A Protocol-Independent Dual Congestion Control Method

Abstract

The invention belongs to the field of network technology, in particular to a protocol-independent dual congestion control method. The method of the present invention includes: determining the network congestion probability Q under the single action of the traditional congestion control algorithm according to the current congestion state parameters of the network n (t); according to Q n (t) and P n (m n ) to determine the network congestion probability P under the action of the congestion control algorithm based on wavelength change alone n (m n ); Q n (t) and P n (m n ) jointly control the congestion state of the network. The advantage of the present invention is that: the variability of the wavelength will reduce the congestion probability of the network, and under certain conditions, the overall throughput and bandwidth utilization of the network can not be affected or improved; the traditional network congestion control algorithm only needs to do very little Improved to fit P n (m n ), ensuring the flexibility and scalability of the method. With the invention, the traditional congestion control algorithm can be optimized, and the network throughput rate and bandwidth utilization rate can be improved.

Description

technical field [0001] The invention belongs to the field of network technology, and in particular relates to a network congestion control method. Background technique [0002] Congestion control has always been a research hotspot in the network field. The main reason is that the state of the network is constantly changing, and the outdated congestion control algorithm is not suitable for the current network environment, or is not optimal, so it needs to be improved. Traditional congestion control algorithms, such as TCP reno, TCP new reno, HSTCP, HTCP, UDT, BIC, CUBIC, BBR, QUIC, etc. are all adjusted based on packet loss, RTT or available bandwidth. These algorithms rely too much on the process parameters in congestion, such as NAK (packet loss feedback signal), ACK (response signal), RTT (round trip delay), Bandwidth, etc., and many parameters are difficult to accurately measure, coupled with network status It is ever-changing, so it is difficult for the traditional netw...

AI Technical Summary

Problems solved by technology

These algorithms rely too much on the process parameters in congestion, such as NAK (packet loss feedback signal), ACK (response signal), RTT (round trip delay), Bandwidth, etc., and many parameters are difficult to accurately measure, coupled with network status Ever-changing, so the traditional network congestion control algorithm is difficult to achieve the global optimal

Especially in a WAN with a large BDP, bandwidth utilization and throughput need to be improved

Images