CN103220052A - Method for detecting frequency spectrum hole in cognitive radio - Google Patents

Abstract

The invention provides a solving method which coordinates a cognitive radio system to enhance robustness detection. The method for detecting a frequency spectrum hole in cognitive radio mainly comprises the following steps: independently sampling a frequency spectrum environment which needs to be detected by a plurality of cognitive users of a cognitive system in different geographical positions so as to obtain perceptive signals; sending perceptive data obtained by means of sampling by the cognitive user through a control channel in the cognitive system to a cognitive based station; receiving a plurality of perceptive signals and splitting the perceptive signals into a plurality of shorter perceptive signal sections with identical slot time by the cognitive based station, and obtaining a perceptive signal matrix through data infusion; and calculating the perceptive signal matrix and a covariance matrix, maximum and minimum eigenvalues of the covariance matrix and eigenratio of the maximum eigenvalues and minimum eigenvalues, comparing the eigenratio with a decision threshold, and judging whether the frequency spectrum hole exists or not. The method for detecting the frequency spectrum hole in the cognitive radio can effectively improve systematic detecting performance under the condition that the number of cooperative users are less.

Description

Detect the method for frequency spectrum cavity-pocket in a kind of cognitive radio

Technical field

The present invention relates to a kind of method that in cognitive radio, detects frequency spectrum cavity-pocket, relate in particular to and adopt a plurality of cognitive user cooperation perception to detect the method for frequency spectrum cavity-pocket.

Background technology

Wireless communication spectrum is a kind of very valuable natural resources, is generally used by government authorization.The wireless frequency spectrum distribution method is that a certain section frequency spectrum licensed to a user regularly.Along with the development of wireless traffic, radio spectrum resources is deficient gradually, and the drawback of fixed spectrum allocation may constantly manifests.In actual applications, what frequency needs was at full stretch mainly concentrates in the radio band that frequency band is hundreds of MHz-3GHz, not user's use has some just occupied once in a while in addition in some frequency band mosts of the time, and the use of some other frequency band is then relatively very frequent.How to improve the availability of frequency spectrum and become the technical problem that people pay special attention to.Therefore proposed at cognitive radio technology, authorized frequency range idly and authorized user is not produced the method for disturbing by opportunistic access and solve the not high problem of frequency spectrum utilization rate.Count situation according to the cognitive user that participates in perception, can be divided into: single user's perception and the multi-user perception of cooperating at the perception of authorized master user signal.

Single user's frequency spectrum perception scheme is that individual equipment is realized the perception judgement to main subscriber signal, and representative has: matched filter detection, energy measuring, the detection of cyclostationary characteristic value etc.

The performance that the operation principle that matched filter detects is meant filter and the characteristic of signal obtain that certain is consistent, and the signal transient power of filter output and the ratio of noise average power are maximized.In cognitive radios, use matched filter, the same with detection signal in the existing communication network, in fact be exactly coherent demodulation, cognitive radio users need be known the prior information of authorized master user, as comprises modulation system, impulse waveform and packet format or the like.This method shortcoming is the situation that the meeting appearance causes the matched filter detector not adjudicate because not knowing main subscriber signal structure in the practical operation.

Energy measuring is to have judged whether that by the received signal energy value of measuring in a period of time or the one section frequency domain authorized user exists, and is a kind of irrelevant detection method.Energy detection technique directly is that to received signal energy value carries out a square processing, does not need to know main user's prior information, and is easy to realize that this is advantageous in cognitive radio technology.Shortcoming is that energy detection technique is subjected to the influence of cognitive user radio frequency environment, radio-frequency front-end performance and mobile environment easily.

In authorisation network, because signal is subjected to the modulation of artificial periodic signal, to the intrinsic periodicity of discrete sampling, scanning, modulation, multiple access, coding and the physical phenomenon of signal the statistical property of signal is presented periodically, so modulation signal have typical cyclo-stationary: all to present cyclic periodicity as its statistical property, average, auto-correlation function.The cyclostationary characteristic detection method is exactly to determine by utilizing the spectrum correlation function to detect the cycle period feature that exists in the received signal whether authorization user signal exists.The cyclostationarity detection method has overcome many shortcomings of matched filter method and energy measuring method, can also be applied to the detection of spread-spectrum signal, the influence that this method need not to know the prior information of signal and can break away from background noise, but the complexity height that it calculates, the observation time of requirement is longer.

Multi-user's perception of cooperating, in cognitive environment, may there be decline or shadow fading even can run into bigger barrier and can cause the main subscriber signal decay that the cognitive user termination receives very big in various degree in channel between main user and cognitive user, thereby cause the false retrieval erroneous judgement, and main telex network is caused interference.And multi-user's cooperative detection can effectively overcome the recessive terminal problem that single user detects, and improves and detects performance.

Summary of the invention

Technical problem: the purpose of this invention is to provide a kind of solution that detects robustness at cooperative cognitive radio system enhanced system, solve the problem that main subscriber signal verification and measurement ratio descends because building stops, cognitive user depleted of energy and cognitive user generation physical damage cause collaboration user to reduce, this method is a kind of tactic method, and the system that the method that the application of the invention proposes can effectively improve under the less situation of collaboration user detects performance.

Technical scheme: method of the present invention is a kind of method of tactic, reaches the purpose that increases the signal correction amount of information by the splitting and reorganizing to the perceptual signal of original cognitive user.

According to the perception scene schematic diagram of shown in Figure 1 cooperative cognitive of the present invention system, suppose to have in the cognitive unit a plurality of cognitive user SU on the diverse geographic location and a main user base station PBS and a cognition of being distributed in from base station SBS, method of the present invention is:

1. the independent cognitive user SU in the cognitive system samples to the spectrum environment that needs detect, and obtains perceptual signal.Perceptual signal is main user SU transmitter signal, and cognition by the analysis to perceptual signal, draws court verdict from base station SBS at last.

2. cognitive user SU sends to the perceptual signal data that sampling obtains cognitive from base station SBS by the control channel in the cognitive system.Control channel is the preserved signal bandwidth in the cognitive system, is generally used for transmitting perceptual signal and control signal.

3. cognition receives a plurality of perceptual signals from base station SBS, and cognition splits into q section short perceptual signal by signal resolver with each perceptual signal from base station SBS, and these signals are stored in respectively in the register, obtains a perceptual signal matrix Y then.The perception time slot of supposing cognitive user is T, and signal resolver is that the perceptual signal of T is divided into the signal segment that q section time slot is t (wherein T is the integral multiple of t, T=q*t) with duration.

4. calculate perceptual signal matrix Y covariance matrix

, and the eigenvalue of maximum of covariance matrix and minimal eigenvalue λ _MaxAnd λ _Min, minimax characteristic value ratio

5. feature ratio and decision threshold are compared, and by following decision rule,

Judge and have or not frequency spectrum cavity-pocket.

Description of drawings

Fig. 1 is the perception scene schematic diagram of cooperative cognitive of the present invention system.

Fig. 2 is a frequency spectrum sensing method flow chart of the present invention.

Fig. 3 is a fractionation cooperation signal frequency spectrum sensing method flow diagram of the present invention.

Embodiment

Below in conjunction with accompanying drawing and embodiment the present invention is described in further detail.

Fig. 1 is the perception scene schematic diagram of cooperative cognitive of the present invention system.What the present invention considered is to have cognitive cooperative sensing scene from the base station in the cooperative cognitive wireless network, has provided the model of cognition wireless network, and wherein each part effect is as follows:

Primary User (PU): represent main user, claim authorized user again, promptly can cognitive system want perception to determine take its mandate frequency spectrum.

Primary Base Station (PBS): represent main user base station, can communicate from base station SBS that main responsible main user PU signal of transmission and the main user PU of management are to authorizing the operating position of frequency spectrum with cognition.

Secondary User (SU): the expression cognitive user, the main user PU signal of being responsible for main user base station PBS is launched carries out perception, and sensing results is sent to cognitive from base station SBS.

Secondary Base Station (SBS): expression is cognitive from the base station, the main perceptual signal of being responsible for receiving each cognitive user SU, and from the SBS of base station, handle the perceptual signal data that receive in cognition, and make then and whether authorize the frequency range free time, promptly whether there is the judgement of frequency spectrum cavity-pocket.

Maximum Minimum Eigenvalue(MME): expression eigenvalue of maximum, minimal eigenvalue decision method, the perceptual signal data that participate in a plurality of cognitive user SU of cooperation perception are become signal matrix through data fusion, eigenvalue of maximum, minimal eigenvalue and minimax characteristic value ratio that the associate(d) matrix theory draws signal matrix utilize feature ratio and decision threshold to make comparisons at last and draw last court verdict.

Main user PU and main user base station PBS communicate, and signal that the mode that cognitive user SU adopts cooperation is launched main user base station PBS and main user PU launch signal and detect, and a signal data that detects are sent to cognitive from base station SBS.Cognition is carried out respective handling from base station SBS someway to these perceptual signal The data, at last according to decision rule, judge in the frequency range that we are concerned about whether have main user PU signal, promptly whether have frequency spectrum cavity-pocket, finally realize the collaborative spectrum sensing function.That consider in the diagram scene is main user base station PBS and cognitive from the separate situation of base station SBS, and cognition network is finished by cognitive user SU the perception of main user base station PBS.Cognitive user SU to main user PU launch result that signal detects exist two kinds may:

$x\left(i\right)=\left\{\begin{array}{cc}n\left(i\right),& H_0\\ h\left(i\right)s\left(i\right)+n\left(i\right),& H_1\end{array}\right.$

Wherein: the unknown signaling that x (i) representative sampling obtains;

It is σ that n (i) represents variance ²Noise;

H (i) is the channel gain coefficient;

S (i) is a signal;

I represents time value;

H ₀Represent and have only noise in the received signal, not main user PU launches the situation of signal;

H ₁Then represent in the received signal and not only contain noise but also contain the situation that main user PU launches signal.

Fig. 2 is a frequency spectrum sensing method flow chart of the present invention, promptly is the key step of MME method.In Fig. 1, M cognitive user SU arranged, be expressed as: SU1, SU2, SU3 ..., SUM, N received signal of each cognitive user SU sampling.Then j cognitive user SU can be expressed as X at detected signal of the k time sampling instant and noise _j(k), n _j(k) (j=1,2..., M; K=1,2..., N).For convenience of explanation, adopt following form to represent relevant parameter: $S=\left[\begin{array}{cccc}{S_1}^T& {S_2}^T& ...& S_M^T\end{array}\right]$ ， $n=\left[\begin{array}{cccc}{n}_1^T& n_2^T& ...& n_M^T\end{array}\right]$ Represent main user PU to launch the background noise vector matrix at signal vector matrix and cognitive user SU place respectively.

Hence one can see that M cognitive user SU, its received signal can be represented with a M * N dimensional signal vector matrix, and is as follows:

Because main user PU launches signal and noise signal is uncorrelated, can draw the covariance matrix of above-mentioned signal matrix X:

R _X＝E(XX ^T)＝E[(Hs)(Hs) ^T]+E(nn ^T)＝R _S+R _n＝R _S+σ ²I _M

Wherein: I _MBe M rank unit matrixs.

More than be theory analysis, and in the reality since the sampled signal number be limited, so also need to define the sampling covariance matrix of some reality:

${\hat{R}}_X\left(N\right)=\frac{1}{N}{\mathrm{XX}}^T$

${\hat{R}}_S\left(N\right)=\frac{1}{N}{\left(\mathrm{Hs}\right)\left(\mathrm{Hs}\right)}^T$

${\hat{R}}_n\left(N\right)=\frac{1}{N}{\mathrm{nn}}^T$

Because signal and noise all are steadily and ergodic random process, then when N → ∞, following relational expression establishment is arranged:

$R_X\≈{\hat{R}}_X\left(N\right)={\hat{R}}_S\left(N\right)+{\hat{R}}_n\left(N\right)$

Make λ _Max, λ _MinBe respectively matrix

Eigenvalue of maximum and minimal eigenvalue;

ρ _Max, ρ _MinBe respectively matrix

Eigenvalue of maximum and minimal eigenvalue.

Work as H ₁During situation, covariance matrix R _XThe minimax eigenvalue is arranged _Max, λ _MinBe respectively λ _Max=ρ _Max+ σ ², λ _Min=ρ _Min+ σ ², and λ _Max/ λ _Min1;

If H ₀Situation, then λ _Max=λ _Min=σ ², λ then _Maxλ _Min=1.

Can judge whether there is main user PU signal in the perceived spectral section by above-mentioned analysis.Carry out frequency spectrum perception according to this method, its perceptual performance is significantly improved than single user's frequency spectrum perception.

The above-mentioned MME method that provides mainly is to form a signal vector matrix by the perception data that merges a plurality of cognitive user SU, obtain the covariance matrix of received signal matrix according to signal and noise irrelevance, the decision value of obtaining under the prevailing condition according to big dimension random theory compares with decision threshold more then, draws court verdict at last.The method effect under the sufficient situation of fixedly cooperative cognitive user SU number and cooperative cognitive user SU is obvious, but owing to participate in the cognitive user of cooperation in the perception scene is constantly to change, so the cooperative cognitive user SU number that is changing can cause the robustness of frequency spectrum perception performance to reduce, be embodied in when the cooperative cognitive number of users reduces, the frequency spectrum detection probability of MME algorithm reduces, and this has reduced the practicality of algorithm.

This method is exactly at above-mentioned shortcoming, is fully understanding on the Random Matrices Theory application foundation, proposes to improve one's methods as shown in Figure 3.Main improvement step is before cognition is merged the received signal matrix from base station SBS, to the processing that splits of each perceptual signal, (wherein T is the integral multiple of t to be exactly with M perceptual signal each subsignal that all splits into q equal in length specifically, T=q*t, q determines during cognitive system in design), j perceptual signal X then _j(i)=h _j(i) S _j(i)+n _j(i) through splitting q the subsignal that obtains be:

$X_{j1}\left(i^{\′}\right)=h_{j1}\left(i^{\′}\right)S_{j1}\left(i^{\′}\right)+n_{j1}\left(i^{\′}\right),i^{\′}=1,...,\frac{N}{q}$

$X_{j2}\left(i^{\′}\right)=h_{j2}\left(i^{\′}\right)S_{j2}\left(i^{\′}\right)+n_{j2}\left(i^{\′}\right),i^{\′}=\frac{N}{q}+1,...,\frac{N}{q}*2$

$...,X_{\mathrm{jm}}\left(i^{\′}\right)=h_{\mathrm{jm}}\left(i^{\′}\right)S_{\mathrm{jm}}\left(i^{\′}\right)+n_{\mathrm{jm}}\left(i^{\′}\right),i^{\′}=\frac{N}{q}*(m-1)+1,...,\frac{N}{q}*m,...$

$X_{\mathrm{jq}}\left(i^{\′}\right)=h_{\mathrm{jq}}\left(i^{\′}\right)S_{\mathrm{jq}}\left(i^{\′}\right)+n_{\mathrm{jq}}\left(i^{\′}\right),i^{\′}=\frac{N}{q}*(q-1)+1,...,\frac{N}{q}*q,$

Through splitting M cooperation perceptual signal, fractionation is obtained the perceptual signal matrix of M*q sub-signal fused Cheng Xin, the benefit of doing like this is exactly the logic number that has increased subsignal, alleviate less cooperation perceptual signal to the frequency spectrum perception Effect on Performance, and then strengthened the robustness of perception algorithm.

Beneficial effect of the present invention is, the present invention program can influence the detection robustness of main user PU signal at the cooperative cognitive user SU quantity that constantly changes in the cooperative cognitive network, has proposed a kind of solution.Adopt the method for multi-user's collaborative spectrum sensing, can effectively avoid of the influence of factors such as the stopping of building in the wireless network, signal fadeout perceptual performance.Secondly, splitting the cooperation perceptual signal is that more subsignal is fused into new perceptual signal matrix, the benefit of doing like this is exactly the logic number that has increased subsignal, alleviated less cooperation perceptual signal to the frequency spectrum perception Effect on Performance, and then strengthened the robustness of perception algorithm, effectively improve the detection probability of cognitive system main user PU signal under the less situation of collaboration user, strengthen cognitive system perceptual performance robustness.

Claims (2)

1. detect the method for frequency spectrum cavity-pocket in the cognitive radio, it is characterized in that comprising following steps:

Step 1: a plurality of cognitive user (SU) that are in diverse geographic location are sampled to the spectrum environment that needs detect independently in the cognitive system, obtain perceptual signal;

Step 2: cognitive user (SU) sends to the perceptual signal data that sampling obtains cognitive from the base station (SBS) by the control channel in the cognitive system;

Step 3: cognitive (SBS) receives a plurality of perceptual signals from the base station, and cognitive (SBS) splits into a plurality of short identical perceptual signal sections of time slot with perceptual signal from the base station, obtains a perceptual signal matrix by the data fusion technology then;

Step 4: calculate perceptual signal matrix covariance matrix, and the eigenvalue of maximum of covariance matrix and minimal eigenvalue, and minimax characteristic value ratio;

Step 5: feature ratio and decision threshold are compared, and by given decision rule

Judge and have or not frequency spectrum cavity-pocket.

2. detect the method for frequency spectrum cavity-pocket in a kind of cognitive radio according to claim 1, it is characterized in that described data fusion technology: suppose to have in the cognitive unit M cognitive user SU to participate in the cooperation perception, work alone between them, and j SU user is expressed as X at i perceptual signal constantly _j(i) (j=1 ..., M), for easy analysis, the signal vector that is defined as follows:

X represents the perceptual signal matrix of SU, wherein X _j(j=1 ..., M) represent the signal vector that j SU sampling obtains.S, n represent PU transmitter signal matrix and noise signal matrix respectively.M * the N that is made up of M perception user's perceptual signal ties up the perceptual signal matrix.With all the signal vector X in the matrix _J(N) (j=1 ..., M) on average split into q(q 〉=1) the long subsignal vector of section L=N/q, then j perceptual signal X _j(i)=h _j(i) S _j(i)+n _j(i) through splitting q the subsignal that obtains be:

Parameter q is system's setting.So decomposition obtains K=M*q sub-signal vector, then all subsignals is formed new matrix form, obtains the matrix of a K * L like this:

。

Images