International Journal of Engineering Trends and Applications (IJETA) – Volume 4 Issue 3, May-Jun 2017 ISSN: 2393-9516 www.ijetajournal.org Page 30 TCP Congestion Control Mechanisms evaluation over Wired / Wireless Environments: A Survey K. Vasudha Rani [1] , Prof.P. Chenna Reddy [2] Research Scholar, RAYALASEEMA UNIVERSITY, Kurnool Director, Academic Audit, JNTU- Ananthapuramu, Ananthapuramu ABSTRACT This paper explores the attributes of TCP agents, namely “Tahoe, Reno, New Reno, Vegas, and Sack”, and computed their throughputs in the simulated environment utilizing the NS-2 test system by shifting different exhibition of parameters, for example, data transmission, simulation time and number of activity sources with a specific end goal to figure out which one of them is the best for which scenario. Simulation was performed to concentrate the execution of the Protocols as far as Bandwidth, Simulation time and number of activity sources. After the effects of the reproduction, demonstrated that Reno is great when bundle of the packets losses are little. New Reno is observed to be valuable when there are more packet losses of information. Sack is found to very poor at the point when different packet losses happen in a same congestion window, while Vegas Is the best while fluctuating data transmission inferable from the utilization of time based data transfer capacity estimation plans to control its congestion window. Keywords:- TCP, NS-2 I. INTRODUCTION The Transmission Control Protocol (TCP) is one of the superior transport Protocol under the transport layer of OSI, which support the World Wide Web, email and records exchange is a basic functionality over the internet. TCP provides better performance in terms of reliability, control flow of information, fast error recovery. TCP mainly uses to communicate host-to-host in the network. Once the host-to-host connection established exchange the data between the segments happen due to segments lost will happen by congestion control. The TCP functionality well defined and implemented in the RFC 793 [1]. Congestion control is a well-recognized problem in the complex computer networks. Congestion control causes two basic problems namely call small-packets, source-quench problems explained under RFC 896[2]. The small packet problem well fixed in the RFC like if any data input from the user and which not acknowledged previous packet in the queue, a new tcp segment data transmitted. Congestion control algorithms (“congestion avoidance, Slow start, Fast retransmit and recovery”) are well explained under RFC 5681 [3]. TCP have lot of issues when network running on high overload. In this paper we discussed congestion control between different TCP variants, such as “Tahoe, Reno, New Reno, Sack, Vegas, Westwood, Fack and Veno” and compared the performance in two stages.in the first stage we compared the performance evaluation between TCP Reno and RQB.in the second stage we assess the execution of four protocols of TCP that is Reno, New Reno, Sack and Vegas. II. TCP/IP Overview This paper will explore the execution correlations of these previously mentioned adaptations of TCP and discover which one is better in which cases. Generally, TCP/IP are the two different layer protocols which works seamlessly. TCP is the upper layer of which responsible for data transfer of a packet whereas IP is more responsible for logical addressing explained in the RFC 791. TCP standardization approved by IETF as RFC 793 [4] in 1981 and have wide range of features like “Interoperability, multi-vendor support, logical addressing, name resolution, routing, multiplexing/DE multiplexing, Flow control and error control”. IP standardized with RFC791 [5] in 1981. IP is belonging to network layer and have unique features, namely “unreliable, support for connectionless, data deliver without Acknowledgments”. 2.1. TCP Congestion Control Congestion control balance the network traffic in the computer networks in order to eliminate the congestion during over huge incoming data. This is gradually decreases throughput of the RESEARCH ARTICLE OPEN ACCESS