Optimized Congestion Control Algorithm for QUIC in Wireless Networks: CubicBytes-N Algorithm

Xiang Xiao, Long Chen, Ming Zhao

Abstract: With the explosive growth of mobile devices, wireless networks play a crucial role in people’s daily lives. This article addresses the issue of network congestion control in wireless environments and proposes an improved algorithm based on the quick UDP Internet connection (QUIC) protocol, namely the CubicBytes-N algorithm. By introducing a bandwidth estimation mechanism, this algorithm successfully distinguishes congestion-induced packet loss from random packet loss, addressing the blind window reduction issue in the traditional CubicBytes algorithms. Afterward, the congestion level is set based on the estimation results, and the slow start threshold is further adjusted to make the algorithm more robust. Through a series of experiments conducted in the network simulator-3 (NS3) simulation environment, we validated the performance of the CubicBytes-N algorithm in wireless networks with random packet loss rates ranging from 0% to 5%. The experimental results demonstrate that the CubicBytes-N algorithm not only excels in improving network bandwidth utilization but also avoids preempting the bandwidth required by the traditional CubicBytes algorithm in high packet loss rate environments, ensuring fairness. Overall, the CubicBytes-N algorithm provides an effective congestion control solution for the application of the QUIC protocol in wireless networks.