An Enhanced TCP Vegas Congestion Control Mechanism for GEO Satellite Return Links

In this paper, we propose an an enhanced TCP Vegas algorithm for the high latency and high BER of the GEO satellite return link. To optimize the congestion window adjustment strategy, the algorithm dynamically evaluates the network bandwidth and RTT variations: firstly, the initial congestion window is increased to accelerate the bandwidth detection, and the weighted average method is used to dynamically update the BaseRTT to inhibit link jitter; secondly, the dynamic adjustment formula of the window is designed based on the bandwidth estimation, which shortens the convergence time; and at the same time, the mechanism of the segmentation parameter of the number of concurrent streams adaption is introduced, and the threshold is dynamically adjusted to balance the throughput and fairness. throughput and fairness. The SNS3 simulation results show that, compared to traditional Vegas and Westwood algorithms, the proposed algorithm not only improves the throughput and fairness significantly in high latency and high BER environments, and also reduces the transmission delay effectively.