الفهرس 
Chapter 1: IntroductionOnly 14 pages are availabe for public view
 |  | from | 32 |  |  |
| | | from | 32 | | |
Abstract
This research work presents a new strategy, called Dynamic Increase Dynamic
Decrease (DIDD), to enhance most of the existing end-to-end congestion control
mechanisms. The main idea is to adjust the congestion window (cwnd) size
dynamically according to the available bandwidth of the network to speed up the
use of the available capacity of the network. In the proposed strategy, instead of
increasing the cwnd size linearly by the Additive Increase Multiplicative Decrease
(AIMD) strategy, the cwnd is increased dynamically according to the available
bandwidth of the network. In addition, instead of decreasing the cwnd to half of
its size as congestion happens, the cwnd is decreased to the average value
between the latest value of the cwnd that was used effectively without losses and
the last active cwnd that causes the losses. The proposed approach is coded in
C++ and implemented in the TCP Tahoe, Reno, Newreno, Vegas and SACK. The
performance is evaluated by using the network simulator NS-٢ considering a
realistic network topology generated by the GT-ITM.
The simulation results show that the proposed DIDD strategy outperforms the
AIMD algorithm in terms of certain metrics which are throughput, delay, packet
loss and CWND, where the throughput increases and the delay time decreases.
Also, the results show that, when the proposed strategy is implemented with the
original TCP Tahoe and the TCP Newreno, the network throughput, delay and
CWND are improved.