Optimization of Veno parameter based on stochastic approximation: OVeno

Abstract

Transmission Control Protocol (TCP) ensures reliable communication between source and destination. However, TCP’s performance is significantly affected by congestion control, which regulates data flow and maintains optimal transfer rates while preventing packet loss. Congestion control is managed by the router that is network-assisted, and another approach is controlled by TCP, which is end-to-end congestion control. A popular TCP congestion control algorithm, Veno has the advantage of distinguishing between random loss and congestion loss. Veno serves as the base algorithm of TCP, performing well even in the presence of a wireless environment. These important features motivate us to redesign the Veno. This paper proposes a modified multiplicative decrease phase of the TCP Veno algorithm based on the stochastic approximation that is used to determine the optimal value of parameters. The performance evaluation of the proposed algorithm is evaluated with recent existing algorithms. The experimental result shows that the proposed approach improves the performance of the existing standard algorithms in terms of loss rate, throughput, and delay. Our proposed algorithm improves throughput by 143%, 131%, 66%, and 42% compared to Reno, Compound TCP, CUBIC, and Veno, respectively. We have also tested the efficacy of our proposed algorithm in the wireless environment.