Congestion control is an effective method to solve network congestion. Traditional loss-based congestion control algorithms consider loss as a sign of network congestion. In 2016, Google proposed the Bottleneck Bandwidth and Round-trip propagation time (BBR) congestion control algorithm. This is a congestion-based congestion control algorithm which can achieve low latency while maintaining high throughput. However, periodic probing bandwidth phase of BBR suffers from slow convergence and high retransmission in practice. The BBR-APG algorithm proposed in this paper can solve the above problems by adaptively adjusting the pacing gain according to the bottleneck buffer size and network congestion. Experimental results show that BBR-APG improves the convergence speed by 21% and reduces retransmission by up to 73% compared with BBR.
Improved Convergence and Low Retransmission of BBR Congestion Control Algorithm Based on Adaptive Pacing Gain
2022-11-24
1903258 byte
Conference paper
Electronic Resource
English
AVLC retransmission mechanism simulation based on adaptive channel estimation algorithm
British Library Online Contents | 2015
|Algorithm improved congestion propagation analysis method
European Patent Office | 2022
|