TCP Congestion Control Mechanisms – an Experimental Analysis and Comparison

Session

Computer Science and Communication Engineering

Description

The incompatibility between traffic load and network processing/communication capacities brings networks to the congestion state. This typically results in the overall network performance deterioration. Various approaches to congestion control have been proposed so far, both at the network and transport layer of networking architecture. However, there is still no such an algorithm that would overperform all the other algorithms in all of the networking scenarios. This study aims to analyze and compare the most popular traditional TCP approaches to congestion control. New Reno, Cubic, and modified SACK TCP were deployed in an OPNET- based simulated topology with varying packet delays and packet loss rate. Although expected (to some extent), SACK version has shown better performances in networks with lower packet delays. Cubic, however, performed surprisingly under the expectations as compared to two other versions, while New Reno has shown consistency along various different scenarios and implementations.

Keywords:

Congestion Control, TCP, CUBIC, New Reno, SACK, OPNET

Proceedings Editor

Edmond Hajrizi

ISBN

978-9951-550-47-5

Location

UBT Kampus, Lipjan

Start Date

30-10-2021 12:00 AM

End Date

30-10-2021 12:00 AM

DOI

10.33107/ubt-ic.2021.372

This document is currently not available here.

Share

COinS
 
Oct 30th, 12:00 AM Oct 30th, 12:00 AM

TCP Congestion Control Mechanisms – an Experimental Analysis and Comparison

UBT Kampus, Lipjan

The incompatibility between traffic load and network processing/communication capacities brings networks to the congestion state. This typically results in the overall network performance deterioration. Various approaches to congestion control have been proposed so far, both at the network and transport layer of networking architecture. However, there is still no such an algorithm that would overperform all the other algorithms in all of the networking scenarios. This study aims to analyze and compare the most popular traditional TCP approaches to congestion control. New Reno, Cubic, and modified SACK TCP were deployed in an OPNET- based simulated topology with varying packet delays and packet loss rate. Although expected (to some extent), SACK version has shown better performances in networks with lower packet delays. Cubic, however, performed surprisingly under the expectations as compared to two other versions, while New Reno has shown consistency along various different scenarios and implementations.