDN-based traffic scheduling algorithm, software defined networking, software-defined network, software-defined networking, Switches, telecommunication scheduling, telecommunication traffic, traffic scheduling, traffic scheduling algorithm |
| Abstract | Compared to ordinary Internet applications, the transfer of scientific data flows often has higher requirements for network performance. The network security devices and systems often affect the efficiency of scientific data transfer. As a new type of network architecture, Software-defined Networking (SDN) decouples the data plane from the control plane. Its programmability allows users to customize the network transfer path and makes the network more intelligent. The Science DMZ model is a private network for scientific data flow transfer, which can improve performance under the premise of ensuring network security. This paper combines SDN with Science DMZ, designs and implements an SDN-based traffic scheduling algorithm considering the load of link. In addition to distinguishing scientific data flow from common data flow, the algorithm further distinguishes the scientific data flows of different applications and performs different traffic scheduling of scientific data for specific link states. Experiments results proved that the algorithm can effectively improve the transmission performance of scientific data flow. |
| DOI | 10.1109/HPCC/SmartCity/DSS.2019.00024 |
| Citation Key | zhou_scientific_2019 |
A Scientific Data Traffic Scheduling Algorithm Based on Software-Defined Networking