CN103888397A - BPSK signal blind identification result validity evaluation method based on HOUGH conversion - Google Patents

Abstract

The invention provides a BPSK signal blind identification result validity evaluation method based on HOUGH conversion. According to the method, on the condition without signal priori knowledge, reference signals are constructed according to signal models corresponding to modulation identification results, a module value curve after correlation accumulation of the reference signals and observation signals is analyzed is converted into a two-dimensional image of binarization, HOUGH conversion is then carried out, straight line detection is carried out, and whether the BPSK signal blind identification results are valid can be judged by detecting the number of aggregation points of the parameter space according to a straight line in the data space. The algorithm can test the validity of the blind identification results of the BPSK signals on the condition of the low signal to noise ratio, and moreover, the priori information of the signals is not needed.

Description

The efficiency assessment method of the blind recognition result of bpsk signal based on HOUGH conversion

Technical field

The present invention relates to a kind of BPSK(Binary Phase Shift Keying based on HOUGH conversion, two-phase PSK) the efficiency assessment method of Signal blind recognition result, especially a kind of compared with the efficiency assessment method of the blind recognition result of bpsk signal under Low SNR.

Background technology

Lacking under signal prior information and Low SNR, sample sequence to observation signal detects, Modulation Identification and parameter Estimation, electronic reconnaissance and cognitive radio signal (CR, Cognitive Radio) process the important step of front end, directly affect the handling property of follow-up signal processing links.In electronic reconnaissance, front end signal result exerts an influence the subsequent treatment links such as the sorting to signal, location and tracking, interference and individual radiation source identification.In cognitive radio, effectively front end frequency spectrum perception, result of spectrum analysis are the effectively prerequisites and basis of operation of the follow-up cognitive link such as frequency spectrum judging, spectrum management.But under non-cooperation condition, Modulation Identification, the parameter Estimation etc. of the sample sequence to observation signal can only be carried out blind identification.But, for whether blind recognition result is correct, effectively Efficient Evaluation method is less.The blind identifying of bpsk signal comprises the link such as Modulation Mode Recognition and decoding.Obviously the prerequisite, being correctly decoded is that Modulation Mode Recognition is correct.

HOUGH conversion is a kind of parameter estimation techniques that uses voting principle, its principle is to utilize point-line duality of image space and HOUGH parameter space, test problems in image space is transformed into parameter space, by carry out simple cumulative statistics in parameter space, then find the method detection of straight lines of accumulator peak value at HOUGH parameter space.Its essence is the pixel in image space with certain relation is carried out to cluster, searching can connect these pixels parameter space accumulation corresponding points by a certain analytical form, in the time adopting HOUGH to convert the shape of detection of straight lines, the number that only need to find the accumulation point in parameter space, the shape that adopts HOUGH to convert detection of straight lines is existing conventional method.

Summary of the invention

The technical problem to be solved in the present invention is to provide a kind of efficiency assessment method of the blind recognition result of bpsk signal based on HOUGH conversion, can realize the blind recognition result of multiple bpsk signal is carried out to efficiency assessment.

In order to address the above problem, the invention provides the efficiency assessment method of the blind recognition result of bpsk signal based on HOUGH conversion, the blind identifying of bpsk signal comprises Modulation Mode Recognition and decoding, it is characterized in that: adopt hypothesis test to assess the validity of the blind recognition result of bpsk signal, establish and in finite observation time, treat that the multiple bpsk signal of blind identification is:

$s\left(t\right)={\mathrm{Ae}}^{j(2\mathrm{\&pi;}{f}_{0}t+\mathrm{\&theta;})}\underset{k=1}{\overset{N_c}{\mathrm{\&Sigma;}}}{e}^{\mathrm{j\&pi;}{c}_k}{\mathrm{\&Pi;}}_{T_c}(t-{\mathrm{kT}}_c),0\&le;t<T$

Wherein, T is observation time, and A is signal amplitude, f ₀for signal carrier frequency, the initial phase that θ is signal, N _cfor the code length of signal, T _cfor the symbol width of signal, c _kbe k code word, value is 0 or 1;

Carrier frequency to bpsk signal, code word, code length and symbol width carry out parameter Estimation respectively, and the estimates of parameters obtaining is respectively ${\hat{f}}_0,{\hat{c}}_k,{\hat{N}}_c$ With

The superposeed sample sequence of observation signal of multiple bpsk signal of noise of foundation is:

$x\left(n\right)=s\left(n\right)+w\left(n\right)={\mathrm{Ae}}^{j(2{\mathrm{<figure-callout\; id="0"\; label="f"\; filenames="HDA0000477228280000012.png,HDA0000477228280000021.png"\; state="\{\{state\}\}">f</figure-callout>}}_0\mathrm{n\&Delta;t}+\mathrm{\&theta;})}\underset{k=1}{\overset{N_c}{\mathrm{\&Sigma;}}}{e}^{\mathrm{j\&pi;}{c}_k}{\mathrm{\&Pi;}}_{T_c}(\mathrm{n\&Delta;t}-{\mathrm{kT}}_c)+w\left(n\right),0\&le;n<N-1$

Wherein, the sampling interval that Δ t is signal, the sampled signal that s (n) is bpsk signal, w (n) is return-to-zero average limit band white Gaussian noise, and N is number of samples, and the signal to noise ratio of signal is SNR=A ²/ 2 σ ², σ ²for sample variance;

Suppose that Modulation Mode Recognition is correct and without decoding error, the modulation system of identifying is BPSK in blind identifying, and do not have decoding error, efficiency assessment method comprises the following steps:

Step 1, builds corresponding reference signal according to the Modulation Mode Recognition result of hypothesis, and the relevant cumulative delivery of sample sequence with reference to signal to observation signal, obtains corresponding relevant cumulative mould value, and concrete steps are as follows:

(1-1) according to the reference signal of bpsk signal structure be:

$y_0\left(n\right)=e^{-j2\mathrm{\&pi;}{\hat{f}}_0\mathrm{n\&Delta;t}}\underset{k=1}{\overset{{\hat{N}}_c}{\mathrm{\&Sigma;}}}{e}^{-\mathrm{j\&pi;}{\hat{c}}_k}{\mathrm{\&Pi;}}_{{\hat{T}}_c}(\mathrm{n\&Delta;t}-k{\hat{T}}_c),0\&le;n\&le;N-1$

(1-2) with reference to signal y ₀(n) do relevant adding up to the sample sequence of observation signal, obtain relevant being accumulated as:

$z\left(n\right)=\underset{m=0}{\overset{n}{\mathrm{\&Sigma;}}}x\left(m\right)y_0\left(m\right)=s_0\left(n\right)+w_0\left(n\right),0\&le;n\&le;N-1$

Wherein, w ₀(n) be noise section, s ₀(n) be signal section,

$w_0\left(n\right)=\underset{m=0}{\overset{n}{\mathrm{\&Sigma;}}}w\left(m\right)y_0\left(m\right)$

$s_0\left(n\right)=\underset{m=0}{\overset{n}{\mathrm{\&Sigma;}}}{\mathrm{Ae}}^{j(2\mathrm{\&pi;\&Delta;fn\&Delta;t}+\mathrm{\&theta;})}\underset{k=1}{\overset{{\hat{N}}_c}{\mathrm{\&Sigma;}}}{e}^{-\mathrm{j\&pi;}{\hat{c}}_k}{\mathrm{\&Pi;}}_{{\hat{T}}_c}[\mathrm{n\&Delta;t}-(k-1){\hat{T}}_c]\&CenterDot;\underset{k=1}{\overset{N_c}{\mathrm{\&Sigma;}}}{e}^{\mathrm{j\&pi;}{c}_k}{\mathrm{\&Pi;}}_{T_c}[\mathrm{n\&Delta;t}-(k-1)T_c]$

Wherein, for carrier frequency evaluated error, in the time that the value of Δ f trends towards 0, the approximation that obtains signal section is:

s ₀(n)≈A(n+1)e ^{j(πnΔfΔt+θ)}

(1-3), to relevant cumulative delivery, the relevant cumulative mould value obtaining is:

g(n)＝|z(n)|=|s ₀(n)+w ₀(n)|≈A(n+1)+ω'(n)

Wherein, A (n+1) is the certainty component of g (n), and ω ' is (n) the equivalent noise component(s) of g (n);

Step 2, carries out HOUGH conversion to relevant cumulative mould value g (n), judgement hypothesis result, and concrete steps are:

(2-1) establish h=g (n), the point (h on the data space that g (n) is described _i, n _i) be converted to the point (ρ on parameter space _i, θ _i), conversion formula is:

ρ _i＝n _icosθ _i+h _isinθ _i

(2-2) number of accumulation point on statistical parameter space, when the number of accumulation point on parameter space is 1, sentences Modulation Mode Recognition correctly and without decoding error, supposes to set up; When the number of accumulation point on parameter space is not 1, sentence Modulation Mode Recognition mistake or have decoded in error, supposing is false;

Step 3, tests to the judged result of step 2, each θ on calculating parameter space _istandard deviation, obtain standard deviation characteristic quantity C and be:

C＝std(θ _i)

Wherein, std () is for asking standard deviation function, as C < C _th, accuracy of judgement, the blind recognition result of bpsk signal is effective; As C>=C _th, misjudgment, the blind recognition result of bpsk signal is invalid, C _thfor decision threshold, 4≤C _th≤ 8.

Beneficial effect of the present invention is: can carry out effective efficiency assessment to the blind recognition result of bpsk signal, effectively tell Modulation Mode Recognition correct and without decoding error and Modulation Mode Recognition mistake or there is decoded in error.

Brief description of the drawings

Fig. 1 is the software flow pattern of the efficiency assessment method of the blind recognition result of bpsk signal based on HOUGH conversion of the present invention;

Fig. 2 is that the rear identification of HOUGH conversion of the relevant cumulative mould value of bpsk signal of the present invention is correct, the emulation schematic diagram of decoding errorless;

Fig. 3 is that the rear BSPK of HOUGH conversion of the relevant mould value that adds up of bpsk signal of the present invention identifies emulation schematic diagram correct but that have a bit-errors to decode;

Fig. 4 is that the rear BPSK mistake of the HOUGH conversion of the relevant cumulative mould value of bpsk signal of the present invention is known the emulation schematic diagram for normal signal;

Fig. 5 is that the rear BPSK mistake of the HOUGH conversion of the relevant cumulative mould value of bpsk signal of the present invention is known for linear FM signal emulation schematic diagram;

Fig. 6 is that the rear BPSK mistake of the HOUGH conversion of the relevant cumulative mould value of bpsk signal of the present invention is known the emulation schematic diagram for QPSK signal;

Fig. 7 is the correct and HOUGH Transformation Graphs of relevant accumulation curve under SNR=6dB during without decoding error of Modulation Identification of the present invention;

Fig. 8 is the correct and HOUGH Transformation Graphs of relevant accumulation curve under SNR=0dB during without decoding error of Modulation Identification of the present invention;

HOUGH Transformation Graphs when Fig. 9 is correct but decoding one bit-errors of Modulation Mode Recognition of the present invention under SNR=6dB;

HOUGH Transformation Graphs when Figure 10 is correct but decoding one bit-errors of Modulation Mode Recognition of the present invention under SNR=0dB;

Figure 11 is that BPSK mistake of the present invention is known the HOUGH Transformation Graphs under SNR=6dB while being normal signal;

Figure 12 is that BPSK mistake of the present invention is known the HOUGH Transformation Graphs under SNR=0dB while being normal signal;

Figure 13 is that bpsk signal mistake of the present invention is known the HOUGH Transformation Graphs under SNR=6dB while being LFM signal;

Figure 14 is that bpsk signal mistake of the present invention is known the HOUGH Transformation Graphs under SNR=0dB while being LFM signal;

Figure 15 is that bpsk signal mistake of the present invention is known the HOUGH Transformation Graphs under SNR=6dB while being QPSK signal;

Figure 16 is that bpsk signal mistake of the present invention is known the HOUGH Transformation Graphs under SNR=6dB while being QPSK signal.

Embodiment

As shown in Figure 1, the invention provides the efficiency assessment method of the blind recognition result of bpsk signal based on HOUGH conversion, the blind identifying of bpsk signal comprises Modulation Mode Recognition and decoding, it is characterized in that: adopt hypothesis test to assess the validity of the blind recognition result of bpsk signal, establish and in finite observation time, treat that the multiple bpsk signal of blind identification is:

$s\left(t\right)={\mathrm{Ae}}^{j(2\mathrm{\&pi;}{f}_{0}t+\mathrm{\&theta;})}\underset{k=1}{\overset{N_c}{\mathrm{\&Sigma;}}}{e}^{\mathrm{j\&pi;}{c}_k}{\mathrm{\&Pi;}}_{T_c}(t-{\mathrm{kT}}_c),0\&le;t<T$

Wherein, T is observation time, and A is signal amplitude, f ₀for signal carrier frequency, the initial phase that θ is signal, N _cfor the code length of signal, T _cfor the symbol width of signal, c _kbe k code word, value is 0 or 1;

Carrier frequency to bpsk signal, code word, code length and symbol width carry out parameter Estimation respectively, and the estimates of parameters obtaining is respectively ${\hat{f}}_0,{\hat{c}}_k,{\hat{N}}_c$ With

The superposeed sample sequence of observation signal of multiple bpsk signal of noise of foundation is:

$x\left(n\right)=s\left(n\right)+w\left(n\right)={\mathrm{Ae}}^{j(2{\mathrm{<figure-callout\; id="0"\; label="f"\; filenames="HDA0000477228280000012.png,HDA0000477228280000021.png"\; state="\{\{state\}\}">f</figure-callout>}}_0\mathrm{n\&Delta;t}+\mathrm{\&theta;})}\underset{k=1}{\overset{N_c}{\mathrm{\&Sigma;}}}{e}^{\mathrm{j\&pi;}{c}_k}{\mathrm{\&Pi;}}_{T_c}(\mathrm{n\&Delta;t}-{\mathrm{kT}}_c)+w\left(n\right),0\&le;n<N-1$

Wherein, the sampling interval that Δ t is signal, the sampled signal that s (n) is bpsk signal, w (n) is return-to-zero average limit band white Gaussian noise, and N is number of samples, and the signal to noise ratio of signal is SNR=A ²/ 2 σ ², σ ²for sample variance;

Suppose that Modulation Mode Recognition is correct and without decoding error, the modulation system of identifying is BPSK in blind identifying, and do not have decoding error, efficiency assessment method comprises the following steps:

Step 1, builds corresponding reference signal according to the Modulation Mode Recognition result of hypothesis, and the relevant cumulative delivery of sample sequence with reference to signal to observation signal, obtains corresponding relevant cumulative mould value, and concrete steps are as follows:

(1-1) according to the reference signal of bpsk signal structure be:

$y_0\left(n\right)=e^{-j2\mathrm{\&pi;}{\hat{f}}_0\mathrm{n\&Delta;t}}\underset{k=1}{\overset{{\hat{N}}_c}{\mathrm{\&Sigma;}}}{e}^{-\mathrm{j\&pi;}{\hat{c}}_k}{\mathrm{\&Pi;}}_{{\hat{T}}_c}(\mathrm{n\&Delta;t}-k{\hat{T}}_c),0\&le;n\&le;N-1$

(1-2) with reference to signal y ₀(n) do relevant adding up to the sample sequence of observation signal, obtain relevant being accumulated as:

$z\left(n\right)=\underset{m=0}{\overset{n}{\mathrm{\&Sigma;}}}x\left(m\right)y_0\left(m\right)=s_0\left(n\right)+w_0\left(n\right),0\&le;n\&le;N-1$

Wherein, w ₀(n) be noise section, s ₀(n) be signal section,

$w_0\left(n\right)=\underset{m=0}{\overset{n}{\mathrm{\&Sigma;}}}w\left(m\right)y_0\left(m\right)$

$s_0\left(n\right)=\underset{m=0}{\overset{n}{\mathrm{\&Sigma;}}}{\mathrm{Ae}}^{j(2\mathrm{\&pi;\&Delta;fn\&Delta;t}+\mathrm{\&theta;})}\underset{k=1}{\overset{{\hat{N}}_c}{\mathrm{\&Sigma;}}}{e}^{-\mathrm{j\&pi;}{\hat{c}}_k}{\mathrm{\&Pi;}}_{{\hat{T}}_c}[\mathrm{n\&Delta;t}-(k-1){\hat{T}}_c]\&CenterDot;\underset{k=1}{\overset{N_c}{\mathrm{\&Sigma;}}}{e}^{\mathrm{j\&pi;}{c}_k}{\mathrm{\&Pi;}}_{T_c}[\mathrm{n\&Delta;t}-(k-1)T_c]$

Wherein,

for carrier frequency evaluated error, in the time that the value of Δ f trends towards 0, the approximation that obtains signal section is:

s ₀(n)≈A(n+1)e ^{j(πnΔfΔt+θ)}

(1-3), to relevant cumulative delivery, the relevant cumulative mould value obtaining is:

g(n)＝|z(n)|=|s ₀(n)+w ₀(n)|≈A(n+1)+ω'(n)

Wherein, A (n+1) is the certainty component of g (n), and ω ' is (n) the equivalent noise component(s) of g (n);

Step 2, carries out HOUGH conversion to relevant cumulative mould value g (n), judgement hypothesis result, and concrete steps are:

(2-1) establish h=g (n), the point (h on the data space that g (n) is described _i, n _i) be converted to the point (ρ on parameter space _i, θ _i), conversion formula is:

ρ _i＝n _icosθ _i+h _isinθ _i

Parameter space is carried out to space quantization, obtain two-dimentional accumulator matrix M (ρ _i, θ _i), M (ρ _i, θ _i) initial value be set as 0, to the each point (h on h=g (n) _i, n _i) change, the present invention can get (h _h, n _i) ∈ (120,120), all quantized value substitution conversion formulas of θ ∈ (pi/2, pi/2) are calculated to the ρ with respect to each point _ivalue, in the time respectively each point being calculated, need to add 1 to unit corresponding with this point in accumulator, i.e. M[ρ _i] [θ _i]=M[ρ _i] [θ _i]+1, resets a threshold value, according to threshold value at two-dimentional accumulator matrix M (ρ _i, θ _i) the middle point that is greater than threshold values of finding, these points that are greater than threshold values are the peak dot of HOUGH conversion, accumulation point in corresponding parameter space;

(2-2) number of accumulation point on statistical parameter space, when the number of accumulation point on parameter space is 1, sentences Modulation Mode Recognition correctly and without decoding error, supposes to set up; When the number of accumulation point on parameter space is not 1, sentence Modulation Mode Recognition mistake or have decoded in error, supposing is false;

Step 3, tests to the judged result of step 2, each θ on calculating parameter space _istandard deviation, obtain standard deviation characteristic quantity C and be:

C＝std(θ _i)

Wherein, std () is for asking standard deviation function, as C < C _th, accuracy of judgement, the blind recognition result of bpsk signal is effective; As C>=C _th, misjudgment, the blind recognition result of bpsk signal is invalid, C _thfor decision threshold, 4≤C _th≤ 8.

As shown in Figure 2, when the blind recognition result of bpsk signal is that Modulation Mode Recognition is while correctly and not existing decoding error, in the time calculating relevant cumulative mould value g (n), if the value of carrier frequency evaluated error Δ f trends towards at 0 o'clock, the approximation that can obtain g (n) is:

g(n)＝|z(n)|=|s ₀(n)+w ₀(n)|≈A(n+1)+ω'(n)

Wherein, A (n+1) is the certainty component of g (n), ω ' is (n) the equivalent noise component of g (n), and relevant cumulative mould value g (n) is approximately one by the straight line of noise pollution, and it is done only to occur a peak dot after HOUGH conversion.

As shown in Figure 3, when Modulation Mode Recognition is correct, but the parameter estimating error of multiple bpsk signal is larger, likely because the accumulation of carrier frequency evaluated error or symbol width evaluated error is larger, or because the reason such as the estimation of code element figure place is wrong will cause decoding error.If there is a bit-errors decoding, at symbol decoding errors present, the curve of relevant cumulative mould value g (n) fractures generation, formation decoded in error interval, the interval curve for relevant cumulative mould value g (n) of this decoded in error produces in decoded in error position to fracture and forms one with the starting point that fractures and the interval of terminal formation.Now, relevant cumulative mould value g (n) curve is by several the rectilinear(-al)s that slope is different, relevant cumulative mould value g (n) is after HOUGH conversion, near the peak-peak of its waveform, also there are several peak values,, in fact, in relevant cumulative mould value g (n), there is the straight line of three Different Slope in unlikely one of visible peak dot.

As shown in Figure 4,5, 6, in the time that multiple bpsk signal bandwidth is less or signal is disturbed generation distortion, the blind recognition result of multiple bpsk signal may be Modulation Mode Recognition mistake, is likely judged to normal signal or other signal at receiving terminal.Suppose the Modulation Mode Recognition mistake of bpsk signal, bpsk signal is known for other signals by mistake, is normal signal if mistake is known, and in like manner can construct reference signal according to normal signal and be:

$y_1\left(n\right)=e^{-j\left(2\mathrm{\&pi;}{\hat{f}}_0\mathrm{n\&Delta;t}\right)},0\&le;n\&le;N-1$

Again with reference to signal y ₁(n) do relevant adding up to the sample sequence of observation signal, obtain relevant being accumulated as:

$z\left(n\right)=\underset{m=0}{\overset{n}{\mathrm{\&Sigma;}}}x\left(m\right)y_1\left(m\right)=s_1\left(n\right)+w_1\left(n\right),0\&le;n\&le;N-1$

Wherein, w ₁(n) be noise section, s ₁(n) be signal section,

Wherein,

for phase error function,

To relevant cumulative delivery, the relevant cumulative mould value obtaining is again:

g(n)＝|z(n)|=|s ₁(n)+w ₁(n)|。

Fig. 4, Fig. 5 and Fig. 6 are respectively mistake to be known while being normal signal, LFM signal and QPSK signal, the waveform of relevant cumulative mould value g (n) after HOUGH conversion, obviously due to Modulation Identification erroneous results, cause relevant cumulative mould value g (n) not to be straight line, there are many straight lines, or curve, thereby make the peak value of HOUGH conversion more than one.

As shown in Fig. 7-16, in actual applications, due to the reason such as signal to noise ratio, sample number, in peak dot position, can produce some pseudo-accumulation points, therefore decision rule should be revised, in each blockage region in Fig. 7,9,11,13 and 15, be wherein accumulation point.The schematic diagram of relevant cumulative mould value g (n) curve after HOUGH conversion under different situations shown in Fig. 7-Figure 16.The simulated conditions that Fig. 7-Figure 16 adopts is: the sample sequence model x (n) that supposes observation signal is the multiple bpsk signal being polluted by additive white Gaussian noise, and signal to noise ratio is 3dB, and carrier frequency is 10.05MHz, and symbol width is 1 μ s, code sequence is 13 Barker codes, and code sequence is [1,1,1,1,1,0,0,1,1,0,1,0,1], sample frequency f _sfor 100MHz, sample length is 1300 points, and the simulation times under three kinds of situations is all 1000 times.From Fig. 7-Figure 16: under different situations, its accumulation point distributes variant identical, as shown in FIG. 7 and 8, correct and decode when errorless under hypothesis in bpsk signal Modulation Mode Recognition, accumulation point overlaps substantially, therefore utilizes the straight line that line detection method detects to be parallel to each other or to overlap; As shown in Fig. 9-16, the various situations in bpsk signal Modulation Mode Recognition mistake or existence under more than one decoded in error hypothesis, accumulation point comparatively disperses, and the straight line detecting intersects, and does not obviously belong to a line.

As shown in Fig. 9-16, in primitive curve, part straight line is not detected, be mainly because HOUGH carry out extreme point while detecting its thresholding select too low due to.Therefore, in order to improve accuracy of detection, to the θ of each point on parameter space _ivalue is calculated standard deviation, obtains standard deviation characteristic quantity C and is:

C＝std(θ _i)

Wherein, std () is for asking standard deviation function, as C < C _th, accuracy of judgement, the blind recognition result of bpsk signal is effective; As C>=C _th, misjudgment, the blind recognition result of bpsk signal is invalid, C _thfor decision threshold, 4≤C _th≤ 8.

Referring to table 1, the performance of the blind recognition result validity check of bpsk signal is added up, the sample sequence x (n) that supposes the observation signal receiving is the multiple bpsk signal being polluted by additive white Gaussian noise, carrier frequency 10.05MHz, symbol width 1 μ s, code sequence is 13 Barker codes, code sequence is [1,1,1,1,1,0,0,1,1,0,1,0,1], sample frequency f _s=100MHZ, sample length is 1300 points, simulation times 1000 times.If definition: H0 to be Modulation Mode Recognition correct and without decoding error, H1 be defined as Modulation Mode Recognition mistake or Modulation Mode Recognition correct but have a decoding error, in table, n ₀₀represent that reality is H0, be judged as the number of times of H0; n ₀₁represent actual H0, be judged to the number of times of H1; n ₁₀represent that reality is H1, be judged to the number of times of H0; n ₁₁represent that reality is H1, be judged to the number of times of H1; Two class error probability sums are P _e=(n ₁₀+ n ₀₁)/1000.

From table 1: this algorithm is within the scope of appropriate signal to noise ratio, and thresholding is selected when suitable, can effectively complete the validity check to the blind recognition result of BPSK.When signal to noise ratio is more than or equal to 0dB, in 1000 emulation, recognizer Modulation Identification result used all correct and parameter Estimation is more accurate, without decoding error, be only 9 o'clock at thresholding, have 1 H ₁mistake mistake is H ₀, two class error probabilities of check algorithm approach 0; When signal to noise ratio is during at-3dB, in 1000 emulation, 971 H ₀identification is correct, 4 identification errors, 23 H ₁identification is correct, 2 H ₁know mistake, two class error probabilities 0.6%, error detection rate 92% by mistake; When signal to noise ratio is during at-4dB, in 1000 emulation, approximately 900 H ₀identification is correct, approximately 20 identification errors, 23 H ₁identification is correct, 2 H ₁know mistake, two class error probabilities approximately 4.0%, error detection rate 72% by mistake; When after Xin Zao Bi≤-5dB, error rate sharply declines.Therefore,, from simulation result, the validity check method of the blind recognition result of bpsk signal based on HOUGH conversion in this paper, compared with still carrying out validity check to the blind recognition result of bpsk signal under Low SNR, possesses good error detection distinguishing ability.But along with the reduction of signal to noise ratio, two class error probabilities of existence sharply increase, also variation thereupon of the performance of validity check.

Table 1

Claims (1)

1. the efficiency assessment method of the blind recognition result of bpsk signal based on HOUGH conversion, the blind identifying of bpsk signal comprises Modulation Mode Recognition and decoding, it is characterized in that: adopt hypothesis test to assess the blind recognition result of bpsk signal, establish and in finite observation time, treat that the multiple bpsk signal of blind identification is:

$s\left(t\right)={\mathrm{Ae}}^{j(2\mathrm{\&pi;}{f}_{0}t+\mathrm{\&theta;})}\underset{k=1}{\overset{N_c}{\mathrm{\&Sigma;}}}{e}^{\mathrm{j\&pi;}{c}_k}{\mathrm{\&Pi;}}_{T_c}(t-{\mathrm{kT}}_c),0\&le;t<T$

Wherein, T is observation time, and A is signal amplitude, f ₀for signal carrier frequency, the initial phase that θ is signal, N _cfor the code length of signal, T _cfor the symbol width of signal, c _kbe k code word, value is 0 or 1;

Carrier frequency to bpsk signal, code word, code length and symbol width carry out parameter Estimation respectively, and the estimates of parameters obtaining is respectively ${\hat{f}}_0,{\hat{c}}_k,{\hat{N}}_c$ With

The superposeed sample sequence of observation signal of multiple bpsk signal of noise of foundation is:

$x\left(n\right)=s\left(n\right)+w\left(n\right)={\mathrm{Ae}}^{j(2f_0\mathrm{n\&Delta;t}+\mathrm{\&theta;})}\underset{k=1}{\overset{N_c}{\mathrm{\&Sigma;}}}{e}^{\mathrm{j\&pi;}{c}_k}{\mathrm{\&Pi;}}_{T_c}(\mathrm{n\&Delta;t}-{\mathrm{kT}}_c)+w\left(n\right),0\&le;n<N-1$

Wherein, the sampling interval that Δ t is signal, the sampled signal that s (n) is bpsk signal, w (n) is return-to-zero average limit band white Gaussian noise, and N is number of samples, and the signal to noise ratio of signal is SNR=A ²/ 2 σ ², σ ²for sample variance;

Suppose that Modulation Mode Recognition is correct and without decoding error, the modulation system of identifying is BPSK in blind identifying, and do not have decoding error, efficiency assessment method comprises the following steps:

Step 1, builds corresponding reference signal according to the Modulation Mode Recognition result of hypothesis, and the relevant cumulative delivery of sample sequence with reference to signal to observation signal, obtains corresponding relevant cumulative mould value, and concrete steps are as follows:

(1-1) according to the reference signal of bpsk signal structure be:

$y_0\left(n\right)=e^{-j2\mathrm{\&pi;}{\hat{f}}_0\mathrm{n\&Delta;t}}\underset{k=1}{\overset{{\hat{N}}_c}{\mathrm{\&Sigma;}}}{e}^{-\mathrm{j\&pi;}{\hat{c}}_k}{\mathrm{\&Pi;}}_{{\hat{T}}_c}(\mathrm{n\&Delta;t}-k{\hat{T}}_c),0\&le;n\&le;N-1$

(1-2) with reference to signal y ₀(n) do relevant adding up to the sample sequence of observation signal, obtain relevant being accumulated as:

$z\left(n\right)=\underset{m=0}{\overset{n}{\mathrm{\&Sigma;}}}x\left(m\right)y_0\left(m\right)=s_0\left(n\right)+w_0\left(n\right),0\&le;n\&le;N-1$

Wherein, w ₀(n) be noise section, s ₀(n) be signal section,

$w_0\left(n\right)=\underset{m=0}{\overset{n}{\mathrm{\&Sigma;}}}w\left(m\right)y_0\left(m\right)$

$s_0\left(n\right)=\underset{m=0}{\overset{n}{\mathrm{\&Sigma;}}}{\mathrm{Ae}}^{j(2\mathrm{\&pi;\&Delta;fn\&Delta;t}+\mathrm{\&theta;})}\underset{k=1}{\overset{{\hat{N}}_c}{\mathrm{\&Sigma;}}}{e}^{-\mathrm{j\&pi;}{\hat{c}}_k}{\mathrm{\&Pi;}}_{{\hat{T}}_c}[\mathrm{n\&Delta;t}-(k-1){\hat{T}}_c]\&CenterDot;\underset{k=1}{\overset{N_c}{\mathrm{\&Sigma;}}}{e}^{\mathrm{j\&pi;}{c}_k}{\mathrm{\&Pi;}}_{T_c}[\mathrm{n\&Delta;t}-(k-1)T_c]$

Wherein,

for carrier frequency evaluated error, in the time that the value of Δ f trends towards 0, the approximation that obtains signal section is:

s ₀(n)≈A(n+1)e ^{j(πnΔfΔt+θ)}

(1-3), to relevant cumulative delivery, the relevant cumulative mould value obtaining is:

g(n)＝|z(n)|=|s ₀(n)+w ₀(n)|≈A(n+1)+ω'(n)

Wherein, A (n+1) is the certainty component of g (n), and ω ' is (n) the equivalent noise component(s) of g (n);

Step 2, carries out HOUGH conversion to relevant cumulative mould value g (n), judgement hypothesis result, and concrete steps are:

(2-1) establish h=g (n), the point (h on the data space that g (n) is described _i, n _i) be converted to the point (ρ on parameter space _i, θ _i), conversion formula is:

ρ _i＝n _icosθ _i+h _isinθ _i

(2-2) number of accumulation point on statistical parameter space, when the number of accumulation point on parameter space is 1, sentences Modulation Mode Recognition correctly and without decoding error, supposes to set up; When the number of accumulation point on parameter space is not 1, sentence Modulation Mode Recognition mistake or have decoded in error, supposing is false;

Step 3, tests to the judged result of step 2, each θ on calculating parameter space _istandard deviation, obtain standard deviation characteristic quantity C and be:

C＝std(θ _i)

Wherein, std () is for asking standard deviation function, as C < C _th, accuracy of judgement, the blind recognition result of bpsk signal is effective; As C>=C _th, misjudgment, the blind recognition result of bpsk signal is invalid, C _thfor decision threshold, 4≤C _th≤ 8.

Images