CN103596220A - Method for flow control combining optimal admission control and elimination control - Google Patents

Abstract

The invention relates to service quality guarantees and performance optimization in a cognitive radio network and provides a QoS guarantee mechanism combining optimal admission control and elimination control. The QoS guarantee mechanism comprises the steps that a system is initialized and the optimal admission probability, the elimination probability and each performance index of a cognitive user are calculated. According to the QoS guarantee mechanism based on the dynamic admission probability and the elimination probability, the QoS requirements of the cognitive user can be strictly satisfied under the condition that the throughput rate of the system is the maximum; a method can be conducted in a centralized network.

Description

A kind of flow control methods of combining best receiving control and rejecting control

Technical field

The present invention relates to the method for flow control in cognitive radio networks, relate in particular to a kind of best strategy of controlling of receiving and reject of combining.

Technical background

Cognitive radio technology can make the cognitive user frequency spectrum hole that authorized user uses in perception cognition network of having the ability, and then chance ground goes to utilize these frequency spectrum holes to reach the object of the raising availability of frequency spectrum.Yet, because making the service quality (QoS) of cognitive user, the uncertainty of authorized user often can not be guaranteed, therefore, in cognition network, guarantee that the qos requirement of cognitive user becomes the emphasis of people's research.

Call Admission Control is an important method that guarantees cognitive user qos requirement.It is base station according to system loading condition at that time, and consider newly to cognitive user, the impact of other QoS of customer to be determined whether to the cognitive user that receiving is newly arrived simultaneously.At present traditional admission control scheme is all based on channel reservation mechanism and switches buffering area mechanism, and based on admission probability mechanism.For channel reservation mechanism, it is exactly base station reserved a part of channel in advance, only for the higher switching user of service priority (user who is tied up due to the arrival of authorized user), while only having channel reservation available in system, to refuse cognitive user connecting system, to guarantee that the drop rate (being forced to the probability of break in service) of cognitive user remains on below a certain threshold value; And machine-processed for switching buffering area, be to switch user will be suspended (or entering in virtual buffering region) wait until there is available channel to occur, in the process being suspended switching user, will refuse all cognitive user connecting systems that newly arrive; In the mechanism of admission probability, all new users of arriving will, with Probability p connecting system (static receiving controlled), with control system flow, reach the object that guarantees cognitive user QoS.

Although these methods can guarantee the qos requirement of cognitive user, but still come with some shortcomings, make the throughput of system not reach optimum.First, even if provided best channel reservation, the drop rate of cognitive user is not necessarily just in given threshold value place (likely much lower than given threshold value) also, will make like this system throughput degradation a lot; Secondly, if the user who waits in buffering area can not get available channel for a long time, handover delay now will be very large, can not meet equally user's qos requirement; Finally, if to all users all with Probability p connecting system,, when system load is very light, control is now unnecessary, has so also reduced the throughput of system.

Summary of the invention

For the deficiencies in the prior art, the object of this invention is to provide a kind of strategy that associating the best is received and rejecting is controlled in cognitive radio networks, this strategy is based on dynamic accommodation probability, be that to the admission probability of cognitive user, the variation along with system load changes in base station, it can make system throughput in the situation that arrival is maximum, guarantees that the drop rate of cognitive user and average handover delay remain on below given thresholding simultaneously.

For realizing above goal of the invention, the present invention by the following technical solutions:

Combine best receiving and control and reject a flow control methods of controlling, be applied to, in cognitive radio networks, comprise the following steps:

Step (a), system initialization;

Step (b), cognitive user are carried out channel-aware, and sensing results is sent to base station;

Corresponding best admission probability vector table or the best probability vector table of rejecting are searched according to sensing results in step (c), base station, wherein, supposing the system current state is (i, j, k), base station will be interval [0,1] in, get a random number t, described i, j, the number of users of waiting in active authorized user number in k expression system respectively, active cognitive user number and buffering area;

(c.1) for newly arriving user, if this random number t is less than corresponding admission probability

this user will be accepted service by idle channel of Random assignment; Otherwise this user will get clogged;

(c.2) for switching user, if the current available free channel of system, i.e. i+j<M, M is total authorization channel number, any control will not done directly to idle channel of this switching user Random assignment in base station; If otherwise random number t is less than corresponding rejecting probability

this switching user will enter in buffering area and wait for until available free channel; Otherwise it will be interrupted service and directly go offline.

Compared with prior art, technique effect of the present invention is: can make the throughput of system maximize, can strictly guarantee drop rate and the average handover delay of cognitive user simultaneously; Can in centralized network, realize.

Accompanying drawing explanation

Fig. 1 is system mode transition diagram;

Fig. 2 is for guaranteeing machine-processed flow chart based on dynamic accommodation probability with the QoS that rejects probability;

Fig. 3 is the blocking rate comparison diagram of cognitive user in performance simulation;

Fig. 4 is the drop rate comparison diagram of cognitive user in performance simulation;

Fig. 5 is the average handover delay comparison diagram of cognitive user in performance simulation;

Fig. 6 is the throughput comparison diagram of cognitive user in performance simulation.

Embodiment

In order to make object of the present invention, technical scheme and advantage clearer, below in conjunction with accompanying drawing and exemplary embodiment, the present invention is further elaborated.Should be appreciated that exemplary embodiment described herein is only in order to explain the present invention, the scope of application being not intended to limit the present invention.

Consider so centralized cognitive radio networks: authorized user has absolute priority to remove to take authorization channel compared with cognitive user, and the priority of the cognitive user that the user Bi Xin in buffering area arrives is high, if and only if, and buffering area is when empty, newly to cognitive user just likely the system of entering accept service.

In the present invention, base station is controlled the real-time status according to system accordingly to cognitive user, to realize dynamic receiving, controls and rejects and control.QoS based on dynamic accommodation probability and rejecting probability guarantees machine-processed flow chart as shown in Figure 2, comprises the following steps:

Step (a), system initialization:

(a.1) given system parameters.Consider that such one has the centralized cognitive system of switching buffering area: comprise that total authorization channel counts M, the arrival rate λ of authorized user and cognitive user _pand λ _sand service rate μ _pand μ _s(supposing that user arrives and leave system with Poisson process).

(a.2) determine best admission probability vector table Ρ _adwith rejecting probability vector table Ρ _ev, it is by the best admission probability under different system state

reject probability with the best

form, $P_{\mathrm{ad}}=\left\{p_{\mathrm{ad}}^{(i,j,k)}\right|0\&le;i\&le;M,0\&le;j\&le;M-i,0\&le;k\&le;M-j\},$ $P_{\mathrm{ev}}=\left\{p_{\mathrm{ev}}^{(i,j,k)}\right|0\&le;i\&le;M,0\&le;j\&le;M-i,0\&le;k\&le;M-j\}.$ In the present invention, best in order to obtain

with

will be by a three-dimensional equine husband model, a state of its state (i, j, k) representative system, i wherein, j, the number of users of waiting in active authorized user number in k expression system respectively, active cognitive user number and buffering area.Its state transition diagram as shown in Figure 1.

(a.2.1) probability of stability of computing system.

According to state transition diagram, obtain an equilibrium equation:

$\begin{array}{c}({\mathrm{\&lambda;}}_p{\mathrm{\&theta;}}_1+{\mathrm{\&lambda;}}_p{\mathrm{\&theta;}}_3+{\mathrm{\&lambda;}}_s{\mathrm{\&theta;}}_1{p}_{\mathrm{ad}}^{(i,j,k)}+{\mathrm{i\&mu;}}_p+j{\mathrm{\&mu;}}_s)p_{i,j,k}\\ =(i+1){\mathrm{\&mu;}}_p{\mathrm{\&theta;}}_2{p}_{i+1,j,k}+(i+1){\mathrm{\&mu;}}_p{\mathrm{\&theta;}}_2{p}_{i+1,j-1,k+1}+(j+1){\mathrm{\&mu;}}_s{\mathrm{\&theta;}}_2{p}_{i,j+1,k}\\ +{\mathrm{\&lambda;}}_p{\mathrm{\&theta;}}_3(1-p_{ev}^{(i,j,k)})p_{i-1,j+1,k}+{\mathrm{\&lambda;}}_p{\mathrm{\&theta;}}_1{p}_{i-1,j,k}+{\mathrm{\&lambda;}}_p{\mathrm{\&theta;}}_3{p}_{\mathrm{ev}}^{(i,j,k-1)}{p}_{i-1,j+1,k-1}\\ +j{\mathrm{\&mu;}}_s(1-{\mathrm{\&theta;}}_2)p_{i,j,k+1}+{\mathrm{\&lambda;}}_s{\mathrm{\&theta;}}_1{p}_{\mathrm{ad}}^{(i,j-1,k)}{p}_{i,j-1,k}\end{array}---\left(1\right)$

θ wherein ₁=1 _i+j<M; θ ₂=1 _k=0; θ ₃=1 _{i+j=M & j>0};

be illustrated in the new best admission probability to cognitive user under current state (i, j, k);

be illustrated in current state (i, j, k) incision and use the best switching probability (entering the probability of waiting in buffering area) at family instead; p _{i, j, k}be illustrated in the probability of stability of system under current state (i, j, k).Associating equilibrium equation (1) and normalization equation the probability of stability p of each state will be obtained _{i, j, k}.Note, due to Ρ _adand Ρ _evunknown vector to be asked, so probability of stability p herein _{i, j, k}be unknown vector Ρ _adand Ρ _evfunction.

(a.2.2) according to the probability of stability of system, the performance index of computing system.The performance index of obviously, now trying to achieve are also unknown vector Ρ _adand Ρ _evfunction.

I) blocking rate of cognitive user: because system is saturated or receive control to cause the probability newly getting clogged to cognitive user.

$P_{\mathrm{block}}(P_{\mathrm{ad}},P_{\mathrm{ev}})=\underset{\&ForAll;i,j,k}{\mathrm{\&Sigma;}}(1-p_{\mathrm{ad}}^{(i,j,k)})\&CenterDot;p_{i,j,k}---\left(2\right)$

Ii) the drop rate of cognitive user: when the cognitive user newly being tied up to authorized user does not perceive idle channel, cause this user to be forced to the probability of break in service owing to rejecting to control.

$P_{\mathrm{drop}}(P_{\mathrm{ad}},P_{\mathrm{ev}})=\frac{{\mathrm{\&lambda;}}_p\underset{i+j=M,j>0}{\mathrm{\&Sigma;}}(1-p_{\mathrm{ev}}^{(i,j,k)})\&CenterDot;p_{i,j,k}}{{\mathrm{\&lambda;}}_s(1-P_{\mathrm{block}})}---\left(3\right)$

Iii) the average handover delay of cognitive user: switch the average waiting time delay of user in buffering area.

$D_{\mathrm{ho}}(P_{\mathrm{ad}},P_{\mathrm{ev}})=\frac{\stackrel{\&OverBar;}{N_{\mathrm{hdof}}}}{\stackrel{\&OverBar;}{N_{\mathrm{colision}}}}\&CenterDot;\frac{\stackrel{\&OverBar;}{N_{\mathrm{buffer}}}}{\stackrel{\&OverBar;}{R_{\mathrm{hdof}}}}---\left(4\right)$

Wherein

represent every authorized user, be switched to the average of the cognitive user of waiting in buffering area;

represent every authorized user, the average of the cognitive user bumping with it;

expression system reaches the average user number in buffering area while stablizing; $\stackrel{\&OverBar;}{R_{\mathrm{hdof}}}={\mathrm{\&lambda;}}_p\&CenterDot;{\mathrm{\&Sigma;}}_{i+j=M,j>0}{p}_{\mathrm{ev}}^{(i,j,k)}\&CenterDot;p_{i,j,k}$ Represent that cognitive user is switched to the Mean Speed in buffering area.

Iv) the throughput of cognitive user: the cognitive user number that completes service in the unit interval.

R _th(Ρ _ad,Ρ _ev)=λ _s·(1-P _block)·(1-P _drop) (5)

(a.2.3) optimization problem that provides system is described, in the hope of best admission probability and rejecting probability.

max R _th(Ρ _ad,Ρ _ev)

$s.t.\left\{\begin{array}{c}{P}_{\mathrm{drop}}(P_{\mathrm{ad}},P_{\mathrm{ev}})\&le;P_d^{*}\\ D_{\mathrm{ho}}(P_{\mathrm{ad}},P_{\mathrm{ev}})\&le;D_h^{*}\\ 0\&le;P_{\mathrm{ad}}\&le;1\\ 0\&le;P_{\mathrm{ad}}\&le;1\end{array}\right.---\left(6\right)$

Wherein

with

be respectively the drop rate of cognitive user and the threshold value of average handover delay.

Step (b): cognitive user is carried out channel-aware, in practice, restriction due to hardware or efficiency, cognitive user may not all channels of perception, even have certain perception mistake, for easy analysis, the present invention will not consider these situations, but it should be understood by those skilled in the art that it is also very easy that the present invention is applied in these actual conditions.For newly to cognitive user and switch user and all will carry out continuous channel-aware, until perceive available free channel or the complete all channels of perception, subsequently, cognitive user sends to ，You base station, base station determine whether agreeing to connecting system or enter buffering area sensing results information.

Step (c): corresponding decision-making is carried out according to sensing results in base station.Base station is after receiving the channel information that cognitive user sends, by the best admission probability vector table corresponding according to this information searching (for new to user) or the best probability vector table (for switching user) of rejecting.

(c.1) for newly arriving user, supposing the system current state is (i, j, k), and a random number t will be got in base station in interval [0,1], if this random number t is less than corresponding admission probability this user will be accepted service by idle channel of Random assignment; Otherwise this user will get clogged.

(c.2), for switching user, same supposing the system current state is (i, j, k), if the current available free channel of system, i.e. i+j<M, any control will not done directly to idle channel of this switching user Random assignment in base station; If otherwise random number t is less than corresponding rejecting probability

this switching user will enter in buffering area and wait for until available free channel; Otherwise it will be interrupted service and directly go offline.

A large amount of l-G simulation tests show, than channel reservation mechanism and static receiving, control, institute of the present invention extracting method there is larger system throughput, and can strictly guarantee the qos requirement of cognitive user.

The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any modifications of doing within the spirit and principles in the present invention, be equal to and replace and improvement etc., within all should being included in protection scope of the present invention.

Claims (6)

1. combine best receiving and control and reject a flow control methods of controlling, be applied to, in cognitive radio networks, comprise the following steps:

Step (a), system initialization;

Step (b), cognitive user are carried out channel-aware, and sensing results is sent to base station;

Corresponding best admission probability vector table or the best probability vector table of rejecting are searched according to sensing results in step (c), base station, wherein, supposing the system current state is (i, j, k), base station will be interval [0,1] in, get a random number t, described i, j, the number of users of waiting in active authorized user number in k expression system respectively, active cognitive user number and buffering area;

(c.1) for newly arriving user, if this random number t is less than corresponding admission probability

this user will be accepted service by idle channel of Random assignment; Otherwise this user will get clogged;

(c.2) for switching user, if the current available free channel of system, i.e. i+j<M, M is total authorization channel number, any control will not done directly to idle channel of this switching user Random assignment in base station; If otherwise random number t is less than corresponding rejecting probability

this switching user will enter in buffering area and wait for until available free channel; Otherwise it will be interrupted service and directly go offline.

2. method according to claim 1, wherein, described step (a) comprising:

(a.1) given system parameters: total authorization channel is counted the arrival rate λ of M, authorized user and cognitive user _pand λ _sand service rate μ _pand μ _s;

(a.2) determine best admission probability vector table Ρ _adwith rejecting probability vector table Ρ _ev, it is by the best admission probability under different system state

reject probability with the best

form, $P_{\mathrm{ad}}=\left\{p_{\mathrm{ad}}^{(i,j,k)}\right|0\&le;i\&le;M,0\&le;j\&le;M-i,0\&le;k\&le;M-j\},$ $P_{\mathrm{ev}}=\left\{p_{\mathrm{ev}}^{(i,j,k)}\right|0\&le;i\&le;M,0\&le;j\&le;M-i,0\&le;k\&le;M-j\}.$

3. method according to claim 2, wherein, described step (a.2) comprising:

(a.2.1) probability of stability of computing system, obtains the probability of stability p of each state of system _{i, j, k};

(a.2.2) according to the performance index of the probability of stability computing system of system;

(a.2.3) optimization problem that provides system is described, in the hope of best admission probability and rejecting probability.

4. method according to claim 3, wherein, described step (a.2.1) comprising:

According to state transition diagram, be balanced equation:

$\begin{array}{c}({\mathrm{\&lambda;}}_p{\mathrm{\&theta;}}_1+{\mathrm{\&lambda;}}_p{\mathrm{\&theta;}}_3+{\mathrm{\&lambda;}}_s{\mathrm{\&theta;}}_1{p}_{\mathrm{ad}}^{(i,j,k)}+{\mathrm{i\&mu;}}_p+j{\mathrm{\&mu;}}_s)p_{i,j,k}\\ =(i+1){\mathrm{\&mu;}}_p{\mathrm{\&theta;}}_2{p}_{i+1,j,k}+(i+1){\mathrm{\&mu;}}_p{\mathrm{\&theta;}}_2{p}_{i+1,j-1,k+1}+(j+1){\mathrm{\&mu;}}_s{\mathrm{\&theta;}}_2{p}_{i,j+1,k}\\ +{\mathrm{\&lambda;}}_p{\mathrm{\&theta;}}_3(1-p_{ev}^{(i,j,k)})p_{i-1,j+1,k}+{\mathrm{\&lambda;}}_p{\mathrm{\&theta;}}_1{p}_{i-1,j,k}+{\mathrm{\&lambda;}}_p{\mathrm{\&theta;}}_3{p}_{\mathrm{ev}}^{(i,j,k-1)}{p}_{i-1,j+1,k-1}\\ +j{\mathrm{\&mu;}}_s(1-{\mathrm{\&theta;}}_2)p_{i,j,k+1}+{\mathrm{\&lambda;}}_s{\mathrm{\&theta;}}_1{p}_{\mathrm{ad}}^{(i,j-1,k)}{p}_{i,j-1,k}\end{array}$

θ wherein ₁=1 _{i+j < M}; θ ₂=1 _k=0; θ ₃=1 _{i+j=M & j > 0};

be illustrated in the new best admission probability to cognitive user under current state (i, j, k);

be illustrated in current state (i, j, k) incision and use the best switching probability (entering the probability of waiting in buffering area) at family instead; p _{i, j, k}be illustrated in the probability of stability of system under current state (i, j, k);

Combine above-mentioned equilibrium equation and normalization equation

solve the probability of stability p that obtains each state _{i, j, k}.

5. method according to claim 4, wherein, described performance index comprise:

I) blocking rate of cognitive user:

$P_{\mathrm{block}}(P_{\mathrm{ad}},P_{\mathrm{ev}})=\underset{\&ForAll;i,j,k}{\mathrm{\&Sigma;}}(1-p_{\mathrm{ad}}^{(i,j,k)})\&CenterDot;p_{i,j,k}$

Ii) the drop rate of cognitive user:

$P_{\mathrm{drop}}(P_{\mathrm{ad}},P_{\mathrm{ev}})=\frac{{\mathrm{\&lambda;}}_p\underset{i+j=M,j>0}{\mathrm{\&Sigma;}}(1-p_{\mathrm{ev}}^{(i,j,k)})\&CenterDot;p_{i,j,k}}{{\mathrm{\&lambda;}}_s(1-P_{\mathrm{block}})}$

Iii) the average handover delay of cognitive user:

$D_{\mathrm{ho}}(P_{\mathrm{ad}},P_{\mathrm{ev}})=\frac{\stackrel{\&OverBar;}{N_{\mathrm{hdof}}}}{\stackrel{\&OverBar;}{N_{\mathrm{colision}}}}\&CenterDot;\frac{\stackrel{\&OverBar;}{N_{\mathrm{buffer}}}}{\stackrel{\&OverBar;}{R_{\mathrm{hdof}}}}$

Wherein

represent every authorized user, be switched to the average of the cognitive user of waiting in buffering area;

represent every authorized user, the average of the cognitive user bumping with it;

expression system reaches the average user number in buffering area while stablizing; $\stackrel{\&OverBar;}{R_{\mathrm{hdof}}}={\mathrm{\&lambda;}}_p\&CenterDot;{\mathrm{\&Sigma;}}_{i+j=M,j>0}{p}_{\mathrm{ev}}^{(i,j,k)}\&CenterDot;p_{i,j,k}$ Represent that cognitive user is switched to the Mean Speed in buffering area;

Iv) the throughput of cognitive user:

R _th(Ρ _ad,Ρ _ev)＝λ _s·(1-P _block)·(1-P _drop)。

6. method according to claim 5, wherein, described optimization problem is described as:

max R _th(Ρ _ad,Ρ _ev)

$s.t.\left\{\begin{array}{c}{P}_{\mathrm{drop}}(P_{\mathrm{ad}},P_{\mathrm{ev}})\&le;P_d^{*}\\ D_{\mathrm{ho}}(P_{\mathrm{ad}},P_{\mathrm{ev}})\&le;D_h^{*}\\ 0\&le;P_{\mathrm{ad}}\&le;1\\ 0\&le;P_{\mathrm{ad}}\&le;1\end{array}\right.,$

Wherein,

with

be respectively the drop rate of cognitive user and the threshold value of average handover delay.

Images