RFBBR: A Rtt Faireness Awared Algorithm Based on BBR

The popularization and application of the Internet of Things (IoT) have produced massive and multi-type data. The information in the network grows explosively, and is transmitted in many forms in the network. The traditional TCP congestion control algorithms are not suitable for long fat pipes, and the transmission rates are very low in some scenarios such as dedicated line transmission. The TCP-BBR algorithm is a TCP congestion control protocol proposed by Google for transmission on the long fat pipeline to improve the transmission throughput and reduce the delay of data transmission. Through simulations, it is found that the competition between TCP-BBR streams with different RTTs will cause serious RTT unfairness problem. The TCP-BBR data streams with small RTTs are always suppressed by the data streams with large RTTs. To solve this problem, the improved BBR algorithm RFBBR is proposed. The RFBBR can guarantee the faireness of BBR streams with different RTTs. The simulation results show that RFBBR can significantly improve fairness compared with BBR and BBQ in the wireless network spindle topology scenarios.