Transmission Control Protocol (TCP) congestion analysis of wired network using network simulator 2

The study investigates the performance of various Transmission Control Protocol (TCP) congestion control mechanisms—TCP Reno, TCP NewReno, and TCP Vegas—within wired network environments using the NS-2 network simulator. The primary research problem focuses on evaluating the efficiency, stability, and throughput of these TCP variants in response to network congestion, with the aim of identifying the most effective protocol for maintaining high performance under different network conditions. The methodology involves simulating a wired network environment in NS-2 and analyzing key performance metrics such as throughput, packet loss, and congestion window behavior for each TCP variant. The analysis reveals that TCP NewReno generally offers better performance compared to TCP Reno in scenarios involving multiple packet losses due to its improved fast recovery algorithm. TCP Vegas, on the other hand, demonstrates superior throughput and lower delays by proactively avoiding congestion through RTT-based adjustments, making it the most efficient protocol in stable network conditions. However, TCP NewReno proves more reliable than TCP Reno in environments with higher packet loss. The findings suggest that TCP Vegas is well-suited for networks requiring consistent performance and minimal delay, while TCP NewReno is preferable in networks where packet loss is a more frequent issue. The study concludes that TCP Vegas, with its congestion avoidance strategy, provides the best overall performance, though TCP NewReno remains a strong contender in more volatile network conditions, thereby offering a nuanced understanding of TCP protocol efficiency in wired networks.