GMC-158 Evaluating TCP Protocol Performance in Cloud Environments

This research investigates the performance of four different TCP algorithms—BBR, Reno, Vegas, and Cubic in a high-latency and congested condition within a cloud-based environment using EC2 instances and Mininet for network simulation. The study aims to evaluate the throughput and congestion window (cwnd) behavior of each algorithm under various network conditions to identify their strengths and weaknesses. By analyzing the performance metrics across different TCP algorithms, we provide insights into their suitability for cloud infrastructure, contributing to optimized network protocol choices for cloud-based applications and services. The results offer valuable guidance for enhancing network performance in dynamic cloud environments.