![]() A Most Recent Used List (MRUL) storing up to N most active connections’ traffic statistics is maintained by SFRED. Notice that a router sees packets from a fast flow more often than a slow flow we propose in this paper the Scalable Fair Random Early Detection (SFRED) to improve the fairness of RED. Fair Queuing ) is the most direct solution to the unfair problems, with the large number of flows possibly sharing a link, it is not scalable for an Internet router. Secondly, when responsive TCP flows shares a RED router with unresponsive UDP flows, unresponsive UDP flows may has unreasonable high throughput than TCP flows. Firstly, when two or more TCP flow with different RTT competing for the bottleneck bandwidth, RED tends to let the flow with shorter RTT use more bandwidth. However, past work shows that unfairness of RED may occur with RED under two conditions. Known as Active Queue Management (AQM), Random Early Detection (RED) is recommended by IETF for queue management in routers. With these research efforts, a number of schemes were proposed to improve the fairness in networks with modifications to the queue management schemes implemented in Internet routers. With the increasing popularity of stream media applications, the fairness of networks has attracted much research attention. Since NLRED is fully compatible with RED, we can easily upgrade/replace the existing RED implementations by NLRED. Simulations demonstrate that NLRED achieves a higher and more stable throughput than RED and REM. the queue size may soon get out of control), NLRED allows more aggressive packet dropping to back off from it. When the load is heavy and the average queue size approaches the pre-determined maximum threshold (i.e. As a result, at light traffic load, NLRED encourages the router to operate in a range of average queue sizes rather than a fixed one. The underlying idea is that, with the proposed nonlinear packet dropping function, packet dropping becomes gentler than RED at light traffic load but more aggressive at heavy load. We call this new scheme Nonlinear RED, or NLRED. Specifically, we replace the linear packet dropping function in RED by a judicially designed nonlinear quadratic function, while original RED remains unchanged. In addition, another issue we address in this paper is queue management in RED. Different from the previous proposals, the complexity of our proposed scheme is proportional to the size of the MRUL list but not coupled with the queue buffer size or the number of active flows, so it is scalable and suitable for various routers. Its performance improvements are demonstrated with extensive simulations. Based on the list, our proposed scheme jointly considers the identification and punish of the fast and unresponsive fast flows, and the protection of slow flows. ![]() In this paper, we propose a scheme using Most Recently Used List (MRUL)-a list storing statistics of limited active flows that sorted in most recently seen first mode-to improve the fairness of RED. This suggests that network fairness may be improved without per-flow information. Internet routers generally see packets from a fast flow more often than a slow flow. Keywords: Random Early Detection, TCP, Unresponsive Flows, Fairness, Queue Management In the context of such activities, he took part and conduct research teams in numerous research projects (PNR, CNEPRU), funded by the Algerian Ministry of Higher Education and Scientific Research (MESRS).1Department of Communication Engineering, Shenzhen University, Shenzhen, ChinaĢChina Telecom Beijing Research Institute, Beijing, ChinaģShenzhen Graduate School, Peking University, Shenzhen, ChinaĤNational Mobile Communications Research Laboratory, Southeast University, Nanjing, ChinaĮmail: Octorevised Decemaccepted December 31, 2008 His research activities include techniques, methodologies and tools for modelling and intelligent control of engineering and computing systems. He received the Magister (M.Sc.) and the Doctorat d'état (Ph.D.) degrees in automatic control both from the École Nationale Polytechnique of Algiers. He is currently a Professor of automatic control at the Computer Systems Engineering Department and a Director of Research at the Laboratoire de Communication dans les Systèmes Informatiques (LCSI) of the École nationale Supérieure d'Informatique (ESI), Algiers, Algeria. (e-mail: Loudini was born in Bouzina, Batna, Algeria in 1966. His work is devoted to the study of active queue management, congestion control, and intelligent feedback control systems. student in the Laboratoire de Communication dans les Systèmes Informatiques (LCSI) of ESI. He received the engineer degree in computer science from the École nationale Supérieure d'Informatique (ESI), Algeria, in 2013. Abderrahmane Boudi (M'16) was born in Algiers, Algeria, in 1990.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |