Performance evaluation of TCP congestion control mechanisms using NS-2

This paper investigates the characteristics of Tahoe, Reno, New Reno, Vegas, and Sack, and calculated their throughputs in simulated environment using NS-2 simulator by varying various performances parameters such as bandwidth, simulation time and number of traffic sources in order to figure out which one of them is the best for which scenario. Simulation was performed to study the performance of the protocols in terms of Bandwidth, simulation time and number of traffic sources. The results of the simulation showed that Reno is good when packet losses are small. New Reno is found to be useful when there are more losses of data. Sack is found to robust when multiple losses occur in one congestion window, while Vegas is the best when varying bandwidth owing to the use of time based bandwidth estimation schemes to control its congestion window.