CN101588191B - Method and device for radio signal recognition - Google Patents

Abstract

The invention discloses a method and a device for radio signal recognition, and belongs to the field of communication. The method comprises the following steps: sampling a radio signal to be recognized to obtain a digital signal of the radio signal to be recognized; performing calculation of a self correlation function on the digital signal to obtain the self correlation function of the digital signal; calculating a decision quantity according to the self correlation function of the digital signal; and determining whether a basic user signal exists in the radio signal to be recognized according to the decision quantity. The device comprises a sampling module, a first computing module, a second computing module and a determining module. The decision quantity is obtained according to the self correlation function of the digital signal to be recognized, and the existence of the basic user information is determined according to the decision quantity, so the method overcomes the defect thatthe detection performance of an energy detection method on OFDM signals is reduced under the conditions that noise has high uncertainty and SNR is low.

Description

The radio signal recognition method and apparatus

Technical field

The present invention relates to the communications field, particularly a kind of radio signal recognition method and apparatus.

Background technology

The frequency spectrum resource of modern wireless communication systems is planned as a whole to distribute by radio regulatory organization, adopts at present the principle and scheme of static allocation.But along with wireless communication system is increasing, even each communication system all improves the availability of frequency spectrum of self, limited frequency spectrum resource also more and more can not satisfy growing user's request, one of the main reasons be the static allocation method of frequency spectrum resource from the double angle in time domain and spatial domain, still have many white spaces that are not fully utilized.In order to solve the frequency spectrum resource deficiency, realize the frequency spectrum dynamic management and to improve the availability of frequency spectrum, cognitive radio (CR, Cognitive Radio) technology has been proposed.

Network take the CR technology as fundamental construction is called as cognitive radio networks.Cognitive radio networks is comprised of basic network and cognition network, the subnet of the various isomeries that basic network is comprised of the elemental user that distributes frequency range (being that structure is various) all, and every kind of subnet has separately changeless network topology.The secondary device that cognition network is used the frequency range of elemental user by hope forms, each secondary device, and namely cognitive user must have cognitive ability.Cognitive user is compared elemental user and is had lower frequency spectrum access priority, and namely when elemental user did not use the frequency range of distribution, cognitive user can be utilized the communication between these idle frequency ranges realization cognitive user; But when elemental user need to utilize the frequency range of these distribution, no matter cognitive user has not at these frequency range transmitting datas, elemental user will directly be realized corresponding communication at these frequency range transmitting datas.When elemental user occurred, in order elemental user not to be caused harmful interference, cognitive user must be monitored out soon the appearance of elemental user signal and be abdicated corresponding frequency range so that elemental user uses; When elemental user did not occur, in order to improve spectrum utilization efficiency, cognitive user must detect the appearance that does not have the elemental user signal soon.So to the reliable Detection of elemental user signal, be to realize take the basic premise of CR technology as the cognition wireless network function on basis.

Ultra broadband (Ultra Wideband) technology combines with the CR technology, has produced the cognitive ultra-wideband technology, becomes one of developing direction of mobile communication.In the cognition wireless network that has adopted super-broadband tech, cognitive user is to the elemental user signal, also namely the detection of the narrow band signal in more existing communication systems become key link.Can fast and effeciently detect these elemental user signals, directly have influence on the overall performance of cognitive radio networks.Therefore, the authorization signal detection method becomes an important issue in the present cognitive super-broadband tech.

Prior art is for the Mb-ofdm (OFDM that needs in the cognitive ultra-wideband system in time to be detected, when Orthogonal Frequency Division Multiplexing) the elemental user signal of form detected, most widely used detection method was the energy measuring method.The energy measuring method passes through to measure the energy of waveform input signal in the setting-up time, compares according to the threshold value of testing and experience is set again with in advance, obtains testing result.As shown in Figure 1, be the system construction drawing of energy detection method in the prior art.X among Fig. 1 (t), y (t), w (t) all can be illustrated as having the Equivalent Base-Band signal form of real part and imaginary part.When having elemental user to occur, x (t) namely is the ofdm signal that elemental user sends, and when elemental user did not occur, x (t) was zero; W (t) is noise signal; Y (t) receives signal; Eth is in advance according to a threshold value of testing and experience is set.By receiving signal y (t) T computing time _TotalIn the input signal gross energy, this signal gross energy calculates with the equivalence of falling into a trap at last at frequency domain in time domain.Calculating energy adopts fast fourier transform (FFT, Fast Fourier Transform) algorithm more in frequency domain.Suppose time T _TotalIn receive N _TotalThe sampled value of individual input signal, sampling time T _s=T _Total/ N _Total, then be N _TotalThe FFT of point obtains frequency-region signal, calculates signal energy E in frequency domain, finally by crossing threshold judgement, obtains testing result.Decision rule is as follows: if E 〉=E _Th, then judge to have the elemental user signal; Otherwise, judge not have the elemental user signal.

In realizing process of the present invention, the inventor finds:

Energy detection method needs as a comparison standard of predetermined threshold, and this threshold value is the threshold value E of energy _Th, E _ThChange to noise level is very responsive, when noise level change Amplitude Ratio is larger, is difficult to rationally determine this threshold value.Because in practical communication system, the impacts such as the multipath fading that becomes when the signal that elemental user sends can be subject to, shadow fading, also to be superimposed with the various impacts that become noise when unforeseen, therefore the noise in the practical communication system always has certain uncertainty, when the uncertainty of noise is larger, the hydraulic performance decline of energy measuring method, even at all can't be for detection of the elemental user signal.

In addition, because wireless communications environment is abominable, the noise in the practical wireless communication systems often has very large uncertainty, the signal to noise ratio (SNR that cognitive user can receive, Signal Noise Ratio) also may be lower, the energy measuring method is difficult to detect the signal of low SNR.

Summary of the invention

For overcome when noise has very large uncertainty and situation that SNR is lower under, the energy measuring method is to the decline of the detection performance of ofdm signal, the embodiment of the invention provides a kind of radio signal recognition method and apparatus.Described technical scheme is as follows:

A kind of radio signal recognition method, described method comprises:

Treat cognitive radio signal and sample, obtain the described digital signal for the treatment of cognitive radio signal;

Described digital signal is carried out the calculating of auto-correlation function, obtain the auto-correlation function of described digital signal;

Auto-correlation function according to described digital signal calculates judgement amount;

Determine according to described judgement amount whether described treating exists the elemental user signal in the cognitive radio signal;

Wherein, the described calculating of described digital signal being carried out auto-correlation function, the auto-correlation function that obtains described digital signal specifically comprises:

Described digital signal is carried out the estimation of auto-correlation function, obtain the middle auto-correlation function of described digital signal, described middle auto-correlation function is a binary function;

Auto-correlation function in the middle of described is carried out variable simplify, obtain the auto-correlation function of described digital signal, described auto-correlation function is a function of a single variable.

A kind of radio signal recognition equipment, described equipment comprises:

Sampling module is used for treating cognitive radio signal and samples, and obtains the described digital signal for the treatment of cognitive radio signal;

The first computing module is used for described digital signal is carried out the calculating of auto-correlation function, obtains the auto-correlation function of described digital signal;

The second computing module is used for calculating judgement amount according to the auto-correlation function of described digital signal;

Determination module is used for determining according to described judgement amount whether the described radio signal of cognition for the treatment of exists the elemental user signal;

Wherein, described the first computing module comprises:

Estimation unit is used for described digital signal is carried out the estimation of auto-correlation function, obtains the middle auto-correlation function of described digital signal;

Computing unit is used for that auto-correlation function in the middle of described is carried out variable and simplifies, and obtains the auto-correlation function of described digital signal.

The beneficial effect of the technical scheme that the embodiment of the invention provides is:

In the embodiment of the invention, carry out the calculating of auto-correlation function by the digital signal for the treatment of cognitive radio signal, and then calculate judgement amount according to auto-correlation function; Determine to treat whether have the elemental user signal in the cognitive radio signal according to judgement amount, overcome when noise has very large uncertainty and situation that SNR is lower under, the energy measuring method is to the decline of the detection performance of ofdm signal.

Description of drawings

Fig. 1 is the system configuration schematic diagram of the energy detection method that provides of prior art;

Fig. 2 is the radio signal recognition method flow diagram that the embodiment of the invention provides;

Fig. 3 is the radio signal recognition method flow schematic diagram that the embodiment of the invention provides

Fig. 4 (a) is one of the radio signal recognition method that provides of the embodiment of the invention and energy measuring method correlation curve figure;

Fig. 4 (b) is two of the radio signal recognition method that provides of the embodiment of the invention and energy measuring method correlation curve figure;

Fig. 4 (c) is three of the radio signal recognition method that provides of the embodiment of the invention and energy measuring method correlation curve figure;

Fig. 4 (d) is four of the radio signal recognition method that provides of the embodiment of the invention and energy measuring method correlation curve figure;

Fig. 5 is the radio signal recognition equipment that the embodiment of the invention provides.

Embodiment

The radio signal recognition method and apparatus that the embodiment of the invention provides is that the elemental user signal to the OFDM form in the communication system detects, and all ofdm signals in time domain are all take its Equivalent Base-Band signal form as the example explanation.

In the embodiment of the invention, treat that cognitive radio signal is specially the cognitive user signal.The embodiment of the invention is at first sampled to the cognitive user signal, after obtaining the digital signal of cognitive user signal, this digital signal is carried out the calculating of auto-correlation function, calculate judgement amount by auto-correlation function, determine whether there is the elemental user signal in the cognitive user signal according to judgement amount again.

In the embodiment of the invention, the ofdm signal that elemental user is sent is designated as x (t), and the signal that cognitive user receives is designated as y (t), and noise signal is designated as w (t).For the ease of analyzing and describing, in the present embodiment, x (t), y (t) and w (t) all are expressed as the Equivalent Base-Band signal form.Wherein, the universal expression formula of x (t) is:

$x\left(t\right)=\underset{l=-\&infin;}{\overset{+\&infin;}{\mathrm{\&Sigma;}}}\underset{k=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{d}_l\left(k\right)q(t-\mathrm{lT}){\mathrm{<figure-callout\; id="2"\; label="e\; j"\; filenames="S2008101124805E00031.png,S2008101124805E00032.png"\; state="\{\{state\}\}">e</figure-callout>}}^j---\left(1\right)$

In the formula (1), t is time domain variable, and N is the sub-carrier number of ofdm signal, and each group ofdm signal has comprised a plurality of OFDM symbols, d ₁(k) data symbol of k subcarrier in l OFDM symbol of expression, value be 1 with-1 two kind of situation, in the present embodiment, they are independent identically distributed, and d ₁(k) probability of getting two values of binary phase shift keying (BPSK, Binary Phase Shift Keying) mapping equates, is 1/2; Q (t) is the pulse shaping function, and q (t) is only in 0≤t≤T value 1 in the time, other times value 0; T is the time that comprises total OFDM symbol of Cyclic Prefix; T _uIt is the time that does not comprise the OFDM symbol of Cyclic Prefix; T _gThe time of Cyclic Prefix, i.e. T=T _u+ T _g

Wherein, the BPSK technology is a kind of modulation technique that adopts in the ofdm signal.D under the different modulation techniques ₁(k) desirable value is also different, adopts BPSK modulation system d ₁(k) value is 1 or-1, and taking this a kind of modulation system herein is the example explanation, does not affect the embodiment of the invention and adopts other modulation systems.Just change corresponding d for other modulation systems ₁(k) value, algorithm is in full accord.

Need to prove, be without loss of generality, the channel that the present embodiment is investigated is rayleigh fading channel, and the baseband signal of transmission can be multiplied by factor-alpha e through behind this channel ^{I θ}, α e wherein ^{I θ}Be the modeling parameters of rayleigh fading channel, it is the rayleigh distributed of σ that α obeys parameter, and the value according to the decline feature σ of Rayleigh channel in the present embodiment is 0.5; θ distributes (0,2) upper obedience evenly.

The noise signal that superposes in the system is w (t), if noise is the zero-mean white Gaussian noise in each short time, and the average power of every section interior noise of time is at [P ₁, P _h] the upper obedience evenly distribute, the uncertainty of noise then is expressed as with U:

$U=\frac{P_h-P_l}{(P_h+P_l)/2}---\left(2\right)$

Wherein, P ₁The lower limit of expression noise power, P _hThe upper limit of expression noise power.

According to the content of the uncertainty of the feature of above-mentioned Rayleigh channel and noise, obtain the Equivalent Base-Band signal that cognitive user receives and be:

y(t)＝H _i*αe ^iθx(t)+w(t)，i＝0，1 (3)

Wherein, during i=1, H _i=1 expression has elemental user signal x (t) to occur; During i=0, H _i=0 expression does not have elemental user signal x (t) to occur.

As shown in Figure 2, the radio signal recognition method for the embodiment of the invention provides specifically comprises the steps:

Step 101: the signal y (t) that cognitive user is received samples, and obtains digital signal y[n];

Wherein, for convenience of description, at first with sampling time T _s=T _u/ N samples to the x (t) in the formula (1), obtains digital signal x[n]:

$x\left[n\right]=\underset{l=-\&infin;}{\overset{+\&infin;}{\mathrm{\&Sigma;}}}\underset{k=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{d}_l\left(k\right)q(n-\mathrm{lP}){\mathrm{<figure-callout\; id="2"\; label="e\; j"\; filenames="S2008101124805E00031.png,S2008101124805E00032.png"\; state="\{\{state\}\}">e</figure-callout>}}^j$

$=\underset{l=-\&infin;}{\overset{+\&infin;}{\mathrm{\&Sigma;}}}\underset{k=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{d}_l\left(k\right)q(n-\mathrm{lP})W_N^{-k(n-\mathrm{lP}-L)}---\left(4\right)$

Wherein, P=T/T _s, L=T _g/ T _s, N=T _u/ T _s,

N is the variable of discrete domain, the counting of expression sampling, n that the hardware resource that can provide with cognitive user is provided is relevant, the connotation of all the other parameters is ditto described.

With sampling time T _s=T _u/ N samples to the y (t) in the formula (3), and by formula (4) gained x[n] can get digital signal y[n]:

$y\left[n\right]=H_i*\mathrm{\&alpha;}{e}^{\mathrm{i\&theta;}}x\left[n\right]+w\left[n\right]$

$=w\left[n\right]+H_i*\mathrm{\&alpha;}{e}^{\mathrm{i\&theta;}}\underset{l=-\&infin;}{\overset{+\&infin;}{\mathrm{\&Sigma;}}}\underset{k=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{d}_l\left(k\right)q(n-\mathrm{lP})W_N^{-k(n-\mathrm{lP}-L)},i=\mathrm{0,1}---\left(5\right)$

Step 102: to digital signal y[n] carry out the estimation of auto-correlation function, obtain y[n] auto-correlation function R _y(n, τ);

Wherein, because y[n] expression formula in contain x[n], therefore at first to x[n] carry out the estimation of auto-correlation function, obtain x[n] auto-correlation function R _x(n, τ):

$R_x(n,\mathrm{\&tau;})=E\left\{x\right[n\left]x^{*}\right[n-\mathrm{\&tau;}\left]\right\}=\underset{l=-\&infin;}{\overset{+\&infin;}{\mathrm{\&Sigma;}}}\underset{k=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{\mathrm{\&sigma;}}_b^{2}q(n-\mathrm{lP})q^{*}(n-\mathrm{lP}-\mathrm{\&tau;})W_N^{-\mathrm{k\&tau;}}---\left(6\right)$

Wherein, τ is independent variable, gets nonnegative integer, the displacement of signal in discrete domain after the expression sampling, x ^*[n-τ] is x[n] through the displacement τ after x[n-τ] the conjugation expression formula, E represents mathematic expectaion, According to formula (5) and by

The above in like manner can obtain y[n] auto-correlation function R _y(n, τ):

1) when elemental user signal x[n is arranged], i.e. H _i=1 o'clock,

$R_y(n,\mathrm{\&tau;})=E\left\{y\right[n\left]y^{*}\right[n-\mathrm{\&tau;}\left]\right\}=E\left\{\left(\mathrm{\&alpha;}{e}^{\mathrm{i\&theta;}}x\right[n]+w[n\left]\right)\left(\mathrm{\&alpha;}{e}^{-\mathrm{i\&theta;}}{x}^{*}[n-\mathrm{\&tau;}]+w^{*}[n-\mathrm{\&tau;}]\right)\right\}$

$={\mathrm{\&alpha;}}^2{R}_x(n,\mathrm{\&tau;})+R_w\left(\mathrm{\&tau;}\right)$ $\left(7\right)$

$={\mathrm{\&alpha;}}^2\underset{l=-\&infin;}{\overset{+\&infin;}{\mathrm{\&Sigma;}}}\underset{k=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{\mathrm{\&sigma;}}_b^{2}q(n-\mathrm{lP})q^{*}(n-\mathrm{lP}-\mathrm{\&tau;})W_N^{-\mathrm{k\&tau;}}+R_w\left(\mathrm{\&tau;}\right)$

Wherein, R _w(τ) be the auto-correlation function of noise sequence.

2) when there not being elemental user signal x[n], i.e. H _i=0 o'clock,

R _y(n，τ)＝R _w(τ) (8)

Wherein, need to prove, auto-correlation function has reflected the similarity of same signal in different time sections.

Step 103: with y[n] auto-correlation function R _y(n, τ) carries out variable and simplifies, and deletion variable n obtains y[n] a new auto-correlation function R _y(τ);

Wherein, can draw R by formula (6) _x(n, τ) is that the cycle about n is the periodic function of P, according to the character of periodic function, for independent variable n since any one cycle P inner function and all equate, therefore the function in the one-period done and, and ask its average within this cycle, can obtain a new auto-correlation function R _x(τ) be:

Wherein,

N value among the φ (τ) is got nonnegative integer, so φ (τ) is the function about independent variable τ.Can draw R by formula (9) _x(τ) in τ=0 with τ=the N place reaches maximum, and τ is at other the R of value place _xBe 0 (τ).

In like manner, can obtain y[n by formula (7)] new auto-correlation function R _y(τ) be:

$R_y=\left(\mathrm{\&tau;}\right)=\frac{1}{P}\underset{n=0}{\overset{P-1}{\mathrm{\&Sigma;}}}{R}_y(n,\mathrm{\&tau;})=\frac{1}{P}\underset{n=0}{\overset{P-1}{\mathrm{\&Sigma;}}}\{{\mathrm{\&alpha;}}^2\underset{l=-\&infin;}{\overset{+\&infin;}{\mathrm{\&Sigma;}}}\underset{k=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{\mathrm{\&sigma;}}_b^{2}q(n-\mathrm{lP})q^{*}(n-\mathrm{lP}-\mathrm{\&tau;})W_N^{-\mathrm{k\&tau;}}+R_w\left(\mathrm{\&tau;}\right)\}$

Wherein, formula (10) is to be white Gaussian noise (in the actual application, noise all is near white Gaussian noise) at noise, has carried out infinite summation when calculating auto-correlation function and obtains in the situation not taking to be similar to.By formula (10) as can be known, R _yOnly about the function of a single variable of τ (τ).According to formula (7), (9) and (10) and above-mentioned analysis as can be known, when elemental user signal x[n is arranged], i.e. H _i=1 o'clock, R _yBe not 0 in τ=0 and τ=N place (τ), and at other τ values R of place _yBe 0 (τ).In the formula (10), according to the characteristic of white Gaussian noise, R _wBe not 0 at τ=0 place only (τ), and at other τ values R of place _w(τ) be 0.

Wherein, in actual application, noise is all near white Gaussian noise, so the noise of the present embodiment is chosen as white Gaussian noise.

Step 104: according to y[n] auto-correlation function R _y(τ) calculate judgement amount;

Carry out the calculating of judgement amount r according to formula (10), the result is:

$r=\left|\frac{\mathrm{real}\left\{R_y\left(N\right)\right\}}{\mathrm{real}\left\{\frac{1}{N-1}\underset{\mathrm{\&tau;}=1}{\overset{\mathrm{\&tau;}=N-1}{\mathrm{\&Sigma;}}}{R}_y\left(\mathrm{\&tau;}\right)\right\}}\right|---\left(11\right)$

Wherein, real represents the expression formula in the braces is thereafter got real part.

By above-mentioned formula (10) as can be known, R _yBe not 0 and at other τ values R of place at τ=0 and τ=N place (τ) _yBe 0 (τ).In the situation that noise is white Gaussian noise, the denominator part in the formula (11)

In actual conditions, when elemental user signal x[n is arranged] when occurring, real{R _y(N) } reach maximum, and denominator is approximately 0, the judgement amount r that obtains by formula (11) is positive infinity.

Step 105: determine whether to exist the elemental user signal according to the judgement amount that obtains.

Particularly, the false alarm probability (namely in fact not having signal to detect the probability of signal) that can bear according to the user of the present embodiment is set a threshold value C _ThJudgement amount r and threshold value C that above-mentioned through type (11) is calculated _ThCompare, as r 〉=C _ThThen judged result is for existing the elemental user signal; Otherwise then judged result is not for existing the elemental user signal.

Wherein, threshold value C _ThChange can have influence on false alarm probability and detection probability, C _ThCross lowly, detection probability can be improved, but false alarm probability equally also can improve; Otherwise C _ThToo high, detection probability and false alarm probability all can reduce.Because the user has certain requirement to false alarm probability and detection probability, so corresponding thresholding can be set.For detection method arbitrarily, the user can obtain false alarm probability and detection probability under the different threshold values by in advance simulation, so require just threshold value to be set according to corresponding false alarm probability and detection probability.

The technical scheme that the present embodiment provides can be represented by radio signal recognition method flow schematic diagram shown in Figure 3.Wherein, ADC represents the sampling to cognitive user signal y (t) in the time domain.

Radio signal recognition method and energy measuring method correlation curve figure that Fig. 4 (a)～Fig. 4 (d) provides for the embodiment of the invention.Wherein, during the estimation auto-correlation function, the maximum of n gets 2048, and namely counting of sampling is 2048 points.The ofdm signal that needs to detect is with micro-wave access to global intercommunication (WiMax, World Interoperability for Microwave Access) signal is the example explanation, do not affect the embodiment of the invention and be applicable to other ofdm signals, this signal is applied in the wireless MAN more.In the embodiment of the invention, this signal bandwidth is 28MHz, and number of sub carrier wave is 256, and the Cyclic Prefix number is 64, adopts the BPSK mapping, and what elemental user sent is the Equivalent Base-Band signal.Investigate false alarm probability and be respectively 0.3,0.4, noise uncertainty U is respectively the present embodiment provides under 2/3,2 the different scenes radio cognitive approach and the detection performance of two kinds of detection methods of energy measuring method, obtain detection performance comparison curve such as Fig. 4 (a) of two kinds of methods, Fig. 4 (b), Fig. 4 (c) is shown in Fig. 4 (d).Wherein, Fig. 4 (a) is that false alarm probability is 0.3, and the noise uncertainty is 2/3 o'clock, the curve that the detection probability of the radio cognitive approach that energy measuring method and the present embodiment provide changes with SNR; Fig. 4 (b) is that false alarm probability is 0.3, and the noise uncertainty is 2 o'clock, the curve that the detection probability of the radio cognitive approach that energy measuring method and the present embodiment provide changes with SNR; Fig. 4 (c) is that false alarm probability is 0.4, and the noise uncertainty is 2/3 o'clock, the curve that the detection probability of the radio cognitive approach that energy measuring method and the present embodiment provide changes with SNR; Fig. 4 (d) is that false alarm probability is 0.4, and the noise uncertainty is 2 o'clock, the curve that the detection probability of the radio cognitive approach that energy measuring method and the present embodiment provide changes with SNR.

By Fig. 4 (a), Fig. 4 (c) can find out, when the uncertainty of noise is smaller (U=2/3), the method performance that the present embodiment provides slightly is better than the energy measuring method, but both gaps are little.With Fig. 4 (a) and Fig. 4 (b), Fig. 4 (c) and Fig. 4 (d) contrast as can be known, when noise uncertainty U increases to 2 by 2/3, the detection performance of the method that the present embodiment provides does not have to change substantially, and the detection performance of energy measuring method sharply descends, and the method that this moment, the embodiment of the invention provided obviously is better than the performance of energy measuring method.

Because the energy measuring method has under the condition of higher SNR and detects preferably performance, so when SNR improved, the detection performance gap of two kinds of methods diminished; But in practical communication system, noise often has larger uncertainty, and the elemental user signal energy that cognitive user receives is also lower.When noise had larger noise uncertainty, the thresholding of energy method can not accurately arrange, and when SNR was low, the performance of energy method can be poor; The method that the embodiment of the invention provides, by formula (10) and formula (11) as can be known judgement amount and noise uncertainty it doesn't matter, so performance can be got well.Therefore, the method for taking under these circumstances the embodiment of the invention to provide detects ofdm signal, and detection probability is than adopting the energy measuring method to be significantly improved.

In the present embodiment, take the abundant y[n that counts with certain sampling time], estimate that afterwards its auto-correlation function line function of going forward side by side simplifies, and then calculate judgement amount, finally by crossing threshold judgement, obtain testing result.Whether the method that the present embodiment uses auto-correlation function ratio to detect is judged the existence of elemental user signal, removed the calculating of energy in the energy measuring method and the setting of energy threshold value from, solved because the energy that the uncertainty of noise is brought calculates inaccuracy, the energy threshold value can't rationally set and the problem of the energy measuring method hydraulic performance decline that caused by the uncertainty of noise; Solved simultaneously the problem that energy measuring method that the uncertainty by noise causes is difficult to detect low SNR signal.

As shown in Figure 5, be the radio signal recognition equipment that the embodiment of the invention provides, this equipment comprises:

Sampling module is used for treating cognitive radio signal and samples, and obtains treating the digital signal of cognitive radio signal;

The first computing module is used for digital signal is carried out the calculating of auto-correlation function, obtains the auto-correlation function of digital signal;

The second computing module is used for calculating judgement amount according to the auto-correlation function of digital signal;

Determination module is used for determining to treat according to judgement amount whether cognitive radio signal exists the elemental user signal.

Wherein, sampling module is treated cognitive radio signal with the default sampling time and is sampled, and this default sampling time equals the ratio of the sub-carrier number of time of the non-OFDM symbol that comprises Cyclic Prefix in the ofdm signal and ofdm signal.

Wherein, the first computing module specifically comprises:

Estimation unit is used for digital signal is carried out the estimation of auto-correlation function, obtains the middle auto-correlation function of digital signal;

Computing unit is used for that middle auto-correlation function is carried out variable and simplifies, and obtains the auto-correlation function of digital signal.

Further, this equipment also comprises:

The threshold value determination module is used for according to the preset rules of user to false alarm probability and detection probability, sets threshold value;

Correspondingly, determination module specifically is used for working as judgement amount more than or equal to described threshold value, determines to treat to have the elemental user signal in the cognitive radio signal.

In the present embodiment, take the abundant y[n that counts by sampling module with certain sampling time], estimate that by the first computing module its auto-correlation function line function of going forward side by side simplifies afterwards, calculate judgement amount by the second computing module, determine finally by crossing determination module, obtain testing result.Whether the method that the present embodiment uses auto-correlation function ratio to detect is judged the existence of elemental user signal, removed the calculating of energy in the energy measuring method and the setting of energy threshold value from, solved because the energy that the uncertainty of noise is brought calculates inaccuracy, the energy threshold value can't rationally set and the problem of the energy measuring method hydraulic performance decline that caused by the uncertainty of noise; Solved simultaneously the problem that energy measuring method that the uncertainty by noise causes is difficult to detect low SNR signal.

The embodiment of the invention can realize by software, and corresponding software can be stored in the storage medium that can read, for example in the hard disk of computer, CD or the floppy disk.

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

Claims (8)

1. a radio signal recognition method is characterized in that, comprising:

Treat cognitive radio signal and sample, obtain the described digital signal for the treatment of cognitive radio signal;

Described digital signal is carried out the calculating of auto-correlation function, obtain the auto-correlation function of described digital signal;

Auto-correlation function according to described digital signal calculates judgement amount;

Determine according to described judgement amount whether described treating exists the elemental user signal in the cognitive radio signal;

Wherein, the described calculating of described digital signal being carried out auto-correlation function, the auto-correlation function that obtains described digital signal specifically comprises:

Described digital signal is carried out the estimation of auto-correlation function, obtain the middle auto-correlation function of described digital signal, described middle auto-correlation function is a binary function;

Auto-correlation function in the middle of described is carried out variable simplify, obtain the auto-correlation function of described digital signal, described auto-correlation function is a function of a single variable.

2. radio signal recognition method according to claim 1 is characterized in that, the described radio signal for the treatment of cognition is sampled and is specially:

Treat cognitive radio signal with the default sampling time and sample,

The described default sampling time equals the ratio of the sub-carrier number of time of the non-OFDM symbol that comprises Cyclic Prefix in the ofdm signal and ofdm signal.

3. radio signal recognition method according to claim 1 is characterized in that, described middle auto-correlation function is the periodic function about a variable in the auto-correlation function in the middle of described.

4. radio signal recognition method according to claim 3 is characterized in that, describedly auto-correlation function in the middle of described is carried out variable simplifies and is specially:

Auto-correlation function in the middle of described is sued for peace in the one-period of described variable, summation gained function is averaging in described one-period, deletion has periodic variable.

5. radio signal recognition method according to claim 4 is characterized in that, described judgement amount is specially

Wherein r represents judgement amount, and real represents to get the function of real part, R _y(τ) auto-correlation function of the digital signal of cognitive radio signal is treated in expression, and τ is independent variable, gets nonnegative integer, the displacement of representative digit signal in discrete domain, and N represents the sub-carrier number of ofdm signal.

6. radio signal recognition method according to claim 1 is characterized in that, described method also comprises:

According to the preset rules of user to false alarm probability and detection probability, set threshold value;

Correspondingly, describedly determine according to described judgement amount whether described treating exists the elemental user signal to be specially in the cognitive radio signal:

If described judgement amount, is determined described treating more than or equal to described threshold value and is had described elemental user signal in the cognitive radio signal.

7. a radio signal recognition equipment is characterized in that, described equipment comprises:

Sampling module is used for treating cognitive radio signal and samples, and obtains the described digital signal for the treatment of cognitive radio signal;

The first computing module is used for described digital signal is carried out the calculating of auto-correlation function, obtains the auto-correlation function of described digital signal;

The second computing module is used for calculating judgement amount according to the auto-correlation function of described digital signal;

Determination module is used for determining according to described judgement amount whether the described radio signal of cognition for the treatment of exists the elemental user signal;

Wherein, described the first computing module comprises:

Estimation unit is used for described digital signal is carried out the estimation of auto-correlation function, obtains the middle auto-correlation function of described digital signal;

Computing unit is used for that auto-correlation function in the middle of described is carried out variable and simplifies, and obtains the auto-correlation function of described digital signal.

8. radio signal recognition equipment according to claim 7 is characterized in that, described equipment also comprises:

The threshold value determination module is used for according to the preset rules of user to false alarm probability and detection probability, sets threshold value;

Correspondingly, described determination module specifically is used for working as described judgement amount more than or equal to described threshold value, determines that there is described elemental user signal in described treating in the cognitive radio signal.

Images