CN106453174A - Cognitive radio network resource allocation method based on signal modulation identification - Google Patents

Abstract

The invention discloses a cognitive radio network resource allocation method based on signal modulation identification, wherein the method mainly settles defects of overlow accuracy and oversmall use range in inter-class and in-class classification method of a cognitive radio network main user signal modulation mode in prior art and overcomes a defect of insufficient cognitive node requirement consideration caused by a fact that existing spectrum allocation technology focuses on consideration for requirement of a cognitive node service quality in existing spectrum allocation technology. The cognitive radio network resource allocation method comprises the main steps of 1, detecting the spectrum; 2, setting a detecting confidence level threshold value; 3, determining whether a main user channel state is idle; 4, acquiring modulation information; 5, reporting information; 6, performing spectrum fusion; and 7, allocating the spectrum resource. According to the cognitive radio network resource allocation method, corresponding parameters of a trusted system is dynamically adjusted according to a communication environment requirement, particularly an information parameter is modulated, thereby realizing higher flexibility and higher efficiency in allocation and use of a cognitive radio network spectrum resource.

Description

Cognition wireless network resource allocation methods based on signal modulate

Technical field

The invention belongs to communication technical field, further relate to one of cognition wireless network technical field and be based on letter The cognition wireless network resource allocation methods of number Modulation Identification.The present invention can be according to communication environment demand, and dynamic adjustment is trusted System relevant parameter, especially modulation parameter are so that the distribution of cognition wireless network frequency spectrum resource and use are more flexible, high Effect.

Background technology

On solving increasingly rare spectrum issue, cognitive radio (Cognitive Radio, abbreviation CR) be one kind very Promising technology.Digital modulation mode automatic identification is the intermediate steps that signal receives with demodulation, cognitive wireless network system In act as dominant role, in non-cooperation and dynamic communication environment, to preventing harmful interference authorized user, improve the profit of frequency spectrum With playing an important role in rate.In cognitive wireless network system, dynamic spectrum resource management is unusual part and parcel, wherein relates to And to modulation system Classification and Identification technology be one of basis of cognition wireless network resource allocation.Modulation Mode Recognition is inaccurate The waste to available resources will be led to, and signal modulation mode is carried out with accurate Classification and Identification and can effectively improve cognitive nothing The allocative efficiency of resource in line network system.

The patent document " sorting technique of digital modulation signals in a kind of cognition network " that University of Electronic Science and Technology applies at it Digital modulation in a kind of cognition network is disclosed in (application number 201210331200.6 application publication number CN 102868654 A) The sorting technique of signal.The method passes through following steps：1st, db5 wavelet transformation and fractional Fourier are carried out to digital modulation signals Conversion obtains data distribution situation.2nd, using the distribution situation in previous step as characteristic of division, the modulation methods of data signal are determined Formula.The method exist weak point be：Although the method proposes the modulated signal sorting technique in a kind of cognition network, It is not abundant to the quantity consideration of signal modulation mode species in practical communication scene.

The paper " research of cognitive radio intermediate frequency spectrum sensing and optimizing and RRM " that Du Hong et al. delivers at it Elaborate in (Beijing University of Post ＆ Telecommunication's doctoral candidate's academic dissertation 2012) in cognition network that one kind considers different demands to recognize Know the resource allocation methods of user.The method passes through following steps：1st, cognitive user obtains the letter of primary user by frequency spectrum perception Road use state information.2nd, sensing results and the business demand of oneself are reported to the base station of cognitive system.3rd, according to primary user Court verdict and cognitive user Qos demand information, cognitive base station be cognitive user distribution frequency spectrum resource.The method exists Weak point is：Although the method proposes a kind of resource allocation side considering different demands cognitive user in cognition network Case, but only propose the Qos demand of user, do not take into full account the demand of the modulation system to channel to be accessed for the subscriber signal, money Source distribution method considers not perfect.

Content of the invention

The present invention is directed to the deficiencies in the prior art, and proposition is a kind of to be divided based on the cognition wireless network resource of signal modulate Method of completing the square, can be according to communication environment demand, dynamic adjustment belief system relevant parameter, and especially modulation parameter, so that recognize Know the distribution of wireless network frequency spectrum resource and use more flexible, efficiently.

Realize the object of the invention concrete steps include as follows：

(1) detect frequency spectrum：

(1a) each cognitive nodes in cognition wireless network, receive the signal that primary user sends respectively；

(1b) according to statistic formula, each cognitive nodes in cognition wireless network, when calculating cognition wireless network respectively Domain statistic；

(2) according to the following formula, each cognitive nodes in cognition wireless network, calculate energy judging threshold Γ respectively_TD：

Wherein, Γ_TDiRepresent the corresponding energy judging threshold of i-th cognitive nodes in cognition wireless network,Represent high This white noise variance, its value represents in cognition wireless network the corresponding sampling number of i-th cognitive nodes, P for 1, M_afiExpression is recognized Know i-th cognitive nodes false-alarm probability in wireless network,Represent and open radical sign operation；

(3) judge whether the detection confidence level Time-domain Statistics amount of present cognitive wireless network primary user is less than threshold value Γ_TD, If so, represent that primary user's channel bands are unoccupied, execution step (5)；Otherwise, represent that primary user's channel status is busy, Execution step (4)；

(4) obtain modulation system information：

(4a) cognition wireless network cognitive nodes pre-process to cognition wireless network primary user's signal；

(4b) adopt temporal characteristics value extracting method of classifying between class, cognition wireless network cognitive nodes calculate normalization respectively Center amplitude spectrum density maximum value parameter, frequency spectrum maximum value parameter, envelope fluctuation parameter, Fei Ruoduan removes linear instantaneous phase in center The big parameter of standard deviation criteria four；

(4c) cognition wireless network cognitive nodes carry out classification between class to primary user's signal modulation mode type；

(4d) classification temporal characteristics value and Higher Order Cumulants characteristic value combined extraction method, cognition wireless network in class are adopted Cognitive nodes calculate single-frequency components degree of highlighting parameters in series, zero center normalized spectral density maximum parameters in series, spectral line respectively With energy density parameter, amplitude envelops standard deviation criteria, the big parameter of Higher Order Cumulants parameters in series five；

(4e) cognition wireless network cognitive nodes carry out classification in class to primary user's signal modulation mode type；

(4f) synthesize cognition wireless network primary user's channel modulation mode information；

(5) reporting information：

In cognition wireless network, each cognitive nodes reports cognition wireless network to cognition wireless network via node respectively Primary user's channel condition information, cognition wireless network primary user's channel modulation mode information and self-demand；

(6) frequency spectrum merges：

Cognition wireless network via node merges the cognition wireless network primary user that cognition wireless network cognitive nodes report Channel condition information, cognition wireless network primary user's channel modulation mode information and cognition wireless network cognitive nodes itself need Ask；

(7) frequency spectrum resource distribution：

Using frequency spectrum resource allocation method, cognition wireless network primary user's frequency spectrum resource is distributed to suitable cognition wireless Network cognitive user.

The present invention compared with prior art has advantages below：

First, because the present invention adopts temporal characteristics value extracting method of classifying between class, refine sorting parameter between class, overcome In prior art between cognition wireless network primary user's signal modulation mode class the too low shortcoming of the sorting technique degree of accuracy so that In the present invention, between cognition wireless network primary user signal modulation mode class, sorting technique improves fineness.

Second, because the present invention adopts classification temporal characteristics value and Higher Order Cumulants characteristic value combined extraction method in class, Refine and expanded sorting parameter in class, overcome in prior art in cognition wireless network primary user's signal modulation mode class The degree of accuracy of sorting technique shortcoming that is too low and being suitable for scene excessively limitation is so that the present invention improves fineness and universality.

3rd, because the present invention adopts frequency spectrum resource allocation method, analyze cognition wireless network primary user's channel simultaneously Modulation system information and cognition wireless network cognitive nodes modulation condition demand, overcome existing frequency spectrum distributing technique and are more heavily weighted toward Consider that the cognitive nodes demand that the demand of cognitive nodes service quality leads to considers not enough shortcoming so that the present invention improves frequency The flexibility of spectrum resource distribution and validity.

Brief description

Fig. 1 is the application scenario diagram of the present invention；

Fig. 2 is the flow chart of the present invention.

Specific embodiment

Below in conjunction with the accompanying drawings invention is described further.

Referring to the drawings 1, the application scenarios of the present invention are cognition wireless network, arrange a master in this cognition wireless network User PU, N number of cognitive nodes CR₁To CR_N, a via node and a base station BS.N represents cognitive in this cognition wireless network The number of node.Wherein, N number of cognitive nodes signal that unidirectional reception primary user PU sends respectively, the unidirectional reception of via node is N number of The information that cognitive nodes send, base station receives via node information and directly controls N number of cognitive nodes.

Below in conjunction with the accompanying drawings 2, the concrete steps of the present invention are further described.

Step 1, detects frequency spectrum.

Each cognitive nodes in cognition wireless network, receive the signal that primary user sends respectively.

According to statistic formula, each cognitive nodes in cognition wireless network, calculate cognition wireless network time domain system respectively Metering.

Step 2, according to the following formula, each cognitive nodes in cognition wireless network, calculate energy judging threshold Γ respectively_TD：

Wherein, Γ_TDiRepresent the corresponding energy judging threshold of i-th cognitive nodes in cognition wireless network,Represent high This white noise variance, its value represents in cognition wireless network the corresponding sampling number of i-th cognitive nodes, P for 1, M_afiExpression is recognized Know i-th cognitive nodes false-alarm probability in wireless network,Represent and open radical sign operation.

Step 3, judges whether the detection confidence level Time-domain Statistics amount of present cognitive wireless network primary user is less than threshold value Γ_TD, if so, represent that primary user's channel bands are unoccupied, execution step 5；Otherwise, represent that primary user's channel status is non-NULL Spare time, execution step 4.

Step 4, obtains modulation system information.

Cognition wireless network cognitive nodes pre-process to cognition wireless network primary user's signal.

The comprising the following steps that of described pretreatment：

The first step, according to the following formula, calculates cognition wireless network primary user's zero center signaling point：

Wherein, S_z(i_s,a_s) represent i-th in cognition wireless network_sThe corresponding a of individual cognitive nodes_sIndividual primary user's zero center Signaling point, S (i_s,j_s) represent i-th in cognition wireless network_sThe corresponding jth of individual cognitive nodes_sIndividual primary user's signal sampling point, j_s Represent any positive integer between 1 to M, M represents i-th_sThe number of the sampled point of the corresponding primary user's signal of individual cognitive nodes, a_s Represent any positive integer between 1 to M, ∑ represents even add operation.

Second step, according to the following formula, calculates cognition wireless network primary user Hilbert Hilbert conversion signaling point：

Wherein, HT (i_h,b_h) represent i-th in cognition wireless network_hThe corresponding b of individual cognitive nodes_hIndividual primary user's Martin Hilb Special Hilbert converts signaling point, b_hRepresent any positive integer between 1 to M, S_z(i_h,a_h) represent i-th in cognition wireless network_h The corresponding a of individual cognitive nodes_hIndividual primary user's zero center signaling point, a_hRepresent any positive integer between 1 to M, * represents convolution Operation, π represents pi.

3rd step, according to the following formula, calculates cognition wireless network primary user's envelope normalized signal point：

Wherein, S_e(i_e,c_e) represent i-th in cognition wireless network_eThe corresponding c of individual cognitive nodes_eIndividual primary user's envelope is returned One change signaling point, c_eRepresent any positive integer between 1 to M, S_z(i_e,a_e) represent i-th in cognition wireless network_eIndividual cognitive nodes Corresponding a_eIndividual primary user's zero center signaling point, a_eRepresent any positive integer between 1 to M, HT (i_e,b_e) represent cognitive nothing I-th in gauze network_eThe corresponding b of individual cognitive nodes_eIndividual primary user Hilbert Hilbert converts signaling point, b_eRepresent 1 arrive M it Between any positive integer,Represent and open radical sign operation；4th step, by cognition wireless network primary user's envelope normalized signal point Synthesis primary user's preprocessed signal.

4th step, cognition wireless network primary user's envelope normalized signal point is synthesized primary user's preprocessed signal.

Classification temporal characteristics value extracting method between using class, cognition wireless network cognitive nodes calculate normalization center respectively Amplitude spectrum density maxima parameter, frequency spectrum maximum value parameter, envelope fluctuation parameter, Fei Ruoduan goes at center linear instantaneous phase standard The difference big parameter of parameter four.

Between described class, classification temporal characteristics value extracting method comprises the following steps that：

The first step, according to the following formula, calculates the normalization center amplitude spectrum density maximum value parameter in cognition wireless network：

Wherein, γ_max(i_γ) represent i-th in cognition wireless network_γIndividual cognitive nodes corresponding normalization center amplitude spectrum is close The maximum value parameter of degree, MAX represents and takes maxima operation, i_γRepresent any positive integer between 1 to N, S_e(α_γ) represent cognitive nothing α in gauze network_γThe corresponding primary user's preprocessed signal of individual cognitive nodes, α_γRepresent any positive integer between 1 to M, M represents I-th_γThe number of the sampled point of the corresponding primary user's signal of individual cognitive nodes.

Second step, according to the following formula, calculates the frequency spectrum maximum value parameter in cognition wireless network：

Wherein, F_max(i_f) represent i-th in cognition wireless network_fIndividual cognitive nodes corresponding frequency spectrum maximum value parameter, MAX table Show and take maxima operation, i_fRepresent any positive integer between 1 to N, | | represent the operation that takes absolute value, DFT represents that Fourier becomes Change operation, S_e(α_f) represent α in cognition wireless network_fThe corresponding primary user's preprocessed signal of individual cognitive nodes, α_fRepresent that 1 arrives N Between any positive integer, μ (κ) represents that in cognition wireless network, the corresponding primary user's preprocessed signal of the κ cognitive nodes is equal Value, κ represents any positive integer between 1 to N.

3rd step, according to the following formula, calculates the envelope fluctuation parameter in cognition wireless network：

Wherein, m_A(i_m) represent i-th in cognition wireless network_mThe corresponding envelope fluctuation parameter of individual cognitive nodes, i_mRepresent 1 Any positive integer between N, σ²(ι) represent that in cognition wireless network, the ι cognitive nodes corresponding primary user pretreatment is believed Number variance, ι represents any positive integer between 1 to N, and μ (κ) represents in cognition wireless network the corresponding master of the κ cognitive nodes User's preprocessed signal average, κ represents any positive integer between 1 to N.

4th step, according to the following formula, linear instantaneous phase standard deviation door is removed at the Fei Ruoduan center calculating in cognition wireless network Limit value：

Wherein, τ_a(i_τ) represent i-th in cognition wireless network_τLinear instantaneous are removed at the corresponding Fei Ruoduan center of individual cognitive nodes Phase standard difference threshold value, i_τRepresent any positive integer between 1 to N, MAX represents and takes maxima operation, MIN represents and takes minimum Value Operations, Φ_NL(i_τ,k_τ) represent i-th in cognition wireless network_τThe corresponding kth of individual cognitive nodes_τIndividual primary user's signal transient phase Place value, k_τRepresent any positive integer between 1 to M.

5th step, according to the following formula, the Fei Ruoduan center in cognition wireless network that calculates goes linear instantaneous phase standard deviation to join Number：

Wherein, σ_dp(i_σ) represent i-th in cognition wireless network_σLinear instantaneous are removed at the corresponding Fei Ruoduan center of individual cognitive nodes Phase standard difference parameter, i_σRepresent any positive integer between 1 to N, Φ_NL(i_σ, k) represent i-th in cognition wireless network_σIndividual recognize Know the corresponding kth of node_σIndividual primary user's signal transient phase value, k_σRepresent any positive integer between 1 to M, C (η) represents cognitive The η cognitive nodes corresponding primary user signal transient phase place average in wireless network, η represents arbitrarily just whole between 1 to N Number.Cognition wireless network cognitive nodes are classified between primary user's class signal.

Classify between described class, comprise the following steps that：

The first step, according to the following formula, calculates threshold value of classifying between class：

Wherein, T represents between class classification threshold value, X and Y represents by two separate subsets of classification threshold T between class respectively, On the premise of P (X (T) | X) represents and is known to be signal subset X, using class between classification threshold T be judged to the empirical probability of X, P (Y (T) | Y) represent and be known to be signal subset Y on the premise of, using class between classification threshold T be judged to the empirical probability of Y, P (T) represents The mean value of probability P (X (T) | X) and probability P (Y (T) | Y).

Whether second step, judge current zero center normalization center amplitude spectrum density maximum value parameter more than its thresholding of classifying Value, if so, then judges primary user's signal modulation mode type as amplitude shift keying MASK and quadrature amplitude modulation MQAM two types One of, execute the 3rd step；Otherwise, it is determined that primary user's signal modulation mode type is phase-shift keying (PSK) MPSK, frequency shift keying One of MFSK and MSK MSK three types, execute the 4th step.

3rd step, judges that linear instantaneous phase standard deviation criteria is removed more than its thresholding of classifying in current Fei Ruoduan center whether Value, if so, then judges primary user's signal modulation mode type as amplitude shift keying MASK；Otherwise, it is determined that primary user signal modulation side Formula type is quadrature amplitude modulation MQAM.

4th step, judges that current envelope fluctuation parameters, whether more than its threshold value of classifying, if so, then judge primary user's signal Modulation system type is phase-shift keying (PSK) MPSK；Otherwise, it is determined that primary user's signal modulation mode type is frequency shift keying MFSK and One of little frequency shift keying MSK both types.

5th step, it is Cl that quadrature amplitude modulation MQAM signal is numbered_A, phase-shift keying (PSK) mpsk signal is Cl_B, by frequency displacement It is Cl that keying MFSK signal and MSK msk signal are numbered jointly_C, it is Cl that amplitude shift keying MASK signal is numbered_D.

Using classification temporal characteristics value in class and Higher Order Cumulants characteristic value combined extraction method, cognition wireless network is cognitive Node calculates single-frequency components degree of highlighting parameters in series, zero center normalized spectral density maximum parameters in series, spectral line and energy respectively Metric density parameter, amplitude envelops standard deviation criteria, the big parameter of Higher Order Cumulants parameters in series five.

In described class, classification temporal characteristics value is referred to Higher Order Cumulants characteristic value combined extraction method, classifies wink in class When the characteristic value and Higher Order Cumulants characteristic value method that carries out Conjoint Analysis, comprise the following steps that：

The first step, according to the following formula, calculates single-frequency components degree of highlighting parameters in series in cognition wireless network：

Wherein, C₂(i_c2) represent i-th in cognition wireless network_c2The corresponding second order single-frequency components of individual cognitive nodes highlight Degree parameter, S_F2Represent the second order normalized power spectral density parameters of cognition network primary user's signal, C₄(i_c) represent cognition wireless I-th in network_c4The corresponding quadravalence single-frequency components degree of highlighting of individual cognitive nodes, S_F4Represent the quadravalence of cognition network primary user's signal Normalized power spectral density, its length is M, S_F2And S (Pos)_F4(Pos) value represents a certain determination between 1 to M for 1, Pos Integer, represents multiplication operation, and ∑ represents even add operation, and Posk represents any positive integer between 1 to M.

Second step, according to the following formula, calculates zero center normalized spectral density maximum parameters in series in cognition wireless network：

Wherein, P_max2(i_p2) represent i-th in cognition wireless network_p2Individual cognitive nodes corresponding second order zero center normalization spectrum Density maxima parameter, P_max4(i_p4) represent i-th in cognition wireless network_p4The corresponding quadravalence zero center normalization of individual cognitive nodes Spectrum density maximum value parameter, i_p2And i_p4Represent any positive integer between 1 to N, MAX represents and take maxima operation, DFT represents Fourier transform operation, S_e2(α_p2) represent α in cognition wireless network_p2The corresponding primary user's preprocessed signal of individual cognitive nodes, S_e4(α_p4) represent α in cognition wireless network_p4The corresponding primary user's preprocessed signal of individual cognitive nodes, α_p2Represent between 1 to N Any positive integer, α_p4Represent any positive integer between 1 to N, M represents that cognition wireless network primary user's signal sampling is counted, ∑ represents Lian Jiacao, | | represent the operation that takes absolute value.

3rd step, according to the following formula, calculates spectral line and energy density parameter in cognition wireless network：

Wherein, Π_s(i_π) represent i-th in cognition wireless network_πThe corresponding spectral line of individual cognitive nodes and energy density parameter, i_πRepresent any positive integer between 1 to N, S_FRepresent the normalized power spectral density of cognition network primary user's signal, ω_bRepresent S_FIt is more thanRough estimate bandwidth, n_ωRepresent S in rough estimate bandwidth_FIt is more thanSpectrum number of lines, represent multiplication behaviour Make, ∑ represents even add operation.

4th step, according to the following formula, calculates amplitude envelops standard deviation criteria in cognition wireless network：

Wherein, E_s(i_e) represent i-th in cognition wireless network_eThe corresponding amplitude envelops standard deviation criteria of individual cognitive nodes, i_e Represent any positive integer between 1 to N, S (i_e,j_e) represent i-th in cognition wireless network_eThe corresponding jth of individual cognitive nodes_eIndividual Primary user's signal sampling point, j_eRepresent any positive integer between 1 to M, M represents i-th_eIndividual cognitive nodes corresponding primary user letter Number sampled point number,Represent and open radical sign operation.

5th step, according to the following formula, calculates Higher Order Cumulants parameters in series in cognition wireless network：

|C₂₁|(i_ε2)=cum [HT (i_ε2,b_ε2),HT(i_ε2,b_ε2)]=M₂₀

Wherein, | C₂₁|(i_ε2) represent i-th in cognition wireless network_ε2Individual cognitive nodes corresponding second order Higher Order Cumulants ginseng Number, i_ε2Represent any positive integer between 1 to N, | C₄₀|(i_ε4) represent i-th in cognition wireless network_ε4Individual cognitive nodes are corresponding Quadravalence Higher Order Cumulants parameter, i_ε4Represent any positive integer between 1 to N, cum [] expression asks Higher Order Cumulants to operate, HT (i_ε2,b_ε2) represent the corresponding b of i-th cognitive nodes in cognition wireless network_ε2Individual primary user Hilbert Hilbert conversion Signaling point, b_ε2Represent any positive integer between 1 to M, HT (i_ε4,b_ε4) represent i-th cognitive nodes pair in cognition wireless network The b answering_ε4Individual primary user Hilbert Hilbert converts signaling point, b_ε4Represent any positive integer between 1 to M, M₂₀Represent The second moment of primary user's signal, M in cognition wireless network₄₀Represent the Fourth-order moment of primary user's signal in cognition wireless network.

Cognition wireless network cognitive nodes are classified in primary user's class signal.

In described class, classification, comprises the following steps that：

The first step, according to the following formula, calculates threshold value of classifying between class：

Wherein, T represents classification threshold value in class, X and Y represents by two separate subsets of classification threshold T in class respectively, On the premise of P (X (T) | X) represents and is known to be signal subset X, it is judged to the empirical probability of X, P (Y using classification threshold T in class (T) | Y) represent and be known to be signal subset Y on the premise of, be judged to the empirical probability of Y using classification threshold T in class, P (T) represents The mean value of probability P (X (T) | X) and probability P (Y (T) | Y).

Second step, judges whether the quadravalence Higher Order Cumulants parameter in current Higher Order Cumulants parameters in series classifies more than it Threshold value, if so, then judges primary user's signal modulation mode type as 64 rank quadrature amplitude modulation 64QAM；Otherwise, judge currently Whether the second order Higher Order Cumulants parameter in Higher Order Cumulants parameters in series, less than its threshold value of classifying, if so, then judges primary Family signal modulation mode type is 8 rank quadrature amplitude modulation 8QAM；Otherwise, it is determined that primary user's signal modulation mode type is 16 ranks Quadrature amplitude modulation 16QAM.

3rd step, judges whether quadravalence single-frequency components degree of the highlighting parameter in current single-frequency components degree of highlighting parameters in series is big In its threshold value of classifying, if so, then judge primary user's signal modulation mode type as 8 rank phase-shift keying (PSK) 8PSK；Otherwise, judge to work as Whether second order single-frequency components degree of the highlighting parameter in front single-frequency components degree of highlighting parameters in series is less than its threshold value of classifying, if so, Then judge primary user's signal modulation mode type as 2 rank phase-shift keying (PSK) BPSK；Otherwise, it is determined that primary user's signal modulation mode type For 4 rank phase-shift keying (PSK) QPSK.

4th step, judges that current spectral line and energy density parameter, whether more than its threshold value of classifying, if so, then judge primary Family signal modulation mode type is MSK MSK；Otherwise, execute the 5th step.

5th step, judges the quadravalence zero center normalization spectrum in current zero center normalized spectral density maximum parameters in series Whether density maxima parameter, more than its threshold value of classifying, if so, then judges primary user's signal modulation mode type as 8 rank frequency displacements Keying 8FSK；Otherwise, judge the second order zero center normalization spectrum in current zero center normalized spectral density maximum parameters in series Whether density maxima parameter, less than its threshold value of classifying, if so, then judges primary user's signal modulation mode type as 2 rank frequency displacements Keying 2FSK；Otherwise, it is determined that primary user's signal modulation mode type is 4 rank frequency shift keying 2FSK.

6th step, it is Cl that 64 rank quadrature amplitude modulation 64QAM signals are numbered_A- 1, by 16 rank quadrature amplitude modulation 16QAM It is Cl that signal is numbered_A- 2, it is Cl that 8 rank quadrature amplitude modulation 8QAM signals are numbered_A-3；8 rank phase-shift keying (PSK) 8PSK signals are numbered For Cl_B- 1, it is Cl that 4 rank phase-shift keying (PSK) QPSK signals are numbered_B- 2, it is Cl that 2 rank phase-shift keying (PSK) bpsk signals are numbered_B-3；By 8 It is Cl that rank frequency shift keying 8FSK signal is numbered_C- 1, it is Cl that 4 rank frequency shift keying 4FSK signals are numbered_C- 2, by 2 rank frequency shift keyings It is Cl that 2FSK signal is numbered_C- 3, it is Cl that MSK msk signal is numbered_C-4；4 rank amplitude shift keying 4ASK signals are numbered For Cl_D- 1, it is Cl that 2 rank amplitude shift keying 2ASK signals are numbered_D-2.

Synthesis cognition wireless network primary user's channel modulation mode information.

Step 5, reporting information.

In cognition wireless network, each cognitive nodes reports cognition wireless network to cognition wireless network via node respectively Primary user's channel condition information, cognition wireless network primary user's channel modulation mode information and self-demand.

Described self-demand includes, and the QoS demand of cognition wireless network cognitive nodes and cognition wireless network cognition save Point modulation condition demand.

Step 6, frequency spectrum merges.

Cognition wireless network via node merges the cognition wireless network primary user that cognition wireless network cognitive nodes report Channel condition information, cognition wireless network primary user's channel modulation mode information and cognition wireless network cognitive nodes itself need Ask.

Step 7, frequency spectrum resource distributes.

Using frequency spectrum resource allocation method, cognition wireless network primary user's frequency spectrum resource is distributed to suitable cognition wireless Network cognitive user.

The comprising the following steps that of described frequency spectrum resource allocation method：

1st step, sets degree of correlation Pr between class_outerWith degree of correlation Pr in class_inner.

Degree of correlation Pr between described class_outerI () sets rule as follows：

The first step, when the modulation system of cognitive user i is MQAM, if it is Cl that primary user's modulation system is numbered_A, then set Pr_outerI () is 40%；If it is Cl that primary user's modulation system is numbered_B, then set Pr_outerI () is 30%；If primary user's modulation methods It is Cl that formula is numbered_C, then set Pr_outerI () is 20%；If it is Cl that primary user's modulation system is numbered_D, then set Pr_outerI () is 10%.

Second step, when the modulation system of cognitive user i is MPSK, if it is Cl that primary user's modulation system is numbered_A, then set Pr_outerI () is 30%；If it is Cl that primary user's modulation system is numbered_B, then set Pr_outerI () is 40%；If primary user's modulation methods It is Cl that formula is numbered_C, then set Pr_outerI () is 20%；If it is Cl that primary user's modulation system is numbered_D, then set Pr_outerI () is 10%.

3rd step, when the modulation system of cognitive user i is any one in MFSK and MSK, if primary user's modulation system Numbering is Cl_A, then set Pr_outerI () is 30%；If it is Cl that primary user's modulation system is numbered_B, then set Pr_outerI () is 20%；If it is Cl that primary user's modulation system is numbered_C, then set Pr_outerI () is 40%；If primary user's modulation system is numbered being Cl_D, then set Pr_outerI () is 10%.

4th step, when the modulation system of cognitive user i is MASK, if it is Cl that primary user's modulation system is numbered_A, then set Pr_outerI () is 30%；If it is Cl that primary user's modulation system is numbered_B, then set Pr_outerI () is 20%；If primary user's modulation methods It is Cl that formula is numbered_C, then set Pr_outerI () is 10%；If it is Cl that primary user's modulation system is numbered_D, then set Pr_outerI () is 40%.

Degree of correlation Pr in described class_innerI () sets rule as follows：

Numbering when primary user's channel modulation mode is Cl_A- 1, Cl_B- 1, Cl_C- 1, Cl_DWhen one of -1, set Pr_inner I () is 50%；Numbering when primary user's channel modulation mode is Cl_A- 2, Cl_B- 2, Cl_C- 2, Cl_DWhen one of -2, set Pr_innerI () is 30%；Numbering when primary user's channel modulation mode is Cl_A- 3, Cl_B- 3, Cl_C- 3, Cl_CWhen one of -4, if Determine Pr_innerI () is 20%.

2nd step, according to the following formula, calculates cognitive nodes access priority parameter in cognition network：

Pr=Pr_outer·Pr_inner

Wherein, Pr represents cognitive nodes access priority parameter, Pr in cognition network_outerRepresent the degree of correlation between class, Pr_innerRepresent the degree of correlation in class, represent multiplication operation.

3rd step, cognition wireless network via node is according to the size of access priority parameter, descending sequence.

4th step, the maximum cognition wireless network of cognition wireless network via node prioritizing selection access priority parameter is recognized Know that node accesses idle channel.

Claims (9)

1. a kind of cognition wireless network resource allocation methods based on signal modulate, comprise the steps：

(1) detect frequency spectrum：

(1a) each cognitive nodes in cognition wireless network, receive the signal that primary user sends respectively；

(1b) according to statistic formula, each cognitive nodes in cognition wireless network, calculate cognition wireless network time domain system respectively Metering；

(2) according to the following formula, each cognitive nodes in cognition wireless network, calculate energy judging threshold Γ respectively_TD：

${\mathrm{\Γ}}_{TDi}={\mathrm{\σ}}_{\mathrm{\ω}}^2(M+\sqrt{2M}{Q}^{-1}(P_{afi}\left)\right)$

Wherein, Γ_TDiRepresent the corresponding energy judging threshold of i-th cognitive nodes in cognition wireless network,Represent white Gaussian Noise variance, its value represents in cognition wireless network the corresponding sampling number of i-th cognitive nodes, P for 1, M_afiRepresent cognitive nothing I-th cognitive nodes false-alarm probability in gauze network,Represent and open radical sign operation；

(3) judge whether the detection confidence level Time-domain Statistics amount of present cognitive wireless network primary user is less than threshold value Γ_TDIf, It is to represent that primary user's channel bands are unoccupied, execution step (5)；Otherwise, represent that primary user's channel status is busy, hold Row step (4)；

(4) obtain modulation system information：

(4a) cognition wireless network cognitive nodes pre-process to cognition wireless network primary user's signal；

(4b) adopt temporal characteristics value extracting method of classifying between class, cognition wireless network cognitive nodes calculate normalization center respectively Amplitude spectrum density maxima parameter, frequency spectrum maximum value parameter, envelope fluctuation parameter, Fei Ruoduan goes at center linear instantaneous phase standard The difference big parameter of parameter four；

(4c) cognition wireless network cognitive nodes carry out classification between class to primary user's signal modulation mode type；

(4d) classification temporal characteristics value and Higher Order Cumulants characteristic value combined extraction method in class are adopted, cognition wireless network is cognitive Node calculates single-frequency components degree of highlighting parameters in series, zero center normalized spectral density maximum parameters in series, spectral line and energy respectively Metric density parameter, amplitude envelops standard deviation criteria, the big parameter of Higher Order Cumulants parameters in series five；

(4e) cognition wireless network cognitive nodes carry out classification in class to primary user's signal modulation mode type；

(4f) synthesize cognition wireless network primary user's channel modulation mode information；

(5) reporting information：

In cognition wireless network, each cognitive nodes reports cognition wireless network primary to cognition wireless network via node respectively Family channel condition information, cognition wireless network primary user's channel modulation mode information and self-demand；

(6) frequency spectrum merges：

Cognition wireless network via node merges cognition wireless network primary user's channel that cognition wireless network cognitive nodes report Status information, cognition wireless network primary user's channel modulation mode information and cognition wireless network cognitive nodes self-demand；

(7) frequency spectrum resource distribution：

Using frequency spectrum resource allocation method, cognition wireless network primary user's frequency spectrum resource is distributed to suitable cognition wireless network Cognitive user.

2. the cognition wireless network resource allocation methods based on signal modulate according to claim 1, its feature exists In the statistic formula described in step (1b) is as follows：

${\hat{V}}_{{\mathrm{TDv}}_i}=\frac{1}{M}{\∫}_0^T{y}^2\left(t\right)dt$

Wherein,Represent v in cognition wireless network_iIndividual cognitive nodes corresponding primary user signal detection confidence cognition wireless Network session statistic, ∫ represents integration operation, and M represents v in cognition wireless network_iThe corresponding sampling number of individual cognitive nodes, v_iRepresent any positive integer between 1 to N.

3. the cognition wireless network resource allocation methods based on signal modulate according to claim 1, its feature exists In comprising the following steps that of, the pretreatment described in step (4a)：

The first step, according to the following formula, calculates cognition wireless network primary user's zero center signaling point：

$S_z(i_s,a_s)=S(i_s,j_s)-\frac{1}{M}\underset{j_s=1}{\overset{M}{\Σ}}S(i_s,j_s)$

Wherein, S_z(i_s,a_s) represent i-th in cognition wireless network_sThe corresponding a of individual cognitive nodes_sIndividual primary user's zero center signal Point, S (i_s,j_s) represent i-th in cognition wireless network_sThe corresponding jth of individual cognitive nodes_sIndividual primary user's signal sampling point, j_sRepresent 1 Any positive integer between M, M represents i-th_sThe number of the sampled point of the corresponding primary user's signal of individual cognitive nodes, a_sRepresent 1 Any positive integer between M, ∑ represents even add operation；

Second step, according to the following formula, calculates cognition wireless network primary user Hilbert Hilbert conversion signaling point：

$HT(i_h,b_h)=S_z(i_h,a_h)*\frac{1}{{\mathrm{\πa}}_h}$

Wherein, HT (i_h,b_h) represent i-th in cognition wireless network_hThe corresponding b of individual cognitive nodes_hIndividual primary user's Hilbert Hilbert converts signaling point, b_hRepresent any positive integer between 1 to M, S_z(i_h,a_h) represent i-th in cognition wireless network_hIndividual The corresponding a of cognitive nodes_hIndividual primary user's zero center signaling point, a_hRepresent any positive integer between 1 to M, * represents that convolution is grasped Make, π represents pi；

3rd step, according to the following formula, calculates cognition wireless network primary user's envelope normalized signal point：

$S_e(i_e,c_e)=\frac{S_z(i_e,a_e)}{\sqrt{S_z^2(i_e,a_e)+{\mathrm{HT}}^2(i_e,b_e)}}$

Wherein, S_e(i_e,c_e) represent i-th in cognition wireless network_eThe corresponding c of individual cognitive nodes_eIndividual primary user's envelope normalization Signaling point, c_eRepresent any positive integer between 1 to M, S_z(i_e,a_e) represent i-th in cognition wireless network_eIndividual cognitive nodes correspond to A_eIndividual primary user's zero center signaling point, a_eRepresent any positive integer between 1 to M, HT (i_e,b_e) represent cognitive wireless I-th in network_eThe corresponding b of individual cognitive nodes_eIndividual primary user Hilbert Hilbert converts signaling point, b_eRepresent between 1 to M Arbitrarily positive integer,Represent and open radical sign operation；

4th step, cognition wireless network primary user's envelope normalized signal point is synthesized primary user's preprocessed signal.

4. the cognition wireless network resource allocation methods based on signal modulate according to claim 1, its feature exists In between step (4b) described class, classification temporal characteristics value extracting method comprises the following steps that：

The first step, according to the following formula, calculates the normalization center amplitude spectrum density maximum value parameter in cognition wireless network：

${\mathrm{\γ}}_{max}\left(i_{\mathrm{\γ}}\right)=\frac{MAX\[{S_e}^2\left({\mathrm{\α}}_{\mathrm{\γ}}\right)\]}{M}$

Wherein, γ_max(i_γ) represent i-th in cognition wireless network_γIndividual cognitive nodes corresponding normalization center amplitude spectrum density is Big value parameter, MAX represents and takes maxima operation, i_γRepresent any positive integer between 1 to N, S_e(α_γ) represent cognitive wireless α in network_γThe corresponding primary user's preprocessed signal of individual cognitive nodes, α_γRepresent any positive integer between 1 to M, M represents i-th_γ The number of the sampled point of the corresponding primary user's signal of individual cognitive nodes；

Second step, according to the following formula, calculates the frequency spectrum maximum value parameter in cognition wireless network：

$F_{max}\left(i_f\right)=\frac{MAX\left|DFT{\[S_e\left({\mathrm{\α}}_f\right)\]}^2\right|}{{\mathrm{\μ}}^2\left(\mathrm{\κ}\right)}$

Wherein, F_max(i_f) represent i-th in cognition wireless network_fIndividual cognitive nodes corresponding frequency spectrum maximum value parameter, MAX represents and takes Maxima operation, i_fRepresent any positive integer between 1 to N, | | represent the operation that takes absolute value, DFT represents that Fourier transformation is grasped Make, S_e(α_f) represent α in cognition wireless network_fThe corresponding primary user's preprocessed signal of individual cognitive nodes, α_fRepresent between 1 to N Any positive integer, μ (κ) represents the corresponding primary user's preprocessed signal average of the κ cognitive nodes, κ in cognition wireless network Represent any positive integer between 1 to N；

3rd step, according to the following formula, calculates the envelope fluctuation parameter in cognition wireless network：

$m_A\left(i_m\right)=\frac{{\mathrm{\σ}}^2\left(\mathrm{\ι}\right)}{{\mathrm{\μ}}^2\left(\mathrm{\κ}\right)}$

Wherein, m_A(i_m) represent i-th in cognition wireless network_mThe corresponding envelope fluctuation parameter of individual cognitive nodes, i_mRepresent 1 arrive N it Between any positive integer, σ²(ι) represent the ι cognitive nodes corresponding primary user preprocessed signal side in cognition wireless network Difference, ι represents any positive integer between 1 to N, and μ (κ) represents in cognition wireless network the corresponding primary user of the κ cognitive nodes Preprocessed signal average, κ represents any positive integer between 1 to N；

4th step, according to the following formula, linear instantaneous phase standard deviation threshold value is gone at the Fei Ruoduan center calculating in cognition wireless network：

${\mathrm{\τ}}_a\left(i_{\mathrm{\τ}}\right)=\frac{MAX\[{\mathrm{\Φ}}_{NL}(i_{\mathrm{\τ}},k_{\mathrm{\τ}})\]+MIN\[{\mathrm{\Φ}}_{NL}(i_{\mathrm{\τ}},k_{\mathrm{\τ}})\]}{2}$

Wherein, τ_a(i_τ) represent i-th in cognition wireless network_τLinear instantaneous phase is removed at the corresponding Fei Ruoduan center of individual cognitive nodes Standard deviation threshold value, i_τRepresent any positive integer between 1 to N, MAX represents and takes maxima operation, MIN represents and takes minimum of a value to grasp Make, Φ_NL(i_τ,k_τ) represent i-th in cognition wireless network_τThe corresponding kth of individual cognitive nodes_τIndividual primary user's signal transient phase value, k_τRepresent any positive integer between 1 to M；

5th step, according to the following formula, linear instantaneous phase standard deviation criteria is removed at the Fei Ruoduan center calculating in cognition wireless network：

${\mathrm{\σ}}_{dp}\left(i_{\mathrm{\σ}}\right)=\sqrt{\frac{\underset{{\mathrm{\Φ}}_{NL}(i_{\mathrm{\σ}},k_{\mathrm{\σ}})>{\mathrm{\τ}}_a\left(i_{\mathrm{\σ}}\right)}{\Σ}{{\mathrm{\Φ}}_{NL}}^2(i_{\mathrm{\σ}},k_{\mathrm{\σ}})}{C\left(\mathrm{\μ}\right)}-{\[\frac{\underset{{\mathrm{\Φ}}_{NL}(i_{\mathrm{\σ}},k_{\mathrm{\σ}})>{\mathrm{\τ}}_a\left(i_{\mathrm{\σ}}\right)}{\Σ}{\mathrm{\Φ}}_{NL}(i_{\mathrm{\σ}},k_{\mathrm{\σ}})}{C\left(\mathrm{\μ}\right)}\]}^2}$

Wherein, σ_dp(i_σ) represent i-th in cognition wireless network_σLinear instantaneous phase is removed at the corresponding Fei Ruoduan center of individual cognitive nodes Standard deviation criteria, i_σRepresent any positive integer between 1 to N, Φ_NL(i_σ, k) represent i-th in cognition wireless network_σIndividual cognitive section The corresponding kth of point_σIndividual primary user's signal transient phase value, k_σRepresent any positive integer between 1 to M, C (η) represents cognition wireless The η cognitive nodes corresponding primary user signal transient phase place average in network, η represents any positive integer between 1 to N.

5. the cognition wireless network resource allocation methods based on signal modulate according to claim 1, its feature exists In that classifies between class described in step (4c) comprises the following steps that：

The first step, according to the following formula, calculates threshold value of classifying between class：

$P\left(T\right)=\frac{P\left(X\right(T)\left|X\right)+P\left(Y\right(T)\left|Y\right)}{2}$

Wherein, T represents between class classification threshold value, X and Y represents respectively by two separate subsets of classification threshold T between class, P (X (T) | X) represent and be known to be signal subset X on the premise of, using class between classification threshold T be judged to the empirical probability of X, P (Y (T) | Y) on the premise of representing and being known to be signal subset Y, using class between classification threshold T be judged to the empirical probability of Y, P (T) represents probability The mean value of P (X (T) | X) and probability P (Y (T) | Y)；

Whether second step, judge current zero center normalization center amplitude spectrum density maximum value parameter more than its threshold value of classifying, If so, then judge primary user's signal modulation mode type as in amplitude shift keying MASK and quadrature amplitude modulation MQAM two types One kind, executes the 3rd step；Otherwise, it is determined that primary user's signal modulation mode type be phase-shift keying (PSK) MPSK, frequency shift keying MFSK and One of MSK MSK three types, execute the 4th step；

3rd step, judges that linear instantaneous phase standard deviation criteria is removed more than its threshold value of classifying in current Fei Ruoduan center whether, if It is then to judge primary user's signal modulation mode type as amplitude shift keying MASK；Otherwise, it is determined that primary user's signal modulation mode type For quadrature amplitude modulation MQAM；

4th step, judges that current envelope fluctuation parameters, whether more than its threshold value of classifying, if so, then judge primary user's signal modulation Mode type is phase-shift keying (PSK) MPSK；Otherwise, it is determined that primary user's signal modulation mode type is frequency shift keying MFSK and minimum frequency Move one of keying MSK both types；

5th step, it is Cl that quadrature amplitude modulation MQAM signal is numbered_A, phase-shift keying (PSK) mpsk signal is Cl_B, by frequency shift keying It is Cl that MFSK signal and MSK msk signal are numbered jointly_C, it is Cl that amplitude shift keying MASK signal is numbered_D.

6. the cognition wireless network resource allocation methods based on signal modulate according to claim 1, its feature exists In in the described class of step (4d), classification temporal characteristics value is referred to Higher Order Cumulants characteristic value combined extraction method, in class point The method that class temporal characteristics value and Higher Order Cumulants characteristic value carry out Conjoint Analysis, comprises the following steps that：

The first step, according to the following formula, calculates single-frequency components degree of highlighting parameters in series in cognition wireless network：

$C_2\left(i_{c2}\right)=\frac{30\·S_{F2}\left(Pos\right)}{\underset{Posk=Pos-15/2}{\overset{Pos+15/2}{\Σ}}{S}_{F2}\left(Posk\right)-S_{F2}\left(Pos\right)}$

$C_4\left(i_{c4}\right)=\frac{30\·S_{F4}\left(Pos\right)}{\underset{Posk=Pos-15/2}{\overset{Pos+15/2}{\Σ}}{S}_{F4}\left(Posk\right)-S_{F4}\left(Pos\right)}$

Wherein, C₂(i_c2) represent i-th in cognition wireless network_c2Individual cognitive nodes corresponding second order single-frequency components degree of highlighting ginseng Number, S_F2Represent the second order normalized power spectral density parameters of cognition network primary user's signal, C₄(i_c) represent cognition wireless network In i-th_c4The corresponding quadravalence single-frequency components degree of highlighting of individual cognitive nodes, S_F4Represent the quadravalence normalizing of cognition network primary user's signal Change power spectral density, its length is M, S_F2And S (Pos)_F4(Pos) value represents that for 1, Pos a certain determination between 1 to M is whole Number, represents multiplication operation, and ∑ represents even add operation, and Posk represents any positive integer between 1 to M；

Second step, according to the following formula, calculates zero center normalized spectral density maximum parameters in series in cognition wireless network：

$P_{max2}\left(i_{p2}\right)=MAX\[\frac{\left|DFT\left(S_{e2}\right({\mathrm{\α}}_{p2}\left)\right)\right|}{\underset{i=1}{\overset{M}{\Σ}}\left|DFT\left(S_{e2}\right({\mathrm{\α}}_{p2}\left)\right)\right|}\]$

$P_{max4}\left(i_{p4}\right)=MAX\[\frac{\left|DFT\left(S_{e4}\right({\mathrm{\α}}_{p4}\left)\right)\right|}{\underset{i=1}{\overset{M}{\Σ}}\left|DFT\left(S_{e4}\right({\mathrm{\α}}_{p4}\left)\right)\right|}\]$

Wherein, P_max2(i_p2) represent i-th in cognition wireless network_p2The corresponding second order zero center normalized spectral density of individual cognitive nodes Maximum value parameter, P_max4(i_p4) represent i-th in cognition wireless network_p4Individual cognitive nodes corresponding quadravalence zero center normalization spectrum is close The maximum value parameter of degree, i_p2And i_p4Represent any positive integer between 1 to N, MAX represents and take maxima operation, DFT represents in Fu Leaf transformation operates, S_e2(α_p2) represent α in cognition wireless network_p2The corresponding primary user's preprocessed signal of individual cognitive nodes, S_e4 (α_p4) represent α in cognition wireless network_p4The corresponding primary user's preprocessed signal of individual cognitive nodes, α_p2Represent between 1 to N Arbitrarily positive integer, α_p4Represent any positive integer between 1 to N, M represents that cognition wireless network primary user's signal sampling is counted, ∑ Represent Lian Jiacao, | | represent the operation that takes absolute value；

3rd step, according to the following formula, calculates spectral line and energy density parameter in cognition wireless network：

${\Π}_s\left(i_{\mathrm{\π}}\right)={\{\frac{n_{\mathrm{\ω}}}{{\mathrm{\ω}}_b}\·\frac{\ΣS_F}{{\mathrm{\ω}}_b}\}}_{S_F>\frac{MAX\[S_F\]}{2}}$

Wherein, Π_s(i_π) represent i-th in cognition wireless network_πThe corresponding spectral line of individual cognitive nodes and energy density parameter, i_πRepresent Any positive integer between 1 to N, S_FRepresent the normalized power spectral density of cognition network primary user's signal, ω_bRepresent S_FIt is more thanRough estimate bandwidth, n_ωRepresent S in rough estimate bandwidth_FIt is more thanSpectrum number of lines, represent multiplication operation, ∑ Represent even add operation；

4th step, according to the following formula, calculates amplitude envelops standard deviation criteria in cognition wireless network：

$E_s\left(i_e\right)=\sqrt{\frac{1}{M-1}\underset{i_e=1}{\overset{M}{\Σ}}{\left(S\right(i_e,i_e)-\frac{1}{M}\underset{i_e=1}{\overset{M}{\Σ}}S(i_e,j_e\left)\right)}^2}$

Wherein, E_s(i_e) represent i-th in cognition wireless network_eThe corresponding amplitude envelops standard deviation criteria of individual cognitive nodes, i_eRepresent 1 Any positive integer between N, S (i_e,j_e) represent i-th in cognition wireless network_eThe corresponding jth of individual cognitive nodes_eIndividual primary user Signal sampling point, j_eRepresent any positive integer between 1 to M, M represents i-th_eThe adopting of the individual corresponding primary user's signal of cognitive nodes The number of sampling point,Represent and open radical sign operation；

5th step, according to the following formula, calculates Higher Order Cumulants parameters in series in cognition wireless network：

|C₂₁|(i_ε2)=cum [HT (i_ε2,b_ε2),HT(i_ε2,b_ε2)]=M₂₀

$\left|C_{40}\right|\left(i_{\mathrm{\ϵ}4}\right)=cum\[HT(i_{\mathrm{\ϵ}4},b_{\mathrm{\ϵ}4}),HT(i_{\mathrm{\ϵ}4},b_{\mathrm{\ϵ}4}),HT(i_{\mathrm{\ϵ}4},b_{\mathrm{\ϵ}4}),HT(i_{\mathrm{\ϵ}4},b_{\mathrm{\ϵ}4})\]=M_{40}-3M_{20}^2$

Wherein, | C₂₁|(i_ε2) represent i-th in cognition wireless network_ε2The corresponding second order Higher Order Cumulants parameter of individual cognitive nodes, i_ε2 Represent any positive integer between 1 to N, | C₄₀|(i_ε4) represent i-th in cognition wireless network_ε4The corresponding quadravalence of individual cognitive nodes Higher Order Cumulants parameter, i_ε4Represent any positive integer between 1 to N, cum [] expression asks Higher Order Cumulants to operate, HT (i_ε2, b_ε2) represent the corresponding b of i-th cognitive nodes in cognition wireless network_ε2Individual primary user Hilbert Hilbert converts signal Point, b_ε2Represent any positive integer between 1 to M, HT (i_ε4,b_ε4) represent that in cognition wireless network, i-th cognitive nodes is corresponding B_ε4Individual primary user Hilbert Hilbert converts signaling point, b_ε4Represent any positive integer between 1 to M, M₂₀Represent cognitive The second moment of primary user's signal, M in wireless network₄₀Represent the Fourth-order moment of primary user's signal in cognition wireless network.

7. the cognition wireless network resource allocation methods based on signal modulate according to claim 1, its feature exists In in class described in step (4e), classification comprises the following steps that：

The first step, according to the following formula, calculates classification threshold value in class：

$P\left(T\right)=\frac{P\left(X\right(T)\left|X\right)+P\left(Y\right(T)\left|Y\right)}{2}$

Wherein, T represents classification threshold value in class, X and Y represents respectively by two separate subsets of classification threshold T in class, P (X (T) | X) represent and be known to be signal subset X on the premise of, be judged to the empirical probability of X using classification threshold T in class, P (Y (T) | Y on the premise of) expression is known to be signal subset Y, it is judged to the empirical probability of Y using classification threshold T in class, P (T) represents probability The mean value of P (X (T) | X) and probability P (Y (T) | Y)；

Whether second step, judge the quadravalence Higher Order Cumulants parameter in current Higher Order Cumulants parameters in series more than its thresholding of classifying Value, if so, then judges primary user's signal modulation mode type as 64 rank quadrature amplitude modulation 64QAM；Otherwise, judge current high-order Second order Higher Order Cumulants parameter in cumulant parameters in series, whether less than its threshold value of classifying, if so, then judges primary user's letter Number modulation system type is 8 rank quadrature amplitude modulation 8QAM；Otherwise, it is determined that primary user's signal modulation mode type is that 16 ranks are orthogonal Amplitude modulation(PAM) 16QAM；

3rd step, judges whether quadravalence single-frequency components degree of the highlighting parameter in current single-frequency components degree of highlighting parameters in series is more than it Classification threshold value, if so, then judges primary user's signal modulation mode type as 8 rank phase-shift keying (PSK) 8PSK；Otherwise, judge currently single Whether second order single-frequency components degree of the highlighting parameter in frequency component degree of highlighting parameters in series, less than its threshold value of classifying, is if so, then sentenced Determining primary user's signal modulation mode type is 2 rank phase-shift keying (PSK) BPSK；Otherwise, it is determined that primary user's signal modulation mode type is 4 Rank phase-shift keying (PSK) QPSK；

4th step, judges that current spectral line and energy density parameter, whether more than its threshold value of classifying, if so, then judge primary user's letter Number modulation system type is MSK MSK；Otherwise, execute the 5th step；

5th step, judges the quadravalence zero center normalized spectral density in current zero center normalized spectral density maximum parameters in series Whether maximum value parameter, more than its threshold value of classifying, if so, then judges primary user's signal modulation mode type as 8 rank frequency shift keyings 8FSK；

Otherwise, judge second order zero center normalized spectral density in current zero center normalized spectral density maximum parameters in series Whether big value parameter, less than its threshold value of classifying, if so, then judges primary user's signal modulation mode type as 2 rank frequency shift keyings 2FSK；Otherwise, it is determined that primary user's signal modulation mode type is 4 rank frequency shift keying 2FSK；

6th step, it is Cl that 64 rank quadrature amplitude modulation 64QAM signals are numbered_A- 1, by 16 rank quadrature amplitude modulation 16QAM signals Numbering is Cl_A- 2, it is Cl that 8 rank quadrature amplitude modulation 8QAM signals are numbered_A-3；8 rank phase-shift keying (PSK) 8PSK signals are numbered and is Cl_B- 1, it is Cl that 4 rank phase-shift keying (PSK) QPSK signals are numbered_B- 2, it is Cl that 2 rank phase-shift keying (PSK) bpsk signals are numbered_B-3；By 8 ranks It is Cl that frequency shift keying 8FSK signal is numbered_C- 1, it is Cl that 4 rank frequency shift keying 4FSK signals are numbered_C- 2, by 2 rank frequency shift keying 2FSK It is Cl that signal is numbered_C- 3, it is Cl that MSK msk signal is numbered_C-4；4 rank amplitude shift keying 4ASK signals are numbered and is Cl_D- 1, it is Cl that 2 rank amplitude shift keying 2ASK signals are numbered_D-2.

8. the cognition wireless network resource allocation methods based on signal modulate according to claim 1, its feature exists In the self-demand described in step (5) includes, and the QoS demand of cognition wireless network cognitive nodes and cognition wireless network are recognized Know node modulation condition demand.

9. the cognition wireless network resource allocation methods based on signal modulate according to claim 1, its feature exists In described in step (7), frequency spectrum resource allocation method comprises the following steps that；

1st step, sets degree of correlation Pr between class_outerWith degree of correlation Pr in class_inner；

2nd step, according to the following formula, calculates cognitive nodes access priority parameter in cognition network：

Pr=Pr_outer·Pr_inner

Wherein, Pr represents cognitive nodes access priority parameter, Pr in cognition network_outerRepresent the degree of correlation between class, Pr_innerRepresent The degree of correlation in class, represents multiplication operation；

3rd step, cognition wireless network via node is according to the size of access priority parameter, descending sequence；

4th step, the cognitive section of the maximum cognition wireless network of cognition wireless network via node prioritizing selection access priority parameter Point accesses idle channel.