CN103857046A - Self-adaptive resource distribution method of cognition OFDM network based on spectrum filling - Google Patents

Abstract

The invention discloses a self-adaptive resource distribution method of a cognition OFDM network based on spectrum filling. The self-adaptive resource distribution method of the cognition OFDM network based on spectrum filling comprises the following steps that a cognition base station detects a spectrum hole and the noise power on a null sub-carrier; the cognition base station obtains channel state information of each cognition user and a real-time service transmission speed requirement; the cognition base station distributes the frequency and power resources in the cognition OFDM network based on spectrum filling by utilizing a quantum particle swarm method; the cognition base station initializes the parameters of the quantum particle swarm method; each particle position vector in a group is initialized and comprises the frequency and power resources; the adaptation value of each particle in the group and the particle with the minimum adaptation degree in the group are determined; the position of each particle in the group is determined again according to each particle position vector and the position vector of the particle with the minimum adaptation degree. The self-adaptive resource distribution method of the cognition OFDM network based on spectrum filling distributes the resources from the two dimensions of frequency and power and can efficiently use the wireless resources in the cognition OFDM network based on spectrum filling.

Description

The cognitive OFDM network self-adapting resource allocation methods of filling based on frequency spectrum

Technical field

The invention belongs to communication technical field, particularly a kind of cognitive OFDM network self-adapting resource allocation methods of filling based on frequency spectrum.

Background technology

In recent years, along with developing rapidly of wireless communication technology, make limited frequency spectrum resource increasingly deficient.But the research report of FCC shows: the service condition of frequency spectrum resource is extremely uneven, is embodied in: the mandate frequency range availability of frequency spectrum of distributing at present only has 15%-85%, and the availability of frequency spectrum of the following frequency range of 3GHz only has 35%.This is because static spectrum allocation may mode for a long time makes the frequency spectrum of a lot of suitable transmission of wireless signals often in idle state, has therefore brought up the significant wastage of frequency spectrum resource.Under this background, the concept that J.Mitola doctor III proposed cognitive radio in 1999; It is by continuous cognitive PERCOM peripheral communication environmental information, and the frequency spectrum cavity-pocket in identification current environment has been realized other and authorized the recycling of frequency spectrum resource on spatial domain, time domain and frequency domain, thereby effectively improved the availability of frequency spectrum.On the basis of J.Mitola doctor III research work, the researcher of Virginia, US engineering college clearly proposed the concept of cognition wireless network first in 2005, cognition wireless network is the expansion of cognitive radio technology, it not only payes attention to the reasonable utilization of frequency spectrum resource, more focuses on the optimization of overall performance of network.

The frequency spectrum cavity-pocket of finding due to cognition wireless network may be positioned at very wide spectral range, and is discontinuous, and therefore researcher improves OFDM technology, has proposed to be applicable to the NC-OFDM technology of cognition network.Access way based on NC-OFDM technology is not that all subcarriers in an OFDM symbol can be used by cognitive user, authorized user can take wherein parton carrier wave, for fear of authorized user is caused to interference, these occupied subcarriers can not be used for transmitting data, by the cognition wireless network based on NC-OFDM technology, referred to as the cognitive OFDM network of filling based on frequency spectrum.

Radio Resource in the cognitive OFDM network of filling based on frequency spectrum is very deficient, and these Radio Resources become the main aspect of restriction system performance, therefore need in the cognitive OFDM network of filling based on frequency spectrum, reasonably distribute these Radio Resources.The cognitive OFDM Resource Allocation in Networks technology of filling based on frequency spectrum is dynamic sub carrier and the power distributing technique from the angle research cognition network of interference coordination between cognition network and authorisation network, its objective is in the situation that not affecting the proper communication of authorized user network, effectively utilize as far as possible these Radio Resources and reach higher spectrum efficiency, also need to ensure the satisfaction of cognitive user simultaneously.But compared with traditional OFDM network, the cognitive OFDM network of filling based on frequency spectrum has many special natures, make existing resource allocation methods be difficult to meet its requirement, the cognitive OFDM Resource Allocation in Networks method of filling based on frequency spectrum that especially meets green communications requirement and real time business demand still lacks research.

Resource allocation methods based on quality of service in patent 1(cognitive OFDM system, Nanjing Univ. of Posts and Telecommunications, publication number CN102291352A, application number CN201110236043.6, applying date 2011.08.17) resource allocation methods in a kind of cognitive OFDM system based on quality of service disclosed, the method is under the total transmitted power of inferior user and primary user's interference threshold restriction, according to inferior user's demand percentage, real time business is carried out to adaptive bit and power allocation, until meet its rate requirement, finally remaining resource is distributed to non-real-time service.The method is the resource allocation methods for the cognitive OFDM network design of Underlay pattern, not the cognitive OFDM Resource Allocation in Networks method of design based on frequency spectrum fill pattern.

Patent 2(channel allocation method limiting based on conflict threshold, Nanjing Univ. of Posts and Telecommunications, publication number CN103326984A, application number CN201310279834.6, applying date 2013.07.04) a kind of channel allocation method based on conflict threshold restriction disclosed, the method combines conflict threshold with effective throughput, according to the transmission data package size of each sub-channels and its idle probability distribution parameters, selection meets the minimal number subchannel that its data requires, thereby allows more time user send data.But the method is only considered the dynamic assignment of resource from frequency category, can not further optimize resource utilization from power dimension.In addition, above-mentioned two methods are all taking maximize throughput as optimization aim, can not meet the green communications resource distribution requirements of cognitive OFDM network.

Summary of the invention

The object of the present invention is to provide adaptive resource allocation method in a kind of cognitive OFDM network efficient, that fill based on frequency spectrum reliably, from frequency and two dimension dynamic on-demand Resources allocation of power, fully excavate available frequency spectrum cavity-pocket resource in the cognitive OFDM network of filling based on frequency spectrum.

The technical solution that realizes the object of the invention is: a kind of cognitive OFDM network self-adapting resource allocation methods of filling based on frequency spectrum, comprises the following steps:

Step 1, the noise power in frequency spectrum cavity-pocket and idle sub-carrier is detected in cognitive base station;

Step 2, cognitive base station obtains each cognitive user channel condition information and real-time service transmission rate requirement;

Step 3, cognitive base station adopts quantum particle swarm method to distribute frequency and power resource in the cognitive OFDM network of filling based on frequency spectrum, comprises the following steps:

Step 3.1, the parameter of cognitive base station initialization quantum particle swarm method;

Step 3.2, the each particle position vector in initialization population, this position vector comprises frequency and power resource;

Step 3.3, determines the particle of fitness minimum in the adaptive value of each particle in population and population;

Step 3.4, redefines the position of each particle in population according to the particle position vector of each particle position vector sum fitness minimum;

Step 3.5, repeating step 3.3～step 3.4N _ginferior, the optimal solution of output frequency and power resource allocation, N _grepresent the iterations of quantum particle swarm method.

Compared with prior art, its remarkable advantage is in the present invention: (1) adopts subcarrier and power distribution method flexibly, meets the demand that cognitive user real time business resource is distributed; (2) cognitive OFDM network sub-carriers and the power resource based on frequency spectrum filling from frequency and two dimension dynamic assignment of power, has fully excavated available frequency spectrum cavity-pocket resource in the cognitive OFDM network of filling based on frequency spectrum; (3) can efficiently utilize frequency and the power resource in the cognitive OFDM network of filling based on frequency spectrum, for promoting the cognitive OFDM network green communications of filling based on frequency spectrum that technical support is provided.

Brief description of the drawings

Fig. 1 is the flow chart that the present invention is based on the cognitive OFDM network self-adapting resource allocation methods of frequency spectrum filling.

Fig. 2 is the schematic diagram that in the present invention, cognitive base station obtains each cognitive user channel condition information and real-time service transmission rate requirement.

Fig. 3 is the flow chart that in the present invention, cognitive base station adopts quantum particle swarm method to carry out resource distribution.

Fig. 4 is that the cognitive OFDM network sub-carriers that the present invention is based on frequency spectrum filling takies situation.

Embodiment

Below in conjunction with drawings and the specific embodiments, the present invention is described in further detail.

In conjunction with Fig. 1, the present invention is based on the cognitive OFDM network self-adapting resource allocation methods that frequency spectrum is filled, comprise the following steps:

Step 1, the noise power in frequency spectrum cavity-pocket and idle sub-carrier is detected in cognitive base station, be specially: the radio spectrum resources in authorisation network is detected in the cognitive base station of filling in cognitive OFDM network based on frequency spectrum, obtain the spectrum utilization state information in current authorisation network through the message processing module analysis in cognitive base station, formation can characterize the frequency spectrum detection report of actual parameter and the reflection frequency spectrum cavity-pocket of frequency spectrum cavity-pocket, cognitive base station is sent to frequency spectrum detection report based on frequency spectrum and is filled the cognitive user in cognitive OFDM network by the mode of broadcast, the noise power in idle sub-carrier is detected according to the examining report of frequency spectrum cavity-pocket in cognitive base station.

Step 2, cognitive base station obtains each cognitive user channel condition information and real-time service transmission rate requirement, be specially: cognitive user estimates to fill based on frequency spectrum the channel condition information of the last OFDM symbol of cognitive OFDM network down link by the message processing module of cognitive terminal, cognitive user by fill based on frequency spectrum cognitive OFDM network uplink link by estimate information feedback give cognitive base station, cognitive user feeds back to cognitive base station by fill cognitive OFDM network uplink link based on frequency spectrum by real-time service transmission rate requirement, as shown in Figure 2.

Step 3, cognitive base station adopts quantum particle swarm method to distribute frequency and power resource in the cognitive OFDM network of filling based on frequency spectrum, comprises the following steps in conjunction with Fig. 3:

Step 3.1, the parameter of cognitive base station initialization quantum particle swarm method, is specially:

(1) number of particles N in initialization quantum particle swarm method population _pand N _pthe iterations N of ∈ [500,600], quantum particle swarm method _gand N _g∈ [1000,1200], the moment t cognitive user m subcarrier on subcarrier k distributes indicator variable

and

the power division indicator variable of moment t cognitive user m on subcarrier k

and

$p_{m,k}^t\∈\left[\mathrm{0,10}\right];$

(2) adopt formula (1) to determine moment t cognitive user m transmission rate on subcarrier k

$b_{m,k}^t=\frac{W}{K}{\log }_2(1+\frac{p_{m,k}^t{h}_{m,k}^t}{N_0})---\left(1\right)$

In formula (1), W represents the cognitive OFDM network bandwidth of filling based on frequency spectrum, and K represents the cognitive OFDM network number of sub carrier wave of filling based on frequency spectrum, N ₀represent the noise power on subcarrier,

represent that moment t cognitive user m gains at subcarrier k upper signal channel; T represents time index, and m represents cognitive user index, and k represents sub-carrier indices;

(3) the desired positions vector of i particle when the gen time iteration of initialization

desired positions vector with all particles in the gen time iteration population

initialization iterations sequence number gen=1.

Step 3.2, the each particle position vector in initialization population, this position vector comprises frequency and power resource, is specially:

Each particle position vector x=[c, p] in initialization population, wherein vectorial c represents that subcarrier distributes indicator variable, vectorial c meets formula (2) and formula (3):

$c_{m,k}^t=\left\{\begin{array}{c}\left[\mathrm{0,1}\right],k\∈K_t\\ \left\{0\right\},\mathrm{else}\end{array}\right.,\∀m,k,t---\left(2\right)$

$\underset{m=1}{\overset{M}{\mathrm{\Σ}}}{c}_{m,k}\≤1,c_{m,k}\≥0,\∀k,t---\left(3\right)$

Vector p represents power division indicator variable, and vectorial p meets formula (4):

$\underset{k=1}{\overset{K}{\mathrm{\Σ}}}{c}_{m,k}^t{b}_{m,k}^t\≥R_m^{\min }---\left(4\right)$

Wherein, K _trepresent the idle sub-carrier set that moment t can use,

the real-time service transmission rate requirement that represents cognitive user m, M represents the sum of cognitive user.

Step 3.3, determine and be specially the particle of fitness minimum in the adaptive value of each particle in population and population:

Adopt formula (5) to determine the adaptive value f (x of each particle in population _i(t));

$f\left(x_i\left(\mathrm{gen}\right)\right)=\underset{m=1}{\overset{M}{\mathrm{\Σ}}}\underset{k=1}{\overset{K}{\mathrm{\Σ}}}{c}_{m,k}^t{p}_{m,k}^t---\left(5\right)$

If $f\left(x_i\left(\mathrm{gen}\right)\right)<f\left(x_i^{\mathrm{best}}\left(\mathrm{gen}\right)\right),$ Order $x_i^{\mathrm{best}}\left(\mathrm{gen}\right)=x_i\left(\mathrm{gen}\right);$

If $f\left(x_i\left(\mathrm{gen}\right)\right)<f\left(x_g^{\mathrm{best}}\left(\mathrm{gen}\right)\right),$ Order $x_g^{\mathrm{best}}\left(\mathrm{gen}\right)=x_i\left(\mathrm{gen}\right);$

Wherein, f (x _i(gen)) represent the adaptive value of i particle the gen time iteration,

represent the adaptive value of i particle desired positions vector in the time of the gen time iteration,

the adaptive value of desired positions vector particle in population while being illustrated in the gen time iteration.

Step 3.4, redefines the position of each particle in population according to the particle position vector of each particle position vector sum fitness minimum, be specially:

The latest position that adopts each particle in formula (6) Population Regeneration is the position x of i particle in the time of the gen+1 time iteration _i(gen+1):

$x_i(\mathrm{gen}+1)=\left\{\begin{array}{cc}{x}_{\mathrm{op}}\left(\mathrm{gen}\right)+\mathrm{\&alpha;}|x_{\mathrm{best}}-x_i\left(\mathrm{gen}\right)|1n(1/\mathrm{\&beta;}),\mathrm{if}& \mathrm{\&mu;}\&GreaterEqual;0.5\\ x_{\mathrm{op}}\left(\mathrm{gen}\right)-\mathrm{\&alpha;}|x_{\mathrm{best}}-x_i\left(\mathrm{gen}\right)|1n(1/\mathrm{\&beta;}),\mathrm{if}& \mathrm{\&mu;}<0.5\end{array}---\left(6\right)\right.$

Wherein, x _op(gen) represent

with between random site, and adopt formula (7) determine x _op(gen):

$x_{\mathrm{op}}\left(\mathrm{gen}\right)=\frac{{\mathrm{\&xi;}}_1{x}_i^{\mathrm{best}}\left(\mathrm{gen}\right)+{\mathrm{\&xi;}}_2{x}_g^{\mathrm{best}}\left(\mathrm{gen}\right)}{{\mathrm{\&xi;}}_1+{\mathrm{\&xi;}}_2}---\left(7\right)$

X _bestrepresent the mean value of all particle desired positions in population, and adopt formula (8) to determine x _best:

$x_{\mathrm{best}}=\frac{1}{N_p}\underset{i=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{x}_i^{\mathrm{best}}\left(\mathrm{gen}\right)---\left(8\right)$

Wherein α, β are x _i(gen+1) the adjustment factor of position vector, μ is for selecting the factor, ξ ₁represent

the weight of position vector, ξ ₂represent the weight of position vector, β, μ, ξ ₁and ξ ₂random generation and α ∈ [0.6,0.65] between interval [0,1].

Step 3.5, repeating step 3.3～step 3.4N _ginferior, the optimal solution of output frequency and power resource allocation, N _grepresent the iterations of quantum particle swarm method.

Embodiment 1

In conjunction with Fig. 1～4, the present invention is based on the cognitive OFDM network self-adapting resource allocation methods that frequency spectrum is filled, step is:

Step 1, the noise power in frequency spectrum cavity-pocket and idle sub-carrier is detected in cognitive base station.

In authorisation network, there is N=16 authorized user, authorisation network system bandwidth W=8MHz, available subcarrier is counted K=128, the noise N of the upper cognitive user of subcarrier k ₀=1 × 10 ^-8w; The probability that frequency band is transferred to frequency spectrum cavity-pocket from authorized user seizure condition is 0.8, and it is 0.8 that frequency band is transferred to the probability of authorizing with needing use from cognitive user seizure condition; The radio spectrum resources in authorisation network is detected in cognitive base station in the cognitive OFDM network of filling based on frequency spectrum, obtain the spectrum utilization state information in current authorisation network through the message processing module analysis in cognitive base station, formation can characterize the frequency spectrum detection report of actual parameter and the reflection frequency spectrum cavity-pocket of frequency spectrum cavity-pocket, frequency spectrum detection report is sent to the cognitive user in the cognitive OFDM network of filling based on frequency spectrum by the mode of broadcast by cognitive base station, and the noise power in idle sub-carrier is detected according to the examining report of frequency spectrum cavity-pocket in cognitive base station;

Fig. 4 is that the cognitive OFDM network subcarrier of filling based on frequency spectrum takies situation, cognitive base station usable spectrum resource K _t={ 8 × (kt-1)+1～8 × kt, kt=1,3,5,7,9,11,13,15}, f _c=990MHz, Δ f=62.5KHz, F ₁=0.5MHz, F ₂=1MHz, F ₃=1.5MHz, F ₄=2MHz; f _crepresent the cognitive OFDM network initial frequency of filling based on frequency spectrum, F ₁, F ₂, F ₃and F ₄the relative initial frequency that represents frequency band 1, frequency band 2, frequency band 3 and frequency band 4, Δ f represents each subcarrier spacing.

Step 2, cognitive base station obtains each cognitive user channel condition information and real-time service transmission rate requirement.

The communication radius of cognitive base station is R=2Km, has M=8 cognitive user in cognition network, and cognitive base station is constant to the channel condition information of cognitive user communication link in an OFDM symbol time; Cognitive user is estimated the channel condition information of the last OFDM symbol of cognitive OFDM network down link of filling based on frequency spectrum by the message processing module of cognitive terminal, cognitive user is given cognitive base station by the cognitive OFDM network uplink link based on frequency spectrum filling by the information feedback of estimating, cognitive user feeds back to cognitive base station by the cognitive OFDM network uplink link of filling based on frequency spectrum by real-time service transmission rate requirement, and Fig. 2 is that the required channel condition information of cognitive OFDM Resource Allocation in Networks of filling based on frequency spectrum of the present invention is estimated schematic diagram.

Step 3, cognitive base station adopts quantum particle swarm method to distribute frequency and power resource in the cognitive OFDM network of filling based on frequency spectrum.

Each cognitive user sends the needed minimum transmission rate demand information of real time business to cognitive base station

fig. 3 is that the resource allocation methods flow process based on quantum particle swarm is carried out in cognitive base station:

First, the parameter of cognitive base station initialization quantum particle swarm method, initialization N _p=560, N _g=1100,

with

adopt formula (1) to determine

initialization

with

initialization gen=1;

Then, each particle position vector in initialization population, this position vector comprises frequency and power resource, each particle position vector x=[c, p] in initialization population, vectorial c represents that subcarrier distributes indicator variable, vector c meets formula (2) and formula (3), vector p represents power division indicator variable, and vectorial p meets formula (4), K _trepresent the idle sub-carrier set that moment t can use,

the real-time service transmission rate requirement that represents cognitive user m, M=3 represents the sum of cognitive user;

Secondly, determine the particle of fitness minimum in the adaptive value of each particle in population and population, adopt formula (5) to determine the adaptive value f (x of each particle in population _i(t)); If

order $x_i^{\mathrm{best}}\left(\mathrm{gen}\right)=x_i\left(\mathrm{gen}\right);$ If $f\left(x_i\left(\mathrm{gen}\right)\right)<f\left(x_g^{\mathrm{best}}\left(\mathrm{gen}\right)\right),$ Order $x_g^{\mathrm{best}}\left(\mathrm{gen}\right)=x_i\left(\mathrm{gen}\right);$

Again, redefine the position of each particle in population according to the particle position vector of each particle position vector sum fitness minimum, adopt the latest position x of each particle in formula (6) Population Regeneration _i(gen+1), adopt formula (7) to determine x _op(gen), adopt formula (8) to determine x _best;

Finally, make gen ← gen+1, repeat above-mentioned steps N _ginferior, output

as optimal solution.

In sum, the present invention is directed to real time business demand in the cognitive OFDM network of filling based on frequency spectrum, to minimize the cognitive OFDM network transmitting power of filling based on frequency spectrum as optimization aim, dynamic resource allocation method has been proposed, meet cognitive user minimum transmission rate demand; The method can fully be excavated available frequency spectrum cavity-pocket resource in the cognitive OFDM network of filling based on frequency spectrum, from frequency and two dimension dynamic on-demand Resources allocation of power, for promoting the cognitive OFDM network green communications of filling based on frequency spectrum that technical support is provided.

Claims (8)

1. a cognitive OFDM network self-adapting resource allocation methods of filling based on frequency spectrum, is characterized in that, comprises the following steps:

Step 1, the noise power in frequency spectrum cavity-pocket and idle sub-carrier is detected in cognitive base station;

Step 2, cognitive base station obtains each cognitive user channel condition information and real-time service transmission rate requirement;

Step 3, cognitive base station adopts quantum particle swarm method to distribute frequency and power resource in the cognitive OFDM network of filling based on frequency spectrum.

2. cognitive OFDM network self-adapting resource allocation methods of filling based on frequency spectrum according to claim 1, it is characterized in that, the noise power in frequency spectrum cavity-pocket and idle sub-carrier is detected in cognitive base station described in step 1, be specially: the radio spectrum resources in authorisation network is detected in the cognitive base station of filling in cognitive OFDM network based on frequency spectrum, obtain the spectrum utilization state information in current authorisation network through the message processing module analysis in cognitive base station, formation can characterize the frequency spectrum detection report of actual parameter and the reflection frequency spectrum cavity-pocket of frequency spectrum cavity-pocket, cognitive base station is sent to frequency spectrum detection report based on frequency spectrum and is filled the cognitive user in cognitive OFDM network by the mode of broadcast, the noise power in idle sub-carrier is detected according to the examining report of frequency spectrum cavity-pocket in cognitive base station.

3. cognitive OFDM network self-adapting resource allocation methods of filling based on frequency spectrum according to claim 1, it is characterized in that, cognitive base station described in step 2 obtains each cognitive user channel condition information and real-time service transmission rate requirement, be specially: cognitive user estimates to fill based on frequency spectrum the channel condition information of the last OFDM symbol of cognitive OFDM network down link by the message processing module of cognitive terminal, cognitive user by fill based on frequency spectrum cognitive OFDM network uplink link by estimate information feedback give cognitive base station, cognitive user feeds back to cognitive base station by fill cognitive OFDM network uplink link based on frequency spectrum by real-time service transmission rate requirement.

4. cognitive OFDM network self-adapting resource allocation methods of filling based on frequency spectrum according to claim 1, it is characterized in that, cognitive base station described in step 3 adopts quantum particle swarm method to distribute frequency and power resource in the cognitive OFDM network of filling based on frequency spectrum, comprises the following steps:

Step 3.1, the parameter of cognitive base station initialization quantum particle swarm method;

Step 3.2, the each particle position vector in initialization population, this position vector comprises frequency and power resource;

Step 3.3, determines the particle of fitness minimum in the adaptive value of each particle in population and population;

Step 3.4, redefines the position of each particle in population according to the particle position vector of each particle position vector sum fitness minimum;

Step 3.5, repeating step 3.3～step 3.4N _ginferior, the optimal solution of output frequency and power resource allocation, N _grepresent the iterations of quantum particle swarm method.

5. cognitive OFDM network self-adapting resource allocation methods of filling based on frequency spectrum according to claim 4, is characterized in that, the parameter of the cognitive base station initialization quantum particle swarm method described in step 3.1, is specially:

(1) number of particles N in initialization quantum particle swarm method population _pand N _pthe iterations N of ∈ [500,600], quantum particle swarm method _gand N _g∈ [1000,1200], the moment t cognitive user m subcarrier on subcarrier k distributes indicator variable

and

the power division indicator variable of moment t cognitive user m on subcarrier k

and

$p_{m,k}^t\&Element;\left[\mathrm{0,10}\right];$

(2) adopt formula (1) to determine moment t cognitive user m transmission rate on subcarrier k

$b_{m,k}^t=\frac{W}{K}{\log }_2(1+\frac{p_{m,k}^t{h}_{m,k}^t}{N_0})---\left(1\right)$

In formula (1), W represents the cognitive OFDM network bandwidth of filling based on frequency spectrum, and K represents the cognitive OFDM network number of sub carrier wave of filling based on frequency spectrum, N ₀represent the noise power on subcarrier,

represent that moment t cognitive user m gains at subcarrier k upper signal channel; T represents time index, and m represents cognitive user index, and k represents sub-carrier indices;

(3) the desired positions vector of i particle when the gen time iteration of initialization

desired positions vector with all particles in the gen time iteration population

initialization iterations sequence number gen=1.

6. cognitive OFDM network self-adapting resource allocation methods of filling based on frequency spectrum according to claim 4, is characterized in that, the each particle position vector in the initialization population described in step 3.2, is specially:

Each particle position vector x=[c, p] in initialization population, wherein vectorial c represents that subcarrier distributes indicator variable, vectorial c meets formula (2) and formula (3):

$c_{m,k}^t=\left\{\begin{array}{c}\left[\mathrm{0,1}\right],k\&Element;K_t\\ \left\{0\right\},\mathrm{else}\end{array}\right.,\&ForAll;m,k,t---\left(2\right)$

$\underset{m=1}{\overset{M}{\mathrm{\&Sigma;}}}{c}_{m,k}\&le;1,c_{m,k}\&GreaterEqual;0,\&ForAll;k,t---\left(3\right)$

Vector p represents power division indicator variable, and vectorial p meets formula (4):

$\underset{k=1}{\overset{K}{\mathrm{\&Sigma;}}}{c}_{m,k}^t{b}_{m,k}^t\&GreaterEqual;R_m^{\min }---\left(4\right)$

Wherein, K _trepresent the idle sub-carrier set that moment t can use, the real-time service transmission rate requirement that represents cognitive user m, M represents the sum of cognitive user.

7. cognitive OFDM network self-adapting resource allocation methods of filling based on frequency spectrum according to claim 4, is characterized in that, the particle of fitness minimum in the adaptive value of each particle and population in the definite population described in step 3.3, is specially:

Adopt formula (5) to determine the adaptive value f (x of each particle in population _i(t));

$f\left(x_i\left(\mathrm{gen}\right)\right)=\underset{m=1}{\overset{M}{\mathrm{\&Sigma;}}}\underset{k=1}{\overset{K}{\mathrm{\&Sigma;}}}{c}_{m,k}^t{p}_{m,k}^t---\left(5\right)$

If $f\left(x_i\left(\mathrm{gen}\right)\right)<f\left(x_i^{\mathrm{best}}\left(\mathrm{gen}\right)\right),$ Order $x_i^{\mathrm{best}}\left(\mathrm{gen}\right)=x_i\left(\mathrm{gen}\right);$

If $f\left(x_i\left(\mathrm{gen}\right)\right)<f\left(x_g^{\mathrm{best}}\left(\mathrm{gen}\right)\right),$ Order $x_g^{\mathrm{best}}\left(\mathrm{gen}\right)=x_i\left(\mathrm{gen}\right);$

Wherein, f (x _i(gen)) represent the adaptive value of i particle the gen time iteration,

represent the adaptive value of i particle desired positions vector in the time of the gen time iteration,

the adaptive value of desired positions vector particle in population while being illustrated in the gen time iteration.

8. cognitive OFDM network self-adapting resource allocation methods of filling based on frequency spectrum according to claim 4, it is characterized in that, the position that redefines each particle in population described in step 3.4 according to the particle position vector of each particle position vector sum fitness minimum, is specially:

The latest position that adopts each particle in formula (6) Population Regeneration is the position x of i particle in the time of the gen+1 time iteration _i(gen+1):

$x_i(\mathrm{gen}+1)=\left\{\begin{array}{cc}{x}_{\mathrm{op}}\left(\mathrm{gen}\right)+\mathrm{\&alpha;}|x_{\mathrm{best}}-x_i\left(\mathrm{gen}\right)|1n(1/\mathrm{\&beta;}),\mathrm{if}& \mathrm{\&mu;}\&GreaterEqual;0.5\\ x_{\mathrm{op}}\left(\mathrm{gen}\right)-\mathrm{\&alpha;}|x_{\mathrm{best}}-x_i\left(\mathrm{gen}\right)|1n(1/\mathrm{\&beta;}),\mathrm{if}& \mathrm{\&mu;}<0.5\end{array}---\left(6\right)\right.$

Wherein, x _op(gen) represent

with

between random site, and adopt formula (7) determine x _op(gen):

$x_{\mathrm{op}}\left(\mathrm{gen}\right)=\frac{{\mathrm{\&xi;}}_1{x}_i^{\mathrm{best}}\left(\mathrm{gen}\right)+{\mathrm{\&xi;}}_2{x}_g^{\mathrm{best}}\left(\mathrm{gen}\right)}{{\mathrm{\&xi;}}_1+{\mathrm{\&xi;}}_2}---\left(7\right)$

X _bestrepresent the mean value of all particle desired positions in population, and adopt formula (8) to determine x _best:

$x_{\mathrm{best}}=\frac{1}{N_p}\underset{i=1}{\overset{N_p}{\mathrm{\&Sigma;}}}{x}_i^{\mathrm{best}}\left(\mathrm{gen}\right)---\left(8\right)$

Wherein α, β are x _i(gen+1) the adjustment factor of position vector, μ is for selecting the factor, ξ ₁represent

the weight of position vector, ξ ₂represent

the weight of position vector, β, μ, ξ ₁and ξ ₂random generation and α ∈ [0.6,0.65] between interval [0,1].

Images