CN117499314A - Network self-adaptive congestion control method based on average queue length change trend - Google Patents
Network self-adaptive congestion control method based on average queue length change trend Download PDFInfo
- Publication number
- CN117499314A CN117499314A CN202410005655.1A CN202410005655A CN117499314A CN 117499314 A CN117499314 A CN 117499314A CN 202410005655 A CN202410005655 A CN 202410005655A CN 117499314 A CN117499314 A CN 117499314A
- Authority
- CN
- China
- Prior art keywords
- queue length
- average queue
- average
- change
- probability
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 106
- 230000008859 change Effects 0.000 title claims abstract description 95
- 239000000872 buffer Substances 0.000 claims abstract description 42
- 238000013210 evaluation model Methods 0.000 claims abstract description 9
- 230000008569 process Effects 0.000 claims description 30
- 230000003044 adaptive effect Effects 0.000 claims description 19
- 230000001105 regulatory effect Effects 0.000 claims 1
- 101150070667 ureD gene Proteins 0.000 description 13
- 238000004088 simulation Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 6
- 230000007246 mechanism Effects 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 238000012888 cubic function Methods 0.000 description 3
- 238000012886 linear function Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000001934 delay Effects 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/12—Avoiding congestion; Recovering from congestion
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/32—Flow control; Congestion control by discarding or delaying data units, e.g. packets or frames
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a network self-adaptive congestion control method based on average queue length change trend, which comprises the following steps: initializing network parameters; waiting for packet data to arrive at a router buffer; calculating the average queue length in the router buffer zone at the current moment according to the average queue length evaluation model; calculating the first-order change rate and the second-order change rate of the average queue length in the router buffer zone at the current moment according to the average queue length change trend model; the real-time updating of the average queue length intermediate threshold value is carried out according to the first-order change rate and the second-order change rate of the average queue length; calculating packet data discarding probability; judging whether the packet data should enter a queue of a router buffer zone or be discarded according to the packet data discarding probability; the algorithm of the invention realizes the effective control of network congestion and shows better performance in the aspects of queue length, delay jitter, throughput and the like.
Description
Technical Field
The invention relates to the technical field of network congestion control, in particular to a network self-adaptive congestion control method based on average queue length change trend.
Background
With the popularization of intelligent terminals and the development of 5G mobile communication networks, the number of access nodes in an internet of things system is multiplied, a large amount of real-time multimedia services capable of generating huge data traffic are generated, and the probability of network congestion is greatly increased by high traffic load, so that the buffer area of a forwarding router overflows, and the network performance is greatly reduced, for example, the network packet loss rate is greatly increased, the throughput is greatly reduced, and the transmission delay fluctuation is large. Therefore, establishing an efficient congestion control mechanism in a network is critical to improving network performance.
At present, in the aspect of network congestion control, research at home and abroad still has some defects: as an effective network congestion control algorithm, active queue management (active queue management, AQM) algorithms mostly perform selective discarding of packet data based on the average queue length of a router buffer, so as to achieve the purpose of network congestion control, but this method cannot characterize the direction of change of the instantaneous queue length, and network congestion control efficiency needs to be improved; some algorithms perform selective discarding of packet data based on the average queue length change rate of the buffer, but cannot characterize the change rate of the instantaneous queue length, and cannot reflect the characteristics of network traffic; in addition, in most algorithms, the queue weight parameter is set to a fixed value, so that congestion is difficult to predict effectively, and network delay jitter, throughput and other performances are poor. Therefore, the network self-adaptive congestion control method based on the average queue length change trend is provided, and has important significance for improving the network congestion control efficiency.
Disclosure of Invention
The purpose of the invention is that: in the network adaptive congestion control method based on the average queue length variation trend, a new average queue length assessment model is provided, wherein the weight parameters can be adaptively adjusted according to the network state. And secondly, introducing a new parameter, namely an average queue length second-order change rate, and establishing an average queue length change trend model to predict the change trend of the future queue length. And finally, updating the intermediate threshold according to the first-order and second-order change rate information of the average queue length, and setting a discarding function by adopting a mode of combining a cubic function and a linear function so as to improve the comprehensive efficiency of network congestion control.
In order to achieve the above functions, the present invention designs a network adaptive congestion control method based on average queue length variation trend, and for a network including a plurality of routers, the following steps S1 to S8 are executed to adjust queues composed of packet data in buffers of the routers, thereby completing congestion control of the network:
step S1: initializing network parameters, wherein the network parameters comprise router parameter initial weights, maximum discarding probability, queue minimum threshold values, queue maximum threshold values and router buffer zone sizes;
step S2: waiting for packet data to reach each router buffer;
step S3: constructing an average queue length evaluation model, and calculating the average queue length in each router buffer zone at the current momentavg；
Step S4: constructing an average queue length change trend model, and calculating the first-order change rate of the average queue length in each router buffer zone at the current momentdavgSecond order rate of change of average queue lengthsdavg；
Step S5: respectively aiming at each router buffer zone at the current moment, according to the first-order change rate of the average queue lengthdavgSecond order rate of change of average queue lengthsdavgCalculating an average queue length intermediate threshold value at the current momentmid th And performs an average queue length intermediate thresholdmid th Is updated in real time;
step S6: based on average queue lengthavgFirst order rate of change of average queue lengthdavgSecond order rate of change of average queue lengthsdavgAverage queue length intermediate thresholdmid th Constructing a discarding probability model for calculating the discarding probability of the packet datap b According to the discarding probabilityp b Judging whether the packet data arrived in the step S2 should enter a queue of a router buffer area or be discarded;
step S7: judging the arrivalPacket data drop probabilityp b Whether 0 is, if so, the packet data enters a queue of a buffer area of the router, otherwise, the packet data is discarded;
step S8: and waiting for new packet data to reach the router buffer area, and repeatedly executing the steps S1-S7 to complete the congestion control of the network.
As a preferred technical scheme of the invention: the initialization setting of each network parameter in step S1 is as follows: router parameter initial weightsw q0 =0.002, maximum drop probabilitymax p =0.1, queue minimum thresholdmin th =24, maximum queue thresholdmax th Router buffer size=72W=120。
As a preferred technical scheme of the invention: the method is characterized in that an average queue length evaluation model constructed in the step S3 is as follows:
；
in the method, in the process of the invention,tindicating the current time of day and,avg(t) For the current momenttIs used to determine the average queue length of the (c),avg(t+1) istThe average queue length at time +1,inst(t+1) istThe instantaneous queue length at time +1,w q for router parameter weights, it calculates the following formula:
；
in the method, in the process of the invention,w q0 as an initial weight for the router parameters,avgfor the average queue length to be the same,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
As a preferred technical scheme of the invention: the average queue length change trend model constructed in the step S4 comprises two parts, namely an average queue length first-order change rate model and an average queue length second-order change rate model, wherein the average queue length first-order change rate model has the following formula:
；
in the method, in the process of the invention,davg(t) For the current momenttIs a first order rate of change of the average queue length,davg(t+1) istThe average queue length first order rate of change at time +1,inst(t) For the current momenttIs used to determine the instantaneous queue length of (1),inst(t+1) istThe instantaneous queue length at time +1,w q weighting router parameters;
the average queue length second order rate of change model is as follows:
；
in the method, in the process of the invention,sdavg(t) For the current momenttIs a second order rate of change of the average queue length,sdavg(t+1) istThe average queue length second order rate of change at time +1,inst(t-1) ist-instantaneous queue length at time 1.
As a preferred technical scheme of the invention: the average queue length intermediate threshold value described in step S5mid th Is calculated and updated as follows:
；
in the method, in the process of the invention,mid th the initial value is set to be 1/2%min th +max th ），min th For the minimum threshold value of the queue,max th as the maximum threshold value of the queue,davgto average the first order rate of change of the queue length,sdavgis the second order rate of change of the average queue length.
As a preferred technical scheme of the invention: the discarding probability model constructed in step S6 is as follows:
；
in the method, in the process of the invention,p b in order to discard the probability of a probability,p 1 as a function of the first drop probability,p 2 as a function of the second drop probability,p 3 as a third drop probability function,p 4 and is a fourth drop probability function.
As a preferred technical scheme of the invention: first discard probability function in discard probability modelp 1 The formula is as follows:
；
in the method, in the process of the invention,max p for the maximum probability of being dropped,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
As a preferred technical scheme of the invention: second discard probability function in discard probability modelp 2 The formula is as follows:
；
in the method, in the process of the invention,max p for the maximum probability of being dropped,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
As a preferred technical scheme of the invention: third discard probability function in discard probability modelp 3 The formula is as follows:
；
in the method, in the process of the invention,max p for the maximum probability of being dropped,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
As a preferred technical scheme of the invention: fourth loss in drop probability modelAbandon probability functionp 4 The formula is as follows:
；
in the method, in the process of the invention,max p for the maximum probability of being dropped,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
The beneficial effects are that: the advantages of the present invention over the prior art include:
(1) The weight self-adaptive mechanism is introduced, and the algorithm can select the response degree of the router to the input flow change according to the network state and the queue change, so that the average queue length evaluation model can more accurately reflect the network congestion condition;
(2) The flow characteristics can be reflected through the first-order and second-order change rate information of the average queue length, so that congestion control of the router is better realized;
(3) The self-adaptive adjustment of the intermediate threshold value of the router buffer zone is carried out through the intermediate threshold value updating model, so that the updating of the threshold value and the setting of the subsequent discarding probability can be more suitable for the change rule of the flow, thereby improving the performance of the algorithm such as throughput, queue length stability and the like;
(4) According to the average queue length and the first-order and second-order change rate values thereof, the setting of the discarding function is carried out by adopting a mode of combining a cubic function and a linear function, so that the stability of the queue can be effectively improved, and the end-to-end delay jitter is reduced;
(5) The network self-adaptive congestion control method based on the average queue length change trend is simple and efficient, can effectively improve the efficiency of network congestion control, and improves the network performance of the system.
Drawings
Fig. 1 is a flowchart of a network adaptive congestion control method based on average queue length variation trend according to an embodiment of the present invention;
FIG. 2 is a network topology provided in accordance with an embodiment of the present invention;
FIG. 3 is a simulation diagram of RED algorithm queue length provided in accordance with an embodiment of the present invention;
FIG. 4 is a simulation diagram of the URED algorithm queue length provided in accordance with an embodiment of the present invention;
FIG. 5 is a simulation diagram of a TRED algorithm queue length provided in accordance with an embodiment of the present invention;
FIG. 6 is a simulation diagram of the queue length of an agrED algorithm provided in accordance with an embodiment of the present invention;
FIG. 7 is a simulation diagram of the queue length of an AQMRD algorithm provided in accordance with an embodiment of the present invention;
FIG. 8 is a simulation diagram of a queue length provided in accordance with an embodiment of the invention;
FIG. 9 is a graph comparing end-to-end delays of the method of the present invention and RED, URED, TRED, agRED and the AQMRD algorithm provided in accordance with an embodiment of the present invention;
fig. 10 is a comparison of the end-to-end delay jitter and the method of the present invention and RED, URED, TRED, agRED and AQMRD algorithm provided in accordance with an embodiment of the present invention;
FIG. 11 is a graph comparing the throughput of the inventive method and RED, URED, TRED, agRED and AQMRD algorithm provided in accordance with an embodiment of the present invention;
fig. 12 is a comparison chart of the packet loss rate of the method and RED, URED, TRED, agRED and the AQMRD algorithm according to the present invention.
Detailed Description
The invention is further described below with reference to the accompanying drawings. The following examples are only for more clearly illustrating the technical aspects of the present invention, and are not intended to limit the scope of the present invention.
The network adaptive congestion control method based on average queue length variation trend provided by the embodiment of the invention, with reference to fig. 1, executes the following steps S1-S8 for a network including a plurality of routers, adjusts queues composed of packet data in each router buffer zone, and completes congestion control of the network:
step S1: initializing network parameters, wherein the network parameters comprise router parameter initial weights, maximum discarding probability, queue minimum threshold values, queue maximum threshold values and router buffer zone sizes;
the initialization setting of each network parameter is as follows: roadInitial weights of router parametersw q0 =0.002, maximum drop probabilitymax p =0.1, queue minimum thresholdmin th =24, maximum queue thresholdmax th Router buffer size=72W=120。
Step S2: waiting for packet data to reach each router buffer;
step S3: constructing an average queue length evaluation model, and calculating the average queue length in each router buffer zone at the current momentavg；
An exponentially weighted moving average (Exponential Weighted Moving Average, EWMA) is used as a low-pass filter to calculate the average queue length, and an adaptive mechanism is added to enable the model to select the degree of response of the router to the input traffic change according to the network state and the average queue change in the buffer, and the average queue length evaluation model constructed in step S3 is as follows:
；
in the method, in the process of the invention,tindicating the current time of day and,avg(t) For the current momenttIs used to determine the average queue length of the (c),avg(t+1) istThe average queue length at time +1,inst(t+1) istThe instantaneous queue length at time +1,w q for router parameter weights, it calculates the following formula:
；
in the method, in the process of the invention,w q0 as an initial weight for the router parameters,avgfor the average queue length to be the same,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
Step S4: constructing an average queue length change trend model, and calculating the first-order change rate of the average queue length in each router buffer zone at the current momentdavgSecond order rate of change of average queue lengthsdavg；
The average queue length change trend model constructed in the step S4 comprises two parts, namely an average queue length first-order change rate model and an average queue length second-order change rate model, wherein the average queue length first-order change rate model has the following formula:
；
in the method, in the process of the invention,davg(t) For the current momenttIs a first order rate of change of the average queue length,davg(t+1) istThe average queue length first order rate of change at time +1,inst(t) For the current momenttIs used to determine the instantaneous queue length of (1),inst(t+1) istThe instantaneous queue length at time +1,w q weighting router parameters;
the average queue length second order rate of change model is as follows:
；
in the method, in the process of the invention,sdavg(t) For the current momenttIs a second order rate of change of the average queue length,sdavg(t+1) istThe average queue length second order rate of change at time +1,inst(t-1) ist-instantaneous queue length at time 1.
Step S5: respectively aiming at each router buffer zone at the current moment, according to the first-order change rate of the average queue lengthdavgSecond order rate of change of average queue lengthsdavgCalculating an average queue length intermediate threshold value at the current momentmid th And performs an average queue length intermediate thresholdmid th Is updated in real time;
the average queue length intermediate threshold value described in step S5mid th Is calculated and updated as follows:
；
in the method, in the process of the invention,mid th the initial value is set to be 1/2%min th +max th ），min th For the minimum threshold value of the queue,max th as the maximum threshold value of the queue,davgto average the first order rate of change of the queue length,sdavgis the second order rate of change of the average queue length.
Step S6: based on average queue lengthavgFirst order rate of change of average queue lengthdavgSecond order rate of change of average queue lengthsdavgAverage queue length intermediate thresholdmid th Constructing a discarding probability model for calculating the discarding probability of the packet datap b According to the discarding probabilityp b Judging whether the packet data arrived in the step S2 should enter a queue of a router buffer area or be discarded;
the discarding probability model constructed in step S6 is as follows:
；
in the method, in the process of the invention,p b in order to discard the probability of a probability,p 1 as a function of the first drop probability,p 2 as a function of the second drop probability,p 3 as a third drop probability function,p 4 and is a fourth drop probability function.
When the first order rate of change of the average queue lengthdavgSecond order rate of change of average queue lengthsdavgAll positive values, the first discarding probability function in the corresponding discarding probability modelp 1 The formula is as follows:
；
in the method, in the process of the invention,max p for the maximum probability of being dropped,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
When the first order rate of change of the average queue lengthdavgPositive value, average queueSecond order rate of change of lengthsdavgWhen the value is not positive, the second discarding probability function in the corresponding discarding probability modelp 2 The formula is as follows:
；
in the method, in the process of the invention,max p for the maximum probability of being dropped,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
When the first order rate of change of the average queue lengthdavgNon-positive value, second order rate of change of average queue lengthsdavgIn the case of positive values, the third discarding probability function in the corresponding discarding probability modelp 3 The formula is as follows:
；
in the method, in the process of the invention,max p for the maximum probability of being dropped,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
When the first order rate of change of the average queue lengthdavgSecond order rate of change of average queue lengthsdavgAll non-positive, the fourth discard probability function in the corresponding discard probability modelp 4 The formula is as follows:
；
in the method, in the process of the invention,max p for the maximum probability of being dropped,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
Step S7: determining a drop probability of arriving packet datap b Whether 0 is, if so, the packet data enters a queue of a buffer area of the router, otherwise, the packet data is discarded;
step S8: and waiting for new packet data to reach the router buffer area, and repeatedly executing the steps S1-S7 to complete the congestion control of the network.
The following is the network self-adaptive congestion control method based on the average queue length change trend designed by the invention, which is based on the performance analysis of the actual application scene:
simulation scenarios of 25, 50, 75 and 100 FTP sources were simulated, respectively, corresponding to low, medium, high and very high traffic load conditions, respectively. On the basis, the validity of the method is verified, and verified performance indexes comprise queue length, end-to-end time delay, time delay jitter, throughput and packet loss rate, and the performance indexes are compared with five algorithms RED, URED, TRED, AQMRD and agRED.
FIG. 2 is a network simulation topology, S 1 To S n D is FTP transmission source 1 To D n Is the destination end. From source to router R 1 Each link of (2) has a capacity of 50Mbps and a propagation delay of 10ms. The bottleneck link has a capacity of 10Mbps and a propagation delay of 40ms. Slave router R 2 Each link to the destination has a capacity of 50Mbps and a propagation delay of 10ms.
Fig. 3-8 are simulated views of the queue length of 100 FTP sources RED, URED, TRED, agRED, AQMRD and the method of the present invention, respectively. The method can maintain the queue length of the buffer area at a lower level, and the stability of the buffer area in terms of average queue length and instantaneous queue length is the highest, so that the jitter of the queue length is effectively reduced.
Fig. 9 is a graph comparing the end-to-end delays of the inventive methods (CT-RED) and RED, URED, TRED, agRED and AQMRD algorithms under different traffic loads. Under the condition of low flow load, the method has the effect inferior to that of the agRED algorithm and is superior to RED, URED, TRED and AQMRD algorithms. The method of the invention can obtain good effect under the condition of medium and high flow load, and is superior to RED, URED and TRED algorithms because the end-to-end time delay performance is affected by the length of the queue, and the lower the length of the queue is, the lower the time delay is. The method of the invention adopts the self-adaptive discarding strategy based on the change trend of the queue length, and can keep the queue length at a lower level, thereby effectively reducing the end-to-end time delay.
Fig. 10 is a graph comparing end-to-end delay jitter of the inventive methods (CT-RED) and RED, URED, TRED, agRED and AQMRD algorithms under different traffic loads. The method can be found that the delay jitter is obviously reduced under the conditions of low, medium and high traffic load, and is superior to other five algorithms. This is because the delay jitter performance is affected by the stability of the queue length, the more stable the queue length, the lower the delay jitter. The method adopts the self-adaptive discarding strategy based on the change trend of the queue length, and can effectively improve the stability of the queue length, thereby effectively reducing the delay jitter.
Fig. 11 is a graph comparing the throughput of the inventive methods (CT-RED) and RED, URED, TRED, agRED and AQMRD algorithms at different traffic loads. The network throughput using the method of the present invention was found to be highest. The method has more remarkable lifting effect under low and medium flow load, and slightly improves the lifting effect under high flow load compared with other five algorithms. The method can effectively improve the network throughput no matter what traffic load is.
Fig. 12 is a graph comparing the packet loss rates of the methods of the present invention (CT-RED) and RED, URED, TRED, agRED and AQMRD algorithms under different traffic loads. Under different flow loads, the method can obtain lower packet loss rate, and particularly, when the FTP source number is 25 and 75, the method can obtain the best effect in six algorithms. The method can ensure lower and stable queue length, and can ensure that the buffer zone reserves more available space to absorb burst traffic, thereby effectively reducing the packet loss rate.
In summary, the present invention provides a network adaptive congestion control method based on an average queue length variation trend, and firstly introduces an adaptive mechanism of queue weights, so that the adaptive mechanism can perform adaptive adjustment according to a network state, and a new average queue length evaluation model is provided based on the adaptive mechanism. And secondly, introducing a new parameter, namely an average queue length second-order change rate, and establishing an average queue length change trend model to predict the change trend of the future queue length. And finally, updating the intermediate threshold according to the first-order and second-order change rate information of the average queue length, and setting a discarding function by adopting a mode of combining a cubic function and a linear function so as to improve the network performance index. The algorithm of the invention realizes the effective control of network congestion and shows better performance in the aspects of queue length, delay jitter, throughput and the like.
The embodiments of the present invention have been described in detail with reference to the drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the spirit of the present invention.
Claims (10)
1. The network self-adaptive congestion control method based on the average queue length change trend is characterized in that for a network comprising a plurality of routers, the following steps S1-S8 are executed, the queues formed by packet data in the buffer areas of the routers are regulated, and the congestion control of the network is completed:
step S1: initializing network parameters, wherein the network parameters comprise router parameter initial weights, maximum discarding probability, queue minimum threshold values, queue maximum threshold values and router buffer zone sizes;
step S2: waiting for packet data to reach each router buffer;
step S3: constructing an average queue length evaluation model, and calculating the average queue length in each router buffer zone at the current momentavg；
Step S4: constructing an average queue length change trend model, and calculating the first-order change rate of the average queue length in each router buffer zone at the current momentdavgSecond order rate of change of average queue lengthsdavg；
Step S5: respectively aiming at each router buffer zone at the current moment, according to the first-order change rate of the average queue lengthdavgSecond order rate of change of average queue lengthsdavgCalculating an average queue length intermediate threshold value at the current momentmid th And performs an average queue length intermediate thresholdmid th Is updated in real time;
step S6: base groupAt average queue lengthavgFirst order rate of change of average queue lengthdavgSecond order rate of change of average queue lengthsdavgAverage queue length intermediate thresholdmid th Constructing a discarding probability model for calculating the discarding probability of the packet datap b According to the discarding probabilityp b Judging whether the packet data arrived in the step S2 should enter a queue of a router buffer area or be discarded;
step S7: determining a drop probability of arriving packet datap b Whether 0 is, if so, the packet data enters a queue of a buffer area of the router, otherwise, the packet data is discarded;
step S8: and waiting for new packet data to reach the router buffer area, and repeatedly executing the steps S1-S7 to complete the congestion control of the network.
2. The network adaptive congestion control method according to claim 1, wherein the initialization of each network parameter in step S1 is set to: router parameter initial weight w q0 =0.002, maximum discard probability max p =0.1, queue minimum threshold min th =24, maximum queue threshold max th Router buffer size w=120.
3. The network adaptive congestion control method based on the average queue length variation trend according to claim 1, wherein the average queue length evaluation model constructed in step S3 has the following formula:
；
in the method, in the process of the invention,tindicating the current time of day and,avg(t) For the current momenttIs used to determine the average queue length of the (c),avg(t+1) istThe average queue length at time +1,inst(t+1) istThe instantaneous queue length at time +1,w q is a routerParameter weights calculated as:
；
in the method, in the process of the invention,w q0 as an initial weight for the router parameters,avgfor the average queue length to be the same,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
4. The network adaptive congestion control method according to claim 1, wherein the average queue length change trend model constructed in step S4 includes two parts, namely an average queue length first-order change rate model and an average queue length second-order change rate model, wherein the average queue length first-order change rate model has the following formula:
；
in the method, in the process of the invention,davg(t) For the current momenttIs a first order rate of change of the average queue length,davg(t+1) istThe average queue length first order rate of change at time +1,inst(t) For the current momenttIs used to determine the instantaneous queue length of (1),inst(t+1) istThe instantaneous queue length at time +1,w q weighting router parameters;
the average queue length second order rate of change model is as follows:
；
in the method, in the process of the invention,sdavg(t) For the current momenttIs a second order rate of change of the average queue length,sdavg(t+1) istThe average queue length second order rate of change at time +1,inst(t-1) ist-instantaneous queue length at time 1.
5. The network adaptive congestion control method according to claim 1, wherein the average queue length median threshold value in step S5 ismid th Is calculated and updated as follows:
；
in the method, in the process of the invention,mid th the initial value is set to be 1/2%min th +max th ），min th For the minimum threshold value of the queue,max th as the maximum threshold value of the queue,davgto average the first order rate of change of the queue length,sdavgis the second order rate of change of the average queue length.
6. The network adaptive congestion control method based on the average queue length variation trend according to claim 1, wherein the drop probability model constructed in step S6 is as follows:
；
in the method, in the process of the invention,p b in order to discard the probability of a probability,p 1 as a function of the first drop probability,p 2 as a function of the second drop probability,p 3 as a third drop probability function,p 4 and is a fourth drop probability function.
7. The network adaptive congestion control method based on average queue length variation trend according to claim 6, wherein the first drop probability function in the drop probability modelp 1 The formula is as follows:
；
in the method, in the process of the invention,max p for the maximum probability of being dropped,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
8. The network adaptive congestion control method based on average queue length variation trend according to claim 6, wherein the second drop probability function in the drop probability modelp 2 The formula is as follows:
；
in the method, in the process of the invention,max p for the maximum probability of being dropped,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
9. The network adaptive congestion control method based on average queue length variation trend according to claim 6, wherein the third drop probability function in the drop probability modelp 3 The formula is as follows:
；
in the method, in the process of the invention,max p for the maximum probability of being dropped,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
10. The network adaptive congestion control method based on average queue length variation trend according to claim 6, wherein a fourth drop probability function in the drop probability modelp 4 The formula is as follows:
；
in the method, in the process of the invention,max p for the maximum probability of being dropped,min th for the minimum threshold value of the queue,max th is the maximum threshold for the queue.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410005655.1A CN117499314B (en) | 2024-01-03 | 2024-01-03 | Network self-adaptive congestion control method based on average queue length change trend |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410005655.1A CN117499314B (en) | 2024-01-03 | 2024-01-03 | Network self-adaptive congestion control method based on average queue length change trend |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117499314A true CN117499314A (en) | 2024-02-02 |
| CN117499314B CN117499314B (en) | 2024-03-19 |
Family
ID=89674847
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410005655.1A Active CN117499314B (en) | 2024-01-03 | 2024-01-03 | Network self-adaptive congestion control method based on average queue length change trend |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117499314B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101958833A (en) * | 2010-09-20 | 2011-01-26 | 云南省科学技术情报研究院 | RED-based network congestion control algorithm |
| US20150244639A1 (en) * | 2014-02-24 | 2015-08-27 | Freescale Semiconductor, Inc. | Method and apparatus for deriving a packet select probability value |
| CN111314240A (en) * | 2018-12-12 | 2020-06-19 | 深圳市中兴微电子技术有限公司 | Congestion control method and device, network equipment and storage medium |
| CN112235813A (en) * | 2020-10-14 | 2021-01-15 | 东北大学秦皇岛分校 | Multi-bottleneck network active queue optimization method based on distributed average tracking algorithm |
| CN114244711A (en) * | 2022-02-24 | 2022-03-25 | 南京信息工程大学 | Self-adaptive active queue management method based on average queue length and change rate thereof |
| CN116389375A (en) * | 2023-03-17 | 2023-07-04 | 华中科技大学 | Network queue management method, device and router for live video stream |
-
2024
- 2024-01-03 CN CN202410005655.1A patent/CN117499314B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101958833A (en) * | 2010-09-20 | 2011-01-26 | 云南省科学技术情报研究院 | RED-based network congestion control algorithm |
| US20150244639A1 (en) * | 2014-02-24 | 2015-08-27 | Freescale Semiconductor, Inc. | Method and apparatus for deriving a packet select probability value |
| CN111314240A (en) * | 2018-12-12 | 2020-06-19 | 深圳市中兴微电子技术有限公司 | Congestion control method and device, network equipment and storage medium |
| CN112235813A (en) * | 2020-10-14 | 2021-01-15 | 东北大学秦皇岛分校 | Multi-bottleneck network active queue optimization method based on distributed average tracking algorithm |
| CN114244711A (en) * | 2022-02-24 | 2022-03-25 | 南京信息工程大学 | Self-adaptive active queue management method based on average queue length and change rate thereof |
| CN116389375A (en) * | 2023-03-17 | 2023-07-04 | 华中科技大学 | Network queue management method, device and router for live video stream |
Non-Patent Citations (1)
| Title |
|---|
| 刘治;章云;: "基于模糊参考模型机制的网络自适应拥塞控制", 计算机工程, no. 07, 5 April 2008 (2008-04-05) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117499314B (en) | 2024-03-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107634911B (en) | Adaptive congestion control method based on deep learning in information center network | |
| CN107171842B (en) | Multipath transmission protocol congestion control method based on reinforcement learning | |
| CN110730469B (en) | Method for predicting bandwidth based on extended Kalman wireless network and congestion control thereof | |
| CN111526036B (en) | Short flow real-time optimization method, system and network transmission terminal | |
| Su et al. | QRED: A Q-learning-based active queue management scheme | |
| CN114143264B (en) | Flow scheduling method based on reinforcement learning under SRv network | |
| CN103312606A (en) | Wireless Mesh network routing based on multi-path concurrence and method thereof | |
| CN114785743B (en) | Active queue management algorithm for enhancing RED stability | |
| CN114244711A (en) | Self-adaptive active queue management method based on average queue length and change rate thereof | |
| Pan et al. | A novel active queue management algorithm based on average queue length change rate | |
| CN110719617B (en) | Q routing method based on arc tangent learning rate factor | |
| WO2024125664A1 (en) | Adaptive routing joint prediction method for flight ad hoc network | |
| Mahawish et al. | Survey on: A variety of AQM algorithm schemas and intelligent techniques developed for congestion control | |
| Deng et al. | An adaptive packet aggregation algorithm for wireless networks | |
| CN117499314B (en) | Network self-adaptive congestion control method based on average queue length change trend | |
| Sun et al. | An adaptive neuron AQM for a stable internet | |
| CN110943927B (en) | Named data network transmission control method, device and equipment based on time delay management | |
| Li et al. | Delay-based congestion control for multipath TCP in heterogeneous wireless networks | |
| CN115118648B (en) | VANET network self-adaptive routing protocol selection method and computer program product | |
| Dai et al. | Towards AI-Driven on-Demand Routing in 6G Wide-Area Networks | |
| Chen et al. | Adaptive packet length strategy to minimize end-to-end latency for time-varying channels | |
| CN118381765B (en) | Lossless network congestion control method, lossless network congestion control device, lossless network congestion control equipment, lossless network congestion control medium and switch system | |
| Pinheiro et al. | A fuzzy queue-aware routing approach for wireless mesh networks | |
| CN118381581B (en) | Tactical communication network service framing method based on deep reinforcement learning | |
| Zhang et al. | Adaptive fast TCP |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |