الفهرس | Only 14 pages are availabe for public view |
Abstract Recently, the Internet became the most important environment for many activities including emails, phone calls and videoconferences. Nevertheless, the incremental growth of the internet traffic leads to a serious problem called congestion. Several Active Queue Management (AQM) algorithms have been developed at the internet routers to avoid congestion before happening and solve the congestion if it happens by actively controlling the average queue length in the routers. However, most of the developed algorithms handle all the internet traffics by the same strategy although the internet traffics, real time and non-real time, require different Quality of Service (QoS). This problem may lead to performance degradation especially for realtime applications as video and audio traffics. This thesis first presents a performance evaluation of the current AQM algorithms. It then presents a new (class-based dynamic scheduling strategy) AQM algorithm, called Dynamic Queue RED (DQRED), to guarantee efficient QoS to both real-time and non-real-time traffics. In the proposed algorithm, three queues are used in the internet router; one queue for each traffic type (data, audio and video). The arrived packets are first queued in the corresponding queue. The queued packets are then scheduled dynamically according to the load (the number of queued packets) of each class type. This strategy guarantees QoS for real time applications as well as service fairness. Finally, a comparative study is done between the proposed DQRED algorithm and the most recent AQM algorithms by using the network simulator (NS-2) considering different QoS Metrics. |