CN105282746A - Cognitive radio network frequency spectrum distribution method based on embedded particle swarm gaming - Google Patents

Abstract

The invention discloses a cognitive radio network frequency spectrum distribution method based on embedded particle swarm gaming. The method includes extracting the characteristic information of cognitive users; establishing a non-cooperative gaming model based on the characteristics of the cognitive users; resolving the gaming model by means of inner and outer particle swarm algorithms, namely, an embedded particle swarm algorithm; resolving and calculating the channel preference set of each cognitive user in the gaming model by means of the inner particle swarm algorithm based on the characteristics of the cognitive user; resolving the non-cooperative gaming model by means of the outer particle swarm algorithm based on the channel preference set of each cognitive user, and outputting the optimal frequency spectrum distribution result and a corresponding power matrix. Under the underplay frequency spectrum sharing manner, an optimal frequency spectrum distribution scheme is obtained by means of the embedded particle swarm algorithm. The method provided combines the game theory method and the particle swarm algorithm, the channel distribution and power control, rapidly distributes frequency spectrum sources, and moreover, maximizes the benefits of the cognitive users and the system profits within the maximum tolerable interference threshold of a main user.

Description

Cognitive wireless network spectrum allocation method based on embedded particle swarm game

The technical field is as follows:

the invention belongs to the technical field of wireless communication, and particularly relates to a cognitive wireless network spectrum allocation method based on an embedded particle swarm game.

Background art:

the cognitive radio is a novel communication technology, can sense the surrounding radio environment, and adaptively changes parameters such as carrier frequency, transmission power and modulation technology through intelligent learning, thereby improving the spectrum utilization rate. In the field of radio communications, management and utilization of spectrum resources are very important, and a static spectrum allocation method commonly used in various countries allocates spectrum to authorized users, while other unauthorized users are not authorized to use the spectrum. When the authorized user does not use the spectrum, the waste of spectrum resources is caused, and the utilization rate of the spectrum is low. Meanwhile, with the development of social informatization, radio communication is becoming more and more popular, and radio spectrum resources are in short supply. In recent years, cognitive radio is becoming a key technology for solving the problem of shortage of spectrum resources.

A patent application "optimal algorithm for power control based on improved particle swarm in cognitive radio network" proposed by the university of ludong (patent application No. 201010532702.6, publication No. CN101982992A) discloses a cognitive radio network power control algorithm based on particle swarm. The algorithm comprises the following implementation steps: the first step, the iteration times of the algorithm, the positions and the speeds of the particles and the basic parameters of the particle swarm are initialized. And secondly, calculating a fitness function value, wherein the position of each individual particle is an initial optimal position, and the particle with the optimal function value in the population is an initial optimal population position. And thirdly, executing search based on a PSO algorithm, updating the optimal positions of the particles and the population, and updating the speed and the position of the particles by using a particle swarm basic formula. And fourthly, setting a termination standard. The method has the disadvantages that only the optimal power control of wireless communication is considered, and the optimal allocation of channels is not considered. In practical scenarios, both channel allocation and power control need to be considered.

In "research on cognitive radio resource allocation algorithm based on particle swarm optimization", a cognitive radio resource allocation algorithm based on particle swarm optimization is proposed by any cellcept. The algorithm comprises the following implementation steps: in a first step, an available matrix, a benefit matrix, an interference matrix and a non-interference allocation matrix are initialized. And secondly, encoding the distribution scheme and the particles. And thirdly, initializing iteration times, the speed and the position of the particles and basic parameters of the particle swarm. And fourthly, calculating a fitness function value, and comparing the fitness function value with the individual historical optimal value and the population optimal value of the particles. And fifthly, updating the position and the speed of the particles. And setting an iteration termination condition. The method has the defect that only the overlay spectrum sharing mode is considered. In practice, however, underlay spectrum sharing is more prevalent. And as the number of spectrum and cognitive users increase, the complexity of the algorithm increases significantly.

The information disclosed in this background section is only for enhancement of understanding of the general background of the invention and should not be taken as an acknowledgement or any form of suggestion that this information forms the prior art already known to a person skilled in the art.

The invention content is as follows:

the invention aims to provide a cognitive wireless network spectrum allocation method based on an embedded particle swarm game, which solves a non-cooperative game model of a cognitive user competing for a spectrum by using an embedded particle swarm algorithm. The algorithm optimizes channel allocation and power control in an underlay spectrum sharing mode. The non-cooperative game model considers the interference threshold value of the cognitive user to the main user and the limitation of the maximum power requirement. The invention can quickly realize the cognitive radio frequency spectrum allocation and give consideration to the requirements of the master user and the cognitive user, thereby overcoming the defects in the prior art.

To achieve the above object, the present invention provides

The cognitive wireless network spectrum allocation method based on the embedded particle swarm game comprises the following steps:

step 1, extracting feature information of a cognitive user;

step 2, establishing a non-cooperative game model according to the characteristics of the cognitive user;

step 3, solving the game model by an inner-layer particle swarm algorithm and an outer-layer particle swarm algorithm, namely an embedded particle swarm algorithm;

3-1, solving and calculating a channel preference set of each cognitive user in the game model by an inner-layer particle swarm algorithm according to the characteristics of each cognitive user;

and 3-2, solving a non-cooperative game model by an outer-layer particle swarm algorithm according to the channel preference set of each cognitive user, and outputting an optimal spectrum allocation result and a corresponding power matrix.

Preferably, in the technical scheme, the cognitive user feature information extracted in step 1 is information of a distance, a relative speed and a geographical position between a transmitter and a receiver of the cognitive user collected through a control channel.

Preferably, in the technical scheme, the mathematical formula of the non-cooperative game model in the step 2 is as follows:

＝{N,{S_i}_i∈N,{U_i}_i∈N}

where N is a finite set of participants, S_iIs a strategy set of a participant i, and defines a strategy space as S- × S_iI ∈ N, then U_tS → R is the set of utility functions; at each participant i in the game, the utility function U_iIs about S_iAnd adversary strategy set S_-iAs a function of (c).

Preferably, in the technical scheme, each game participant makes a decision independently and is influenced by the decisions of other participants, and the game problem analysis is most critical to analyzing the Nash equilibrium of the game model; when the game reaches Nash equilibrium, any game participant can not change the action strategy of the game participant; thus, for a set of policies S ═ S for game participants₁,s₂,…,s_NIf and only if U_i(S)≥U_i(s_i',s_-i)，s_i'∈s_iAnd when the game reaches Nash equilibrium, the group of action strategies is Nash equilibrium.

Preferably, in the technical scheme, the inner-layer particle swarm algorithm in the step 3-1 comprises the following steps:

(1) the fitness value of each particle is calculated,

$Fitness=u_n=B\log =(1+\frac{p_n^{st}{G}_{{\mathrm{nn}}^{\′}m}^{st}}{{\mathrm{\σ}}_0+p^{pt}{G}_{{\mathrm{nn}}^{\′}m}^{pt}});$

wherein,b is the channel bandwidth, σ₀For background noise power, p^ptAndrespectively corresponding to main users PR and CR_nThe transmission power of the antenna is set to be,andcorresponding to the gains of PR and CR on channel m respectively,for the main user to CR_nInterference of

(2) Numbering the cognitive users and the channels, and updating the speed and the position of each particle by using a particle updating formula;

the discrete particle i updates the formula as follows:

$v_i^{t+1}=c_1\·v_i^t\⊕c_2\·\left(P_i^t{\mathrm{\Θx}}_i^t\right)\⊕c_3\·\left(P_g^t{\mathrm{\Θx}}_i^t\right)$

$x_i^{t+1}=x_i^{t+1}\⊕v_i^{t+1};$

wherein the dimension of the particle is 1 and the position of the particleRepresents a channel number;

(3) and if the current iteration times reach the preset maximum times or the final result is smaller than the preset convergence precision, stopping the iteration and outputting the channel preference set phi of each current cognitive user.

Preferably, in the technical scheme, the outer-layer particle swarm algorithm in the step 3-2 comprises the following steps:

(1) the fitness value of each particle is calculated,

$Fitness=U=\underset{n=1}{\overset{N}{\Σ}}\underset{m=1}{\overset{M}{\Σ}}{a}_{n,m}{u}_{n,m};$

wherein a is_n,m1 represents CR_nOccupation of the channel m, a_n,mWhen 0 denotes CR_nNo occupied channel m;

$u_{n,m}=Blog(1+\frac{1.5}{\ln (0.2/{\mathrm{BER}}_n^{tar})}\frac{p_n^{st}{G}_{{\mathrm{nn}}^{\′}m}^{st}}{{\mathrm{\σ}}_0+\underset{i=1,i\≠n}{\overset{N}{\Σ}}{p}_i^{st}{a}_{n,m}{G}_{{\mathrm{in}}^{\′}m}^{st}+p^{pt}{G}_{{\mathrm{nn}}^{\′}m}^{pt}});$

wherein,b is the channel bandwidth, σ₀For background noise power, p^ptAndrespectively correspond to PR and CR_nThe transmission power of the antenna is set to be,is CR_nThe target bit error rate of (a) is,andcorresponding to the gains of PR and CR on channel m respectively,for the main user to CR_nThe interference of (a) with the other,respectively correspond to CR_iThe transmission power of the antenna is set to be,represents CR_iTransmitter to CR_nThe gain of the receiver over the channel m,representing other cognitive user pairs CR_nInterference of (2);

(2) numbering the cognitive users and the channels, and updating the speed and the position of each particle by using a particle updating formula; because the channel number is discrete and the power is continuous, different updating formulas are respectively adopted for discrete particles and continuous particles;

the particle i update formula is as follows:

wherein, when d is an odd number,is shown asNumbering channels where the cognitive users are after t iterations; when d is an even number, the number of the transition metal,is shown asAnd (4) the transmitting power of the channel where each cognitive user is located after t iterations.

(3) And if the current iteration times reach the preset maximum times or the final result is smaller than the preset convergence precision, stopping the iteration and outputting the current optimal spectrum allocation result and the corresponding power matrix.

Compared with the prior art, the invention has the following beneficial effects:

and under the underlay frequency spectrum sharing mode, an optimal frequency spectrum allocation scheme is obtained through an embedded particle swarm algorithm. The game theory method and the particle swarm algorithm are combined, channel allocation and power control are combined, spectrum resources are allocated rapidly, and the benefit and the system benefit of the cognitive user are maximized within the maximum interference tolerance threshold of the master user.

Description of the drawings:

FIG. 1 is a schematic flow chart of an implementation of the cognitive wireless network spectrum allocation method based on embedded particle swarm gaming of the present invention;

FIG. 2 is a schematic diagram of an algorithm flow of the cognitive wireless network spectrum allocation method based on the embedded particle swarm game;

the specific implementation mode is as follows:

the following detailed description of specific embodiments of the invention is provided, but it should be understood that the scope of the invention is not limited to the specific embodiments.

Throughout the specification and claims, unless explicitly stated otherwise, the word "comprise", or variations such as "comprises" or "comprising", will be understood to imply the inclusion of a stated element or component but not the exclusion of any other element or component.

As shown in fig. 1, a cognitive wireless network spectrum allocation method based on an embedded particle swarm game comprises the following steps:

step 1, extracting feature information of a cognitive user;

step 2, establishing a non-cooperative game model according to the characteristics of the cognitive user;

step 3, solving the game model by an inner-layer particle swarm algorithm and an outer-layer particle swarm algorithm, namely an embedded particle swarm algorithm;

3-1, solving and calculating a channel preference set of each cognitive user in the game model by an inner-layer particle swarm algorithm according to the characteristics of each cognitive user;

and 3-2, solving a non-cooperative game model by an outer-layer particle swarm algorithm according to the channel preference set of each cognitive user, and outputting an optimal spectrum allocation result and a corresponding power matrix.

As shown in FIG. 2, pbest represents the optimum of the particles in the particle group, and gbest represents the global optimum of the particle group.

In step 1, the cognitive user feature information is extracted, that is, the information of the distance, the relative speed and the geographical position between the transmitter and the receiver of the cognitive user is collected through a control channel.

The mathematical formula of the non-cooperative game model in the step 2 is as follows:

＝{N,{S_i}_i∈N,{U_i}_i∈N}

where N is a finite set of participants, S_iIs a strategy set of a participant i, and defines a strategy space as S- × S_iI ∈ N, then U_tS → R is the set of utility functions; at each participant i in the game, the utility function U_iIs about S_iAnd adversary strategy set S_-iAs a function of (c).

Due to each timeOne game participant independently makes decisions and is influenced by the decisions of other participants, and the most important point of the game problem analysis is to analyze the Nash equilibrium of the game model; when the game reaches Nash equilibrium, any game participant can not change the action strategy of the game participant; thus, for a set of policies S ═ S for game participants₁,s₂,…,s_NIf and only if U_i(S)≥U_i(s_i',s_-i)，s_i'∈s_iAnd when the game reaches Nash equilibrium, the group of action strategies is Nash equilibrium.

With the goal of maximizing the overall utility of the system, we can describe the game model as an optimization problem:

A^opt＝argmaxU^tot(A)

constraint conditions are as follows:

$\underset{m=1}{\overset{M}{\Σ}}{a}_{n,m}=0or1,a_{n,m}\∈\{0,1\},\∀n\∈\[1,N\]$

$a_{n,m}(R_{req}^{st}-R_{n,m}^{st})\≤0,\∀n\∈\[1,N\],\∀m\∈\[1,M\]$

$R_{req}^{pt}\≤R_m^{pt},\∀m\∈\[1,M\]$

$p_n^{st}<p^{\max },\&ForAll;n\&Element;\&lsqb;1,N\&rsqb;;$

wherein,andrespectively representing the minimum transmission rate requirements of the primary user and the cognitive user,andrespectively representing primary users PR_mAnd cognitive user CR_nActual transmission rate on channel m, p^maxThe maximum transmission power of the cognitive user.

The power of the cognitive user depends on the transmission power of the master user; for each PR_mThe rate requirement is satisfied, and it is ensured that the bit error rate is smaller than the bit error rate upper limit, then the maximum interference allowed on the occupied channel can be represented as:

$I\left(m\right)=\frac{p^{pt}{G}_{mm}^{pt}}{(2^{R_p^{req}}-1)\&times;\left(\ln \right(0.2/{\mathrm{BER}}_m^{tar}\left)\right)/1.5}$

wherein,indicating the target bit error rate of the primary user on the channel m.

Then, the transmission power of the cognitive user satisfies:

$\underset{n=1}{\overset{N}{\&Sigma;}}{p}^{st}\left(n\right)a_{n,m}{G}_{{\mathrm{nn}}^{\&prime;}}^{st}\&le;I\left(m\right)$

the model provides a spectrum allocation scheme of the cognitive user on the available channel, so that the cognitive user does not influence the transmission of the main user when using the spectrum, and the overall utility of the system is maximized.

The inner-layer particle swarm algorithm in the step 3-1 comprises the following steps:

(1) the fitness value of each particle is calculated,

$Fitness=u_n=B\log =(1+\frac{p_n^{st}{G}_{{\mathrm{nn}}^{\&prime;}m}^{st}}{{\mathrm{\&sigma;}}_0+p^{pt}{G}_{{\mathrm{nn}}^{\&prime;}m}^{pt}});$

wherein,b is the channel bandwidth, σ₀For background noise power, p^ptAndrespectively corresponding to main users PR and CR_nThe transmission power of the antenna is set to be,andcorresponding to the gains of PR and CR on channel m respectively,for the main user to CR_nInterference of (2);

(2) numbering the cognitive users and the channels, and updating the speed and the position of each particle by using a particle updating formula;

the discrete particle i updates the formula as follows:

$v_i^{t+1}=c_1\&CenterDot;v_i^t\&CirclePlus;c_2\&CenterDot;\left(P_i^t{\mathrm{\&Theta;x}}_i^t\right)\&CirclePlus;c_3\&CenterDot;\left(P_g^t{\mathrm{\&Theta;x}}_i^t\right)$

$x_i^{t+1}=x_i^t\&CirclePlus;v_i^{t+1};$

wherein the dimension of the particle is 1 and the position of the particleRepresents a channel number;

(3) and if the current iteration times reach the preset maximum times or the final result is smaller than the preset convergence precision, stopping the iteration and outputting the channel preference set phi of each current cognitive user.

The outer-layer particle swarm algorithm in the step 3-2 comprises the following steps:

(1) the fitness value of each particle is calculated,

$Fitness=U=\underset{n=1}{\overset{N}{\&Sigma;}}\underset{m=1}{\overset{M}{\&Sigma;}}{a}_{n,m}{u}_{n,m};$

wherein a is_n,m1 represents CR_nOccupation of the channel m, a_n,mWhen 0 denotes CR_nNo occupied channel m;

$u_{n,m}=Blog(1+\frac{1.5}{\ln (0.2/{\mathrm{BER}}_n^{tar})}\frac{p_n^{st}{G}_{{\mathrm{nn}}^{\&prime;}m}^{st}}{{\mathrm{\&sigma;}}_0+\underset{i=1,i\&NotEqual;n}{\overset{N}{\&Sigma;}}{p}_i^{st}{a}_{n,m}{G}_{{\mathrm{in}}^{\&prime;}m}^{st}+p^{pt}{G}_{{\mathrm{nn}}^{\&prime;}m}^{pt}});$

wherein,b is the channel bandwidth, σ₀For background noise power, p^ptAndrespectively correspond to PR and CR_nTransmit power of，Is CR_nThe target bit error rate of (a) is,andcorresponding to the gains of PR and CR on channel m respectively,for the main user to CR_nThe interference of (a) with the other,respectively correspond to CR_iThe transmission power of the antenna is set to be,represents CR_iTransmitter to CR_nThe gain of the receiver over the channel m,representing other cognitive user pairs CR_nInterference of (2);

(2) numbering the cognitive users and the channels, and updating the speed and the position of each particle by using a particle updating formula; because the channel number is discrete and the power is continuous, different updating formulas are respectively adopted for discrete particles and continuous particles;

the particle i update formula is as follows:

wherein, when d is an odd number,is shown asNumbering channels where the cognitive users are after t iterations; when d is an even number, the number of the transition metal,is shown asAnd (4) the transmitting power of the channel where each cognitive user is located after t iterations.

(3) And if the current iteration times reach the preset maximum times or the final result is smaller than the preset convergence precision, stopping the iteration and outputting the current optimal spectrum allocation result and the corresponding power matrix.

The foregoing descriptions of specific exemplary embodiments of the present invention have been presented for purposes of illustration and description. It is not intended to limit the invention to the precise form disclosed, and obviously many modifications and variations are possible in light of the above teaching. The exemplary embodiments were chosen and described in order to explain certain principles of the invention and its practical application to enable one skilled in the art to make and use various exemplary embodiments of the invention and various alternatives and modifications as are suited to the particular use contemplated. It is intended that the scope of the invention be defined by the claims and their equivalents.

Claims (6)

1. The cognitive wireless network spectrum allocation method based on the embedded particle swarm game is characterized by comprising the following steps: the method comprises the following steps:

step 1, extracting feature information of a cognitive user;

step 2, establishing a non-cooperative game model according to the characteristics of the cognitive user;

step 3, solving the game model by an inner-layer particle swarm algorithm and an outer-layer particle swarm algorithm, namely an embedded particle swarm algorithm;

3-1, solving and calculating a channel preference set of each cognitive user in the game model by an inner-layer particle swarm algorithm according to the characteristics of each cognitive user;

and 3-2, solving a non-cooperative game model by an outer-layer particle swarm algorithm according to the channel preference set of each cognitive user, and outputting an optimal spectrum allocation result and a corresponding power matrix.

2. The cognitive wireless network spectrum allocation method based on the embedded particle swarm game is characterized in that: in step 1, the cognitive user feature information is extracted, that is, the information of the distance, the relative speed and the geographical position between the transmitter and the receiver of the cognitive user is collected through a control channel.

3. The cognitive wireless network spectrum allocation method based on the embedded particle swarm game is characterized in that: the mathematical formula of the non-cooperative game model in the step 2 is as follows:

＝{N,{S_i}_i∈N,{U_i}_i∈N}

where N is a finite set of participants, S_iIs a strategy set of a participant i, and defines a strategy space as S- × S_iI ∈ N, then U_tS → R is the set of utility functions; at each participant i in the game, the utility function U_iIs about S_iAnd adversary strategy set S_-iAs a function of (c).

4. The cognitive wireless network spectrum allocation method based on the embedded particle swarm game is characterized in that: when a set of policies S ═ S₁,s₂,…,s_NIf and only if U_i(S)≥U_i(s_i',s_-i)，s_i'∈s_iAnd when the game reaches Nash equilibrium, the group of action strategies is Nash equilibrium.

5. The cognitive wireless network spectrum allocation method based on the embedded particle swarm game is characterized in that: the inner-layer particle swarm algorithm in the step 3-1 comprises the following steps:

(1) the fitness value of each particle is calculated,

$Fitness=u_n=Blog(1+\frac{p_n^{st}{G}_{{\mathrm{nn}}^{\&prime;}m}^{st}}{{\mathrm{\&sigma;}}_0+p^{pt}{G}_{{\mathrm{nn}}^{\&prime;}m}^{pt}});$

wherein,b is the channel bandwidth, σ₀For background noise power, p^ptAndrespectively corresponding to main users PR and CR_nThe transmission power of the antenna is set to be,andcorresponding to the gains of PR and CR on channel m respectively,for the main user to CR_nInterference of (2);

(2) numbering the cognitive users and the channels, and updating the speed and the position of each particle by using a particle updating formula;

the discrete particle i updates the formula as follows:

$v_i^{t+1}=c_1\&CenterDot;v_i^t\&CirclePlus;c_2\&CenterDot;\left(P_i^t{\mathrm{\&Theta;x}}_i^t\right)\&CirclePlus;c_3\&CenterDot;\left(P_g^t{\mathrm{\&Theta;x}}_i^t\right)$

$x_i^{t+1}=x_i^{t+1}\&CirclePlus;v_i^{t+1};$

wherein the dimension of the particle is 1 and the position of the particleRepresents a channel number;

(3) and if the current iteration times reach the preset maximum times or the final result is smaller than the preset convergence precision, stopping the iteration and outputting the channel preference set phi of each current cognitive user.

6. The cognitive wireless network spectrum allocation method based on the embedded particle swarm game is characterized in that: the outer-layer particle swarm algorithm in the step 3-2 comprises the following steps:

(1) the fitness value of each particle is calculated,

$Fitness=U=\underset{n=1}{\overset{N}{\&Sigma;}}\underset{m=1}{\overset{M}{\&Sigma;}}{a}_{n,m}{u}_{n,m};$

wherein a is_n,m1 represents CR_nOccupation of the channel m, a_n,mWhen 0 denotes CR_nNo occupied channel m;

$u_{n,m}=Blog(1+\frac{1.5}{\ln (0.2/{\mathrm{BER}}_n^{tar})}\frac{p_n^{st}{G}_{{\mathrm{nn}}^{\&prime;}m}^{st}}{{\mathrm{\&sigma;}}_0+\underset{i=1,i\&NotEqual;n}{\overset{N}{\&Sigma;}}{p}_i^{st}{a}_{n,m}{G}_{{\mathrm{in}}^{\&prime;}m}^{st}+p^{pt}{G}_{{\mathrm{nn}}^{\&prime;}m}^{pt}});$

wherein,b is the channel bandwidth, σ₀For background noise power, p^ptAndrespectively correspond to PR and CR_nThe transmission power of the antenna is set to be,is CR_nThe target bit error rate of (a) is,andcorresponding to the gains of PR and CR on channel m respectively,for the main user to CR_nThe interference of (a) with the other,respectively correspond to CR_iThe transmission power of the antenna is set to be,represents CR_iTransmitter to CR_nThe gain of the receiver over the channel m,representing other cognitive user pairs CR_nInterference of (2);

(2) numbering the cognitive users and the channels, and updating the speed and the position of each particle by using a particle updating formula; because the channel number is discrete and the power is continuous, different updating formulas are respectively adopted for discrete particles and continuous particles;

the particle i update formula is as follows:

wherein, when d is an odd number,is shown asNumbering channels where the cognitive users are after t iterations; when d is an even number, the number of the transition metal,is shown asAnd (4) the transmitting power of the channel where each cognitive user is located after t iterations.

(3) And if the current iteration times reach the preset maximum times or the final result is smaller than the preset convergence precision, stopping the iteration and outputting the current optimal spectrum allocation result and the corresponding power matrix.