Implementation and comparison of active queue management algorithms in traditional and SDN networks
محورهای موضوعی : شبکه های کامپیوتریKhoshnam Salimi Beni 1 , Mohammad Reza Soltan Aghaei 2 , Rasool Sadeghi 3
1 - Department of Computer Engineering, Isfahan (Khorasgan) Branch, Islamic Azad University, Isfahan 81551-39998, Iran
2 - 2Institute of Artificial Intelligence and Social and Advanced Technologies, Isf.C., Islamic Azad University, Isfahan, Iran
3 - Department of Electrical Engineering, Dolatabad Branch, Islamic Azad University, Isfahan, Iran
کلید واژه:
چکیده مقاله :
In the past decade, networks have experienced significant improvements in scale and data transfer rates, and network traffic rates will soon increase dramatically. Network management and traffic control play key roles in real-time data transmission (such as video conferencing, high-bandwidth streams, video calls, etc.) and data transmission in the Internet of Things (IoT). Although technologies such as SSD storage and virtualization are very effective in meeting network traffic needs. Future networks will require centralized management, easy upgradeability, application optimization, efficient resource allocation, and dynamic routing. To meet these requirements, the benefits of software-defined networking (SDN) must be used. By separating the control part from the data part, SDN will lead to scalability, flexibility and centralized management of the network. With excessive demands on limited network resources, it is inevitable to create long queues of information packets in intermediate routers, and the use of active queue management (AQM) algorithms of TCP/IP network in order to make more use of available bandwidth and reduce Transmission delay is necessary. In this article, we examine some of the most important active queue management algorithms including PFIFO_fast, ARED, CoDel, FQ-CoDel and PIE in traditional and SDN networks. The results of the simulation show that the use of AQM algorithms in the SDN network reduces the average delay and packet loss rate and increases the network efficiency.
In the past decade, networks have experienced significant improvements in scale and data transfer rates, and network traffic rates will soon increase dramatically. Network management and traffic control play key roles in real-time data transmission (such as video conferencing, high-bandwidth streams, video calls, etc.) and data transmission in the Internet of Things (IoT). Although technologies such as SSD storage and virtualization are very effective in meeting network traffic needs. Future networks will require centralized management, easy upgradeability, application optimization, efficient resource allocation, and dynamic routing. To meet these requirements, the benefits of software-defined networking (SDN) must be used. By separating the control part from the data part, SDN will lead to scalability, flexibility and centralized management of the network. With excessive demands on limited network resources, it is inevitable to create long queues of information packets in intermediate routers, and the use of active queue management (AQM) algorithms of TCP/IP network in order to make more use of available bandwidth and reduce Transmission delay is necessary. In this article, we examine some of the most important active queue management algorithms including PFIFO_fast, ARED, CoDel, FQ-CoDel and PIE in traditional and SDN networks. The results of the simulation show that the use of AQM algorithms in the SDN network reduces the average delay and packet loss rate and increases the network efficiency.
