Method and device for adjusting contention window value adaptively

Abstract

A method and device for adjusting contention window value adaptively are disclosed. The method comprises the following steps: the number n of stations participated in a contention currently is measured; according to the number n of stations participated in the contention currently, a practical collision probability Pc of a current contention window value CWold and optimal collision probability Pcopt, under the maximum network throughput are acquired; when a data frame is sent successfully, an initial contention window value CWinit is adjusted according to the practical collision probability Pc and optimal collision, probability Pcopt; when a data frame is sent unsuccessfully, a retransmission contention window value CWnew adjusted according to said practical collision probability Pc. The above solution can better reflect network collision situation, adaptively adjust an initial contention window value and retransmission contention window value jointly.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS/INCORPORATION BY REFERENCE

This application is a 371 of International Patent Application No, PCT/CN2013/082181, filed Aug. 23, 2013, entitled “CONTENTION WINDOW VALUE ADAPTIVE ADJUSTING METHOD AND EQUIPMENT”, which claims priority to Chinese Patent Application No. 201210325365.2, filed Sep. 5, 2012, entitled “SELF-ADAPTIVE CONTENTION WINDOW VALUE ADJUSTMENT METHOD AND DEVICE”. The above-identified applications are hereby incorporated herein by reference in their entirety.

TECHNICAL FIELD

The disclosure relates to the technical field of network communications, and in particular to a method and device for adjusting a contention window value adaptively.

BACKGROUND

An access protocol based on contention in 802.11 (being a wireless local area network standard initially formulated by IEEE) standard constitutes the basis of 802.11 media access control (referred to as MAC) protocol, and the standard specifies a distributed coordination function (referred to as DCF) for a contention phase for the media access control layer (MAC) of the WLAN. The DCF uses the technology of Carrier Sense Multiple Access with Collision Detection (referred to as CSMA/CA) to detect a media state by means of physical and virtual carrier sense functions, and each terminal may independently determine to access a channel and enter into a back-off process when the access is failed so as to reconnect the channel, thereby providing a more flexible wireless communication manner.

In the wireless communication field, each terminal may independently determine to enter into the channel and enter into the back-off process when the access is failed so as to reconnect the channel. However, only one shared channel exists in the wireless local network, and therefore, a station with services to be sent needs to perform channel contention; when stations participated in channel contention are sensed to be idle, duration of a distributed coordination inter frame space (DIFS) is delayed, and a random back-off duration is further waited, and if the channel is still idle at the moment, then the channel is accessed. The random back-off duration is also called a contention window (CW), and in order to reduce the collision probability; the size of an optimal contention window depends on the quantity of stations participating in channel contention in a channel at the same time slot. If the number of stations participated in the contention is less, then the collision probability is low, and now a smaller contention window may be used so as to reduce delay; and if the number of stations participated in the contention is much, then the collision probability is high, and now it is suitable to use a larger contention window so as to reduce collision and improve throughput of the whole network.

In a wireless local area network standard, a back-off method of binary index in a unit of time slot is usually used to acquire a contention window value, and CW is a random integer selected from equal distribution on the section [0, CW−1], and CW ε [CW_min, CW_max], and during retransmission for an ith time, CW(i)=min[2ⁱ·CW_min, CW_max], and min( ) represents taking a minimum value within brackets. The CW is reset to be CW_min after being sent successfully every time. However, in the case of network congestion, if the CW is reset to be CW_min after successful sending, it would easily cause collision and retransmission, and reduce network throughput; and in the case of network congestion being relieved, the transmission would be doubled and the number of network idle time slots would be increased, and unnecessary delay would be introduced and the network throughput would be reduced.

SUMMARY

The main object of the embodiments of the disclosure is to provide a method and device for adjusting a contention window value adaptively so as to reduce network delay and improve network throughput.

The embodiments of the disclosure propose a method for adjusting a contention window value adaptively, and the method comprises the following steps:

measuring the number n of stations participated in a contention currently;

acquiring, according to the number n of stations participated in a contention currently, the practical collision probability P_c of the current contention window value CW_old and optimal collision probability P_copt under the maximum network throughput;

adjusting, when a data frame is transmitted successfully, initial contention window value CW_init according to the practical collision probability P_c and optimal collision probability P_copt; and

adjusting, when a data frame is transmitted unsuccessfully, a retransmission contention window value CW_new according to the practical collision probability P_c.

Preferably, acquiring, according to the number n of stations participated in the contention currently, the practical collision probability P_c of the current contention window value CW_old and optimal collision probability P_copt under the maximum network throughput specifically comprises:

when the number of stations participated in the contention currently is n and current contention window value is CW_old, determining the practical collision probability P_c=1−(1−τ)ⁿ⁻¹[1+(n−1)τ], where τ is a sending probability of a station when a time slot of a channel begins and the current contention window value is CW_old; and

when the number of stations participated in the contention currently is n and network throughput is maximum, determining the optimal collision probability P_copt=1−(1−τ_opt)ⁿ⁻¹[1+(n−1)τ_opt], where τ_opt is a sending probability of a station when a time slot of a channel begins and network throughput is maximum.

Preferably, adjusting the initial contention window value CW_init according to the practical collision probability P_c and optimal collision probability P_copt specifically comprises:

comparing the practical collision probability P_c with a lower limit, which is P_cmax=P_copt−D_L, of a collision probability threshold value and an upper limit, which is P_cmin=D_L, of the collision probability threshold value, wherein D_L is an interactive tolerance threshold;

when P_c<_cmin, determining a current network to be idle, and reducing the initial contention window value CW_init=min[CW_old−CW_step1, CW_max], where CW_step1 is a step length of a first initial contention window value;

when P_c>P_cmax, determining a current network to be congested, and increasing the initial contention window value CW_init=min[CW_old+CW_step2, CW_max], where CW_step2 is a step length of a second initial contention window value;

when P_cmin≦P_c≦P_cmax, determining a current network to be good, and maintaining the initial contention window value CW_init=min[CW_old, CW_max].

Preferably, determining a current network to be idle, and reducing the initial contention window value CW_init when the P_c<P_cmin specifically comprises:

when P_c<P_cmin, detecting whether the current contention window value CW_old is less than a contention window threshold value Thr(n);

if yes, determining the current contention window value CW_old to be appropriate with respect to the current network, and the step length of the initial contention window value CW_step1=0, and maintaining the initial contention window value CW_init=min[CW_old, CW_max];

if no, determining the current contention window value CW_old to be too large with respect to the current network, and the step length of the initial contention window value CW_step1>0, and reducing the initial contention window value CW_init=min[CW_old−CW_step1, CW_max].

Preferably, adjusting the retransmission contention window value CW_new according to the practical collision probability P_c specifically comprises:

acquiring a Packet Error rate (PER) from a media access control layer;

detecting whether the PER is much greater than the practical collision probability P_c;

if yes, determining that a reason for data frame transmission failure is caused by bad network environment, and maintaining the retransmission contention window value CW_new=min[CW_old, CW_max]; and

if no, then determining that the reason for data frame transmission failure is caused by collision, and increasing the retransmission contention window value CW_new=min[CW_old+CW_step3], where CW_step3 is a step length of the retransmission contention window value.

The disclosure further provides a device for adjusting a contention window value adaptively, and the device comprises:

a measurement component configured to measure the number of stations participated in a contention currently:

an acquisition component configured to, according to the number n of stations participated in the contention currently, acquire the practical collision probability P_c of the current contention window value CW_old and optimal collision probability P_copt under the maximum network throughput;

a first adjustment component configured to, when a data frame is transmitted successfully, adjust an initial contention window value CW_init according to the practical collision probability P_c and optimal collision probability P_copt; and

a second adjustment component configured to, when data frame is transmitted unsuccessfully, according to the practical collision probability P_c, adjust retransmission contention window value CW_new.

Preferably, the acquisition component specifically comprises:

a first acquisition unit configured to, when the number of stations participated in the contention currently is n and the current contention window value is CW_old, determine the practical collision probability P_c=(1−(1+τ)ⁿ⁻¹[1+(n−1)τ], where τ is a sending probability of a station when a time slot of a channel begins and the current contention window value is CW_old; and

a second acquisition unit configured to, when the number of stations is participated in current contention is n and network throughput is maximum, determine the optimal collision probability P_copt=1−(1−τ_opt)ⁿ⁻¹[1+(n−1)τ_opt], where τ_opt is a sending probability of a station when a time slot of a channel begins and network throughput is maximum.

Preferably, the first adjustment component specifically comprises:

a first comparison unit configured to compare the practical collision probability P_c with a lower limit, which is P_cmax=P_copt−D_L, of a collision probability threshold value and an upper limit, which is P_cmin=P_copt+D_L, of the collision probability threshold value upper limit, where D_L is an interactive tolerance threshold;

a first adjustment unit configured to, when P_c<P_cmin, determine a current network to be idle, and reduce the initial contention window value CW_init=min[CW_old−CW_step1, CW_max], where CW_step1 is a step length of a first initial contention window value;

a second adjustment unit configured to, when P_c>P_cmax, determine a current network to be congested, and increase the initial contention window value CW_init=min[CW_old−CW_step2, CW_max], where CW_step2 is a step length of a second initial contention window value; and

a third adjustment unit configured to, when P_cmin≦P_c≦P_cmax, determine a current network to be good, and maintain the initial contention window value CW_init=min[CW_old, CW_max].

Preferably, the first adjustment unit is specifically configured to

when P_c<P_cmin, detect whether the current contention window value CW_old is less than a contention window threshold value Thr(n);

if yes, then determine the current contention window value CW_old to be appropriate with respect to the current network, and the step length of the initial contention window value CW_step1=0, and maintaining the initial contention window value CW_init=min[CW_old, CW_max].

if no, then determine the current contention window value CW_old to be too large with respect to the current network, and the step length of the initial contention window value CW_step1>0, and reduce the initial contention window is value CW_init=min[CW_old−CW_step1, CW_max].

Preferably, the second adjustment component specifically comprises:

a packet error rate acquisition unit configured to acquire a PER from a media access control layer;

a second comparison unit configured to detect whether the PER is much greater than the practical collision probability P_c;

a fourth adjustment unit configured to, when the PER is greater than P_c, determine that a reason for data frame transmission failure is caused by bad network environment, and maintain the retransmission contention window value CW_new=min[CW_old, CW_max]; and

a fifth adjustment unit configured to, when the PER is not much greater than P_c, then determine that a reason for data frame transmission failure is caused by collision, and increase the retransmission contention window value CW_new=min[CW_old+CW_step3, CW_max], where CW_step3 is a step length of the retransmission contention window value.

The embodiments of the disclosure can better reflect network collision situation, dynamically adjust initial contention window value associating with retransmission contention window value, thus making the contention window value finally approaching to optimal value, improve the network throughout, and reduce the delay.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a diagram showing relationships between the number of contention stations and collision probability in a method for adjusting a contention window value adaptively according to the disclosure;

FIG. 2 is a flowchart of a method for adjusting a contention window value adaptively according to a first embodiment of the disclosure;

FIG. 3 is a flowchart showing acquiring practical collision probability and optimal collision probability in a method for adjusting a contention window value adaptively according to a first embodiment of the disclosure;

FIG. 4 is a flowchart showing adjusting an initial contention window value in a is method for adjusting a contention window value adaptively according to a first embodiment of the disclosure;

FIG. 5 is a flowchart showing adjusting an initial contention window value in a method for adjusting a contention window value adaptively according to a second embodiment of the disclosure;

FIG. 6 is a flowchart showing adjusting, a retransmission contention window value in a method for adjusting a contention window value adaptively according to a first embodiment of the disclosure;

FIG. 7 is a flowchart of a method for adjusting a contention window value adaptively according to a second embodiment of the disclosure;

FIG. 8 is a structural schematic diagram of a device for adjusting a contention window value adaptively according to a first embodiment of the disclosure;

FIG. 9 is as structural schematic diagram of a device for adjusting a contention window value adaptively according to a second embodiment of the disclosure;

FIG. 10 is a structural schematic diagram of a device for adjusting a contention window value adaptively according to a third embodiment of the disclosure; and

FIG. 11 is a structural schematic diagram of a device for adjusting a contention window value adaptively according to a fourth embodiment of the disclosure;

Objectives and advantages related to the disclosure will be illustrated in the subsequent descriptions and appended drawings.

DESCRIPTION OF EMBODIMENTS

It should be understood that specific embodiments described here are only used for illustrating the disclosure and not intended to limit the disclosure.

As shown in FIG. 1, FIG. 1 is a diagram showing relationships between the number of contention stations and collision probability in a method for adjusting a contention window value adaptively according to the disclosure.

The diagram showing relationships in FIG. 1 in the present embodiment is obtained by simulation according, to theoretical calculation in the case of CW being a certain value. In the case of CW being 32 assuming that the number of stations is n, and in the case of CW>16·n the collision probability is less than 0.05, and in the case of the CW being of other lengths, the relationship is also approximately satisfied.

As shown in FIG. 2, FIG. 2 is a flowchart of a method for adjusting a contention window value adaptively according to a first embodiment of the disclosure. The method for adjusting a contention window value adaptively according to a first embodiment of the disclosure comprises the steps as follows:

step S10, the number n of stations participated in a contention currently is measured;

step S20, according to the number n of stations participated in the contention currently, a practical collision probability P_c of the current contention window value CW_old and optimal collision probability P_copt under the maximum network throughput are acquired;

step S30, when a data frame is transmitted successfully, an initial contention window value CW_init is adjusted according to the practical collision probability P_c and optimal collision probability P_copt; and

the practical collision probability and optimal collision probability are calculated according to the number of stations participated in the contention in the current network, and when the data frame is transmitted successfully, the practical collision probability is compared with the optimal collision probability under the maximum network throughput so as to dynamically adjust the initial contention window value.

Step S40, when the data frame is transmitted unsuccessfully, a retransmission contention window value CW_new is adjusted according to the practical collision probability P_c.

When the data frame is transmitted unsuccessfully, the practical collision probability is compared with the packet error rate so as to dynamically adjust the retransmission contention window value.

The present embodiment may better reflect network collision situation, dynamically adjust the initial contention window value and retransmission contention window value jointly, which may make the contention window value is finally approaching to optimal value, improve the network throughout, and reduce the delay.

As shown in FIG. 3, FIG. 3 is a flowchart showing acquiring practical collision probability and optimal collision probability in a method for adjusting a contention window value adaptively according to a first embodiment of the disclosure. Step S20 specifically comprises:

step S21, when the number of stations participated in the contention is n and the current contention window value is CW_old, the practical collision probability P_c=1−(1−τ)ⁿ⁻¹[1+(n−1)τ] is determined, where τ is a sending probability of a station at the beginning of a time slot of a channel when the current contention window value is CW_old; and

step S22, when the number of stations participated in the contention currently is n and network throughput is maximum, the optimal collision probability P_copt=1−(1−τ_opt)ⁿ⁻¹[1+(n−1)τ_opt] is determined, where τ_opt is a sending probability of a station at the beginning of a time slot of a channel when the network throughput is maximum.

The τ in the present embodiment is related to the number n of stations participated in the contention currently and current contention window value CW_old, and τ_opt is related to the number n of stations participated in the contention currently and an optimal CW value, i.e. it being obtained according to maximum network throughput. The two parameters may be obtained through statistic calculation. According to the number n of stations participated in the contention in the current network, the practical collision probability P_c of the current contention window value CW_old and optimal collision probability P_copt, under the maximum network throughput are calculated so as to dynamically adjust the initial contention window value CW_init and retransmission contention window value CW_new, which may better reflect network collision situation, and make the contention window value close to an optimal value so as to improve the network throughout, and reduce the delay.

As shown in FIG. 4, FIG. 4 is a flowchart showing adjusting an initial contention window value in a method for adjusting a contention window value adaptively according to a first embodiment of the disclosure.

Step S30 specifically comprises:

step S31 the practical collision probability P_c is compared with a lower limit, which is P_cmax=P_copt−D_L, the collision probability threshold value and with a upper limit, which is P_cmin=P_copt+D_L, of the collision probability threshold value, where D_L is an interactive tolerance threshold;

wherein the interactive tolerance threshold D_L may be obtained through simulation;

step S32, when P_c<P_cmin, a current network is determined to be idle, and the initial contention window value CW_init=min[CW_old−CW_step1, CW_max] is reduced, where CW_step1 is a step length of a first initial contention window value;

when P_c<P_cmin, it represents that the collision is lighter and the network is relative idle, and the contention window value should be reduced to improve time slot utilization ratio and reduce delay.

Step S33, when P_c>P_cmax, a current network is determined to be congested, and the initial contention window value CW_init=min[CW_old+CW_step2, CW_max] is increased, where CW_step2 is a step length of a second initial contention window value;

when P_c>P_cmax, it represents that collision is rather severe, and the network is rather congested, and the contention window value should be increased to reduce the collision.

Step S34, when P_cmin≦P_c≦P_cmax, a current network is determined to be good, and the initial contention window value CW_init=min[CW_old, CW_max] is maintained.

In the present embodiment, the practical collision probability P_c is compared with the optimal collision probability P_copt, under the maximum network throughput, and the initial contention window value CW_init is adjusted dynamically which may better reflect network collision situation, and make the contention window value close to an optimal value so as to improve the time slot utilization rate and reduce delay when the network is relatively idle, and reduce collisions when the network is relatively congested.

As shown in FIG. 5, FIG. 5 is a flowchart showing adjusting an initial contention window value in a method for adjusting a contention window value adaptively according to a second embodiment of the disclosure.

Step S30 specifically comprises:

step S31, the practical collision probability P_c is compared with a lower limit, which is P_cmax=P_copt−D_L, of the collision probability threshold value and with a upper limit, which is P_cmin=P_copt+D_L, of the collision probability threshold value, where D_L an interactive tolerance threshold;

Step S321, when P<P_cmin, it is detected whether a current contention window value CW_old is less than a contention window threshold value Thr(n); if yes, turn to step S322; and if no, turn to step S323;

The contention window threshold value Thr(n) is a function of the number n of stations participated in the contention currently, and may be an experience value obtained through relationship curve of the number of stations in the contention and collision probability according to the embodiment shown in FIG. 1.

Step S322, the current contention window value CW_old is determined to be appropriate with respect to the current network, and the step length of the initial contention window value CW_step1=0, and the initial contention window value CW_init=min[CW_old, CW_max] is maintained;

step S323, the current contention window value CW_old is determined to be too large with respect to the current network, and the step length of the initial contention window value CW_step1<0, and the initial contention window value CW_init=min[CW_old−CW_step1, CW_max] is reduced;

and since current contention window value CW_old is too large with respect to the current network, and the step length CW_step=2⁻¹·CW_old of regulated first initial contention window value should be decreased substantially, and the initial contention window value CW_init=min[2⁻¹·CW_old, CW_max] is decreased by half, so as to alleviate the decrease of throughput caused by the contention window value being too large as soon as possible.

Step S33, when P_c>P_cmax, a current network is determined to be congested, and the initial contention window value CW_init=min[CW_old−CW_step2, CW_max] is increased, where CW_step2 is a step length of a second initial contention window value;

at the moment, CW_step2=f(CW_old)·L(n), where f(CW_old) is a decreasing function of CW_old, and L(n) is an increasing function of n, and CW_step2 obtained through the two functions enables CW_init to approximately satisfy a relationship curve of the number of contention stations and collision probability in the embodiment as shown in FIG. 1. It can be seen that when the number n of stations participated in contention is certain, L(n) is constant, and with the increase of CW_old value, f(CW_old) decreases and CW_step2 decreases. In the case of the CW_old value being rather large, the collision probability decreases, and at the moment, CW needs to be regulated with a step length with a smaller value, i.e. slowly adjusting the CW, which is more beneficial to adjust the CW to an optimal value; however, in the case of the CW value being smaller, the collision probability is higher, and the CW needs to be regulated with a step length with as larger value, i.e. quickly regulating the CW, which is beneficial to adjust the CW to the optimal value more quickly.

Step S34, when P_cmin≦P_c≦P_cmax, a current network is determined to be good, and the initial contention window value CW_init=min[CW_old, CW_max] is maintained.

According to results of the practical collision probability P_c respectively comparing with the optimal collision probability P_copt and the contention window threshold value Thr(n), the initial contention window value CW_init is regulated in the present embodiment. When the network is relatively idle and the current contention window value CW_old is too large with respect to the current network, the contention window value is decreased substantially so as to alleviate the decreasing of throughput caused by the contention window value being too large; when the network is congested, the contention window value is increased to reduce collision.

As shown in FIG. 6, FIG. 6 is a flowchart showing adjusting a retransmission contention window value in a method for adjusting a contention window value adaptively according to a first embodiment of the disclosure.

Step S40 specifically comprises:

step S41, a Packet Error Rate (PER) is acquired from a Media Access Control (MAC) layer;

wherein PER=the number of frames which are verified wrongly number of totally verified frames.

Step S42, it is detected the PER is much greater than the practical collision probability P_c, and if yes, turn to step S43; and if no, turn to step S44.

PER>a·P_c may be used for determining that PER is much greater than P_c, where a is a multiple and may be determined according to the practical network condition, for example, when a=100, then when PER is greater than 100 multiples of P_c, the PER is determined to be much greater than P_c.

Step S43, a reason for data frame transmission failure is determined to be caused by bad network environment, and the retransmission contention window value CW_new=min[CW_old, CW_max] is maintained; and

step S44, the reason for data frame transmission failure is determined to be caused by collision, and the retransmission contention window value CW_new=min[CW_old+CW_step3, CW_max] is increased, where CW_step3 is the step length of the retransmission contention window value;

at the moment, CW_step3=f(CW_old)·L′(n), where f(CW_old) is a decreasing function of CW_old, and L′(n) is a second increasing function of n, and CW_step3 obtained through the two functions enables CW_new to approximately satisfy a relationship curve of the number of contention stations and collision probability in the embodiment as shown in FIG. 1. It can be seen that when the number n of stations participated in the contention is certain, L′(n) is constant, and with the increase of CW_old value, f(CW_old) decreases and CW_step3 decreases. In the case of the CW_old value being rather large, the collision probability decreases, and at the moment, CW needs to be regulated with a step length with a smaller value, i.e. slowly adjusting the CW, which is more beneficial to adjust the CW to an optimal value; however, in the case of the CW value being smaller, the collision probability is higher, and the CW needs to be regulated with a step length with a larger value, i.e. quickly regulating the CW, which is beneficial to adjust the CW to the optimal value more quickly. In addition, L′(n) may the same as L(n) in the embodiment shown in FIG. 5.

By comparing the practical collision probability with the PER, in the present embodiment the network collision situation may be reflected better, the retransmission contention window value may dynamically be adjusted to make the contention window value finally approaching to optimal value, improve the network throughout, and reduce the delay.

As shown in FIG. 7, FIG. 7 is method for adjusting a contention window value adaptively according to a second embodiment of the disclosure.

Step S501, CW_init=min[CW_min, CW_max] configured; and in the protocol of 802.11n, CW_min is specified to be 16.

Step S502, a random number is selected from the equal distribution on section [0, CW−1] as a random back-off duration;

Step S503, a station senses a channel to be idle, the DIPS duration is delayed;

Step S504, it is detected whether the channel is idle; if no, turn to step S505; and if yes, turn to step S506;

Step S505, a back-off counter is frozen, and back-off counting is paused, and step S503 is performed;

Step S506, the back-off counting, is continued, and it is detected whether the back-off counter reaches 0; if no, turn to step S504; and if yes, turn to step S507;

During a back-off period, if the channel maintains an idle state, then back-off counting is continued, and it is detected whether the back-off counter reaches 0, and if the channel is busy, then the back-off counting is paused until the channel is idle again, then the counting is continued.

Step S507, a data frame is transmitted.

Step S505, it is detected whether the data frame is transmitted successfully; if yes, turn to step S509; and if no, turn to step S510.

Step S509, the flow of adjusting the initial contention window value is performed;

and the flow of adjusting the initial contention window value may be referred to embodiments as shown in FIG. 1 to FIG. 5, which is not described in detail herein,

Step S510, it is detected whether a maximum retransmission time is reached; if yes, turn to step S511; and if no, turn to step S512.

Step S511, the data frame is discarded.

Step S512, the flow of adjusting the retransmission contention window value is performed.

The flow of adjusting the retransmission contention window value may be referred to embodiments as shown in FIG. 1, FIG. 2 and FIG. 6, which is not described in detail herein.

Since the flow of adjusting the initial contention window value and the flow of adjusting the retransmission contention window value m the present embodiment use all the technical solutions of the above embodiments as shown in FIG. 1 to FIG. 6, the present embodiment may likewise better reflect network collision situation, dynamically adjust initial contention window value and retransmission contention window value jointly, which may make the contention window value finally approaching to optimal value, improve the network throughout, and reduce the delay.

As shown in FIG. 8, FIG. 8 is a structural schematic diagram a device for adjusting a contention window value adaptively according to a first embodiment of the disclosure. The device for adjusting a contention window value adaptively in the present embodiment comprises:

a measurement component 10 configured to measure number n of stations participated in a contention currently;

an acquisition component 20 configured to acquire the practical collision probability P_c of the current contention window value CW_old and optimal collision probability P_copt under the maximum network throughput according to the number n of stations participated in the contention currently;

a first adjustment component 30 configured to according to the practical collision probability P_c and optimal collision probability P_copt when a data frame is transmitted successfully, adjust a initial contention window value CW_init; and

a second adjustment component 40 configured to adjust a retransmission contention window value CW_new according to the practical collision probability P_c to when the data frame is transmitted unsuccessfully.

In the present embodiment, the practical collision probability and optimal collision probability are calculated according to the number of stations participated in contention in the current network, and when the data frame is transmitted successfully, the practical collision probability is compared with the optimal is collision probability under the maximum network throughput so as to dynamically adjust the initial contention window value; and when the data frame is transmitted unsuccessfully, the practical collision probability is compared with the packet error rate so as to dynamically adjust the retransmission contention window value. In the disclosure, the network collision situation may be reflected better, and the initial contention window value and retransmission contention window value may be adjusted dynamically. Through the above solution the contention window value may approach to an optimal value, the network throughout may be improved, and the delay maybe reduced.

As shown in FIG. 9, FIG. 9 is a structural schematic diagram a device for adjusting a contention window value adaptively according to a second embodiment of the disclosure.

The acquisition component 20 specifically comprises:

a first acquisition unit 21 configured to, when the number of stations participated in the contention currently is ii and a current contention window value is CW_old, determine the practical collision probability P_c=1−(1−τ)ⁿ⁻¹[1+(n−1)τ], where τ is a sending probability of a station at the beginning of a time slot of a channel when the current contention window value is CW_old; and

a second acquisition unit 22 configured to, when the number of stations participated in the contention currently is n and network throughput is maximum, determine the optimal collision probability P_copt=1−(1−τ_opt)ⁿ⁻¹[1+(n−1)τ_opt], where τ_opt is a sending probability of a station at the beginning of a time slot of a channel when the network throughput is maximum.

The τ in the present embodiment is related to the number n of stations participated in the contention currently and current contention window value CW_old, and τ_opt is related to the number it of stations participated in the contention currently and an optimal CW value, i.e. it being obtained according to maximum network throughput. The two parameters may be obtained through statistic calculation. According to the number it of stations participated in contention in the current network, the practical collision probability P_c of the current contention window value CW_old and optimal collision probability P_copt, under the maximum network throughput are calculated so as to dynamically adjust the initial contention window value CW_init and retransmission contention window value CW_new, which may better reflect network collision situation, and make the contention window value close to an optimal value so as to improve the network throughout, and reduce the delay.

As shown in FIG. 10, FIG. 10 is a structural schematic diagram of a device for adjusting a contention window value adaptively according to a third embodiment of the disclosure.

The first adjustment component 30 specifically comprises:

a first comparison unit 31 configured to compare the practical collision probability P_c, with a lower limit, which is P_cmax=P_copt−D_L, of a collision probability threshold value and with a upper limit, which is P_cmin=P_copt+D_L, of the collision probability threshold value, where D_L an interactive tolerance threshold, and the interactive tolerance threshold D_L may be obtained through simulation;

a first adjustment unit 32 configured to, when P_c<P_cmin, determine a current network to be idle, and reduce the initial contention window value CW_init=min[CW_old−CW_step1, CW_max], where CW_step1 is a step length of a first initial contention window value;

a second adjustment unit 33 configured to, when P_c>P_cmax, determine the current network to be congested, and increase the initial contention window value CW_init=min[CW_old+CW_step2, CW_max], wherein the CW_step2 is a step length of a second initial contention window value; at the moment CW_step2=f(CW_old)·L(n), where f(CW_old) is a decreasing function of CW_old, and L(n) is an increasing function of n, and the CW_step2 obtained, through the two functions enables CW_init to approximately satisfy a relationship curve illustrating relationships between the number of contention stations and collision probability in the embodiment as shown in FIG. 1. It can be seen that when the number n of stations participated in contention is certain, L(n) is constant, and with the increase of CW_old value, f(CW_old) decreases and CW_step2 decreases. In the case of the CW_old value being rather large, the collision probability decreases, and at the moment, CW needs to be regulated with a step length with a smaller value, i.e. slowly adjusting the CW, which is more beneficial to adjust the CW to an optimal value; however, in the case is of the CW value being smaller, the collision probability is higher, and the CW needs to be regulated with a step length with a larger value, i.e. quickly regulating the CW, which is beneficial to adjust the CW to the optimal value more quickly.

a third adjustment unit 34 configured to, when P_cmin≦P_c≦P_cmax, determine a current network to be good, and maintain the initial contention window value CW_init=min[CW_old, CW_max].

In the present embodiment, when P_c<P_cmin, it represents that the collision is lighter and the network is relative idle, and the contention window value should be reduced to improve time slot utilization ratio and reduce delay; when P_c>P_cmax, it represents that collision is rather severe, and the network is rather congested, and the contention window value should be increased to reduce the collision. In the present embodiment, the practical collision probability P_c is compared with the optimal collision probability P_copt under the maximum network throughput, and the initial contention window value CW_init is adjusted dynamically. The above solution can better reflect network collision situation, and make the contention window value close to an optimal value so as to improve the time slot utilization rate and reduce delay when the network is relatively idle, and reduce collisions when the network is relatively congested.

The first adjustment unit 32 in the embodiment of the disclosure is specifically configured to:

when P_c<P_cmin, detect whether the current contention window value CW_old is less than a contention window threshold value Thr(n), where the contention window threshold value Thr(n) is a function of the number n of stations participated in the contention currently, and may be an experience value obtained through relationship curve illustrating relationships between the number of stations in contention and collision probability according to the embodiment shown in FIG. 1;

if yes, determine the current contention window value CW_old to be appropriate with respect to the current network, and set a step length of the initial contention window value CW_step1=0, and maintain the initial contention window value CW_init=min[CW_old, CW_max];

and if no, determine the current contention window value CW_old to be too large with respect to the current network, set a step length of the initial contention window value CW_step1>0, and reduce the initial contention window value CW_init=min[CW_old−CW_step1, CW_max].

In the present embodiment, when the CW_old is greater than the contention window threshold value Thr(n), and since current contention window value CW_old is too large with respect to the current network, and the step length CW_step=2⁻¹·CW_old of regulated first initial contention window value should be decreased substantially, and the initial contention window value CW_init=min[2⁻¹·CW_old, CW_max] is decreased by half, so as to alleviate the decrease of throughput caused by the contention window value being too large as soon as possible. According to results of the practical collision probability P_c respectively comparing with the optimal collision probability P_copt and the contention window threshold value Thr(n), the present embodiment regulates the initial contention window value CW_init; when the network is relatively idle and the current contention window value CW_old is too large with respect to the current network, the contention window value is decreased substantially so as to alleviate the decreasing of throughput caused by the contention window value being too large; when the network is congested, the contention window value is increased to reduce collision.

As shown in FIG. 11, FIG. 11 is a structural schematic diagram of a device for adjusting a contention window value adaptively according to a fourth embodiment of the disclosure.

The second adjustment component 40 specifically comprises:

a packet error rate acquisition unit 41 configured to acquire a PER from a MAC layer, wherein PER=the number of frames which are verified wrongly÷the number of totally verified frames;

a second comparison unit 42 configured to detect whether the PER is much greater than the practical collision probability P_c;

a fourth adjustment unit 43 configured to, when the PER is greater than P_c, determine that the reason for data frame transmission failure is caused by bad network environment, and maintain the retransmission contention window value CW_new=min[CW_old, CW_max]; and

a fifth adjustment unit 44 configured to, when the PER is not much greater than P_c, determine that the reason for data frame transmission failure is caused by collision, and increase the retransmission contention window value CW_new=min[CW_old+CW_step3, CW_max], where CW_step3 is a step length of the retransmission contention window value. At the moment, CW_step3=f(CW_old)−L′(n), where f(CW_old) is a decreasing function of CW_old, and L′(n) is a second increasing function of n, and CW_step3 obtained according to the two functions enables CW_new to approximately satisfy a relationship curve showing relationships between the number of contention stations and collision probability in the embodiment as shown in FIG. 1. It can be seen that when the number n of stations participated in contention is certain, L′(n) is constant, and with the increase of L′(n), f(CW_old) decreases and CW_step3 decreases. In the case of the CW_old value being rather large, the collision probability decreases, and at the moment, CW needs to be regulated with a step length with a smaller value, i.e. slowly adjusting the CW, which is more beneficial to adjust the CW to an optimal value; however, in the case of the CW value being smaller, the collision probability is higher, and the CW needs to be regulated with a step length with a larger value, i.e. quickly regulating the CW, which is beneficial to adjust the CW to the optimal value more quickly. In addition, L′(n) may be the same as L(n) in the embodiment shown in FIG. 5.

In the present embodiment, PER>a·P_c, may be used for further determining that PER is much greater than P_c, where a is a multiple and may be determined according to the practical network condition, for example, when a=100, then when PER is greater than 100 multiples of P_c, the PER is determined to be much greater than P_c. By comparing the practical collision probability with the packet error rate, the present embodiment may better reflect network collision situation, dynamically adjust retransmission contention window value to make the contention window to value finally approaching to optimal value, improve the network throughout, and reduce the delay.

Any equivalent replacements of the structure or the flow based on the contents of the specification and drawings of the disclosure, or direct or indirect application of the equivalent replacements in other related technical fields shall fall within the scope of protection of the disclosure.

INDUSTRIAL APPLICABILITY

The above-mentioned technical solutions provided in the disclosure may be applied to contention window value adaptive adjustment, and based on the technical solutions provided in the disclosure the network collision situation may be reflected better, the initial contention window value and retransmission contention window value are dynamically adjusted, thus making the contention window value finally approaching to optimal value, improving the network throughout, and reducing the delay.

Claims

1. A method for adjusting contention window value adaptively, comprising: measuring a number n of stations participated in a contention currently;acquiring a practical collision probability Pc of a current contention window value CWold and optimal collision probability Pcopt under maximum network throughput according to the number n of stations participated in the contention currently;when a data frame is transmitted successfully, adjusting an initial contention window value CWinit according to the practical collision probability Pc and optimal collision probability Pcopt; andwhen the data frame is transmitted unsuccessfully, adjusting a retransmission contention window value CWnew according to the practical collision probability Pc.

2. The method for adjusting contention window value adaptively according to claim 1, wherein acquiring the practical collision probability Pc of the current contention window value CWold and optimal collision probability Pcopt under the maximum network throughput according to the number n of stations participated in the contention currently comprises: when the number of stations participated in the contention currently is n and the current contention window value is CWold, determining the practical collision probability Pc=1−(1−τ)n−1[1+(n−1)τ], where τ is a sending probability of a station at a beginning of a time slot of a channel when the current contention window value is CWold; andwhen the number of stations participated in the contention currently is n and network throughput is maximum, determining the optimal collision probability Pcopt=1−(1−τopt)n−1[1+(n−1)τopt], where τopt is a sending probability of a station at a beginning of a time slot of a channel when the network throughput is maximum.

3. The method for adjusting contention window value adaptively according to claim 1, wherein adjusting the initial contention window value CWinit according to the practical collision probability Pc and the optimal collision probability Pcopt comprises: comparing the practical collision probability Pc with a lower limit, which is Pcmax=Pcopt−DL, of a collision probability threshold value and an upper limit, which is Pcmin=Pcopt+DL, of the collision probability threshold value, where DL is an interactive tolerance threshold value;when Pc<Pcmin, determining a current network to be idle, and reducing the initial contention window value CWinit=min[CWold−CWstep1, CWmax], where CWstep1 is a step length of a first initial contention window value, and wherein CWmax is a maximum contention window value;when Pc>Pcmax, determining a current network to be congested, and increasing the initial contention window value CWinit=min[CWold+CWstep2, CWmax], where CWstep2 is a step length of a second initial contention window value;when Pcmin≦Pc≦Pcmax, determining a current network to be good, and maintaining the initial contention window value CWinit=min[CWold, CWmax].

4. The method for adjusting contention window value adaptively according to claim 3, wherein determining the current network to be idle and reducing the initial contention window value CWinit when Pc<Pcmin comprises: when Pc<Pcmin, detecting whether the current contention window value CWold is less than a contention window threshold value Thr(n);based on the CWold is less than the Thr(n), determining the current contention window value CWold is appropriate with respect to the current network, and the step length of the first initial contention window value CWstep1=0, and maintaining the initial contention window value CWinit=min[CWold, CWmax];based on the CWold is not less than the Thr(n), determining the current contention window value CWold is too large with respect to the current network, and the step length of the first initial contention window value CWstep1>0, and reducing the initial contention window value CWinit=min[CWold−CWstep1, CWmax].

5. The method for adjusting contention window value adaptively according to claim 1, wherein adjusting the retransmission contention window value CWnew according to the practical collision probability Pc comprises: acquiring a Packet Error Rate (PER) from a media access control layer;detecting whether the PER is much greater than the practical collision probability Pc;when the PER is much greater than the Pc, determining that a reason for data frame transmission failure is caused by bad network environment, and maintaining the retransmission contention window value CWnew=min[CWold, CWmax], and wherein CWmax is a maximum contention window value; andwhen the PER is not much greater than the Pc, determining that a reason for data frame transmission failure is caused by collision, and increasing the retransmission contention window value CWnew=min[CWold+CWstep3, CWmax], where CWstep3 is a step length of the retransmission contention window value.

6. A device for adjusting contention window value adaptively, comprising: a memory; anda hardware processor, configured to execute programming components stored in the memory, wherein the programming components comprise:a measurement component configured to measure a number n of stations participated in a contention currently;an acquisition component configured to acquire a practical collision probability Pc of a current contention window value CWold and optimal collision probability Pcopt under maximum network throughput according to the number n of stations participated in the contention currently;a first adjustment component configured to adjust an initial contention window value CWinit according to the practical collision probability Pc and optimal collision probability Pcopt when a data frame is transmitted successfully; anda second adjustment component configured to adjust a retransmission contention window value CWnew according to the practical collision probability Pc when a data frame is transmitted unsuccessfully.

7. The device for adjusting contention window value adaptively according to claim 6, wherein the acquisition component comprises: a first acquisition unit configured to determine the practical collision probability Pc=1−(1−τ)n−1[1+(n−1)τ] when the number of stations participated in the contention currently is n and the current contention window value is CWold, where τ is a sending probability of a station at a beginning of a time slot of a channel and the current contention window value is CWold; anda second acquisition unit configured to determine the optimal collision probability Pcopt=1−(1−τopt)n−1[1+(n−1)τopt] when the number of stations participated in the contention currently is n and network throughput is maximum, where τopt is a sending probability of a station at a beginning of a time slot of a channel when the network throughput is maximum.

8. The device for adjusting contention window value adaptively according to claim 6, wherein the first adjustment component comprises: a first comparison unit configured to compare the practical collision probability Pc with a lower limit, which is Pcmax=Pcopt−DL, of a collision probability threshold value and with an upper limit, which is Pcmin=Pcopt+DL, of a collision probability threshold value, where DL is an interactive tolerance threshold value;a first adjustment unit configured to, when Pc<Pcmin, determine a current network to be idle, and reduce the initial contention window value CWinit=min[CWold−CWstep1, CWmax], where CWstep1 is a step length of a first initial contention window value, and wherein CWmax is a maximum contention window value;a second adjustment unit configured to, when Pc>Pcmax, determine the current network to be congested, and increase the initial contention window value CWinit=min[CWold+CWstep2, CWmax], where CWstep2 is a step length of a second initial contention window value; anda third adjustment unit configured to, when Pcmin≦Pc≦Pcmax, determine a current network to be good, and maintain the initial contention window value CWinit=min[CWold, CWmax].

9. The device for adjusting contention window value adaptively according to claim 8, wherein the first adjustment unit is configured to: when Pc<Pcmin, detect whether the current contention window value CWold is less than a contention window threshold value Thr(n);when the CWold is less than the Thr(n), determine that the current contention window value CWold is appropriate with respect to the current network, and the step length of the first initial contention window value CWstep1=0, and maintain the initial contention window value CWinit=min[CWold, CWmax], and wherein CWmax is a maximum contention window value;when the CWold is not less than the Thr(n), determine that the current contention window value CWold is too large with respect to the current network, and the step length of the first initial contention window value CWstep1>0, and reduce the initial contention window value CWinit=min[CWold−CWstep1, CWmax].

10. The device for adjusting contention window value adaptively according to claim 6, wherein the second adjustment component comprises: a packet error rate acquisition unit configured to acquire a Packet Error Rate (PER) from a media access control layer;a second comparison unit configured to detect whether the PER is much greater than the practical collision probability Pc;a fourth adjustment unit configured to, when the PER is much greater than Pc, determine that a reason for data frame transmission failure is caused by bad network environment, and maintain the retransmission contention window value CWnew=min[CWold, CWmax] and wherein CWmax is a maximum contention window value; anda fifth adjustment unit configured to, when the PER is not much greater than Pc, determine that a reason for data frame transmission failure is caused by collision, and increase the retransmission contention window value CWnew=min[CWold+CWstep3, CWmax], where CWstep3 is a step length of the retransmission contention window value.

11. The method for adjusting contention window value adaptively according to claim 2, wherein adjusting the initial contention window value CWinit according to the practical collision probability Pc and the optimal collision probability Pcopt comprises: comparing the practical collision probability Pc with a lower limit, which is Pcmax=Pcopt−DL, of a collision probability threshold value and an upper limit, which is Pcmin=Pcopt+DL, of the collision probability threshold value, where DL is an interactive tolerance threshold value;when Pc<Pcmin, determining a current network to be idle, and reducing the initial Penin, contention window value CWinit=min[CWold−CWstep1, CWmax], where CWstep1 is a step length of a first initial contention window value, and wherein CWmax is a maximum contention window value;when Pc>Pcmax, determining a current network to be congested, and increasing the initial contention window value CWinit=min[CWold+CWstep2, CWmax], where CWstep2 is a step length of a second initial contention window value;when Pcmin≦Pc≦Pcmax, determining a current network to be good, and maintaining the initial contention window value CWinit=min[CWold, CWmax].

12. The method for adjusting contention window value adaptively according to claim 2, wherein adjusting the retransmission contention window value CWnew according to the practical collision probability Pc comprises: acquiring a Packet Error Rate (PER) from a media access control layer;detecting whether the PER is much greater than the practical collision probability Pc;when the PER is much greater than the Pc, determining that a reason for data frame transmission failure is caused by bad network environment, and maintaining the retransmission contention window value CWnew=min[CWold, CWmax], and wherein CWmax is a maximum contention window value; andwhen the PER is not much greater than the Pc, determining that a reason for data frame transmission failure is caused by collision, and increasing the retransmission contention window value CWnew=min[CWold+CWstep3, CWmax], where CWstep3 is a step length of the retransmission contention window value.

13. The device for adjusting contention window value adaptively according to claim 7, wherein the first adjustment component comprises: a first comparison unit configured to compare the practical collision probability Pc with a lower limit, which is Pcmax=Pcopt−DL, of a collision probability threshold value and with an upper limit, which is Pcmin=Pcopt+DL, of a collision probability threshold value, where DL is an interactive tolerance threshold value;a first adjustment unit configured to, when Pc<Pcmin, determine a current network to be idle, and reduce the initial contention window value CWinit=min[CWold−CWstep1, CWmax], where CWstep1 is a step length of a first initial contention window value, and wherein CWmax is a maximum contention window value;a second adjustment unit configured to, when Pc>Pcmax, determine the current network to be congested, and increase the initial contention window value CWinit=min[CWold+CWstep2, CWmax], where CWstep2 is a step length of a second initial contention window value; anda third adjustment unit configured to, when Pcmin≦Pc≦Pcmax, determine a current network to be good, and maintain the initial contention window value CWinit=min[CWold, CWmax].

14. The device for adjusting contention window value adaptively according to claim 7, wherein the second adjustment component comprises: a packet error rate acquisition unit configured to acquire a Packet Error Rate (PER) from a media access control layer;a second comparison unit configured to detect whether the PER is much greater than the practical collision probability Pc;a fourth adjustment unit configured to, when the PER is much greater than Pc, determine that a reason for data frame transmission failure is caused by bad network environment, and maintain the retransmission contention window value CWnew=min[CWold, CWmax]; anda fifth adjustment unit configured to, when the PER is not much greater than Pc, determine that a reason for data frame transmission failure is caused by collision, and increase the retransmission contention window value CWnew=min[CWold+CWstep3, CWmax], where CWstep3 is a step length of the retransmission contention window value.