Font Size: a A A

Research On The IP QoS Technology Based On Diffserv Model

Posted on:2013-02-20Degree:MasterType:Thesis
Country:ChinaCandidate:J LiFull Text:PDF
GTID:2218330371957633Subject:Circuits and Systems
Abstract/Summary:PDF Full Text Request
With the rapid expansion of the Internet scale and the growing number of services, the problem that traditional network lacks guaranteed QoS(quality of service) exposed gradually, IP QoS was born in that case. Aimed at the QoS problem, DiffServ model becomes the first choice depending on its simplicity and expandability, and relative differentiation is the direction of the DiffServ model with the simplicity of deployment and management, proportional differentiated services is one of them. In the DiffServ, queue scheduling algorithms is still a hot topic to achieve QoS.At the beginning, the basic principle of DiffServ and the proportional delay differentiated model which is predictable and controllable are introduced in this thesis, familiar queue scheduling algorithms in DiffServ are discussed in detail, and then summarize the merits and the demerits of these algorithms. Next, a dynamically queue scheduling algorithm based DWRR is proposed, which combines the proportional delay differentiated model—PDDRR, the algorithm adjusts each queue's weight according to the average packet arrival rate to guarantee that the ratio of the average queuing delay keeping in the constant proportion when there is no burst traffic, neglect the proportion fairness principle and increase all the queue's weights linearly under burst traffic environment. At last, PDDRR and its associated algorithms are tested on NS2 simulation platform, the results demonstrate that this strategy not only merely realize the proportion delay differentiated under proper parameters, but also improve the capabilities to handle the burst traffic, and reflect high reliability in delay and average loss rate.
Keywords/Search Tags:DiffServ, Queue scheduling, Proportional delay, Quality of Service, Burst traffic
PDF Full Text Request
Related items