تخفيض تأخير الرتل باستخدام البارامترات المتكيفة للكشف العشوائي المبكر

Various technologies are used for congestion processing in computer networking, but RED (Random Early Detection) technology is the great one of new technologies, aiming to avoid difficulties and obstacles which may occur by using other technologies such as drop tail which has global synchronization for TCP flows. So RED technology has a wide range of applications in transaction from congestion control to congestion avoidance and which accomplishing with advantages and changes on network parameters. This research aims to present a model that takes the current queue size of RED gateway of the random load network as an input signal to be handled within the model that is consist of the Low Pass Filter with adjustable parameters. And then calculate the mean queue size at different filter constant EWMA (Exponential Weighted Moving Average). And then compared with the values of queue size calculated by RED gate within the network and determine the filter constants in accordance with the desired queue size for control the delay rate within the network, which is one of the most important parameters to assess the behavior of the traffic.