CN108718223B - Blind spectrum sensing method for non-cooperative signals - Google Patents
Blind spectrum sensing method for non-cooperative signals Download PDFInfo
- Publication number
- CN108718223B CN108718223B CN201810291498.XA CN201810291498A CN108718223B CN 108718223 B CN108718223 B CN 108718223B CN 201810291498 A CN201810291498 A CN 201810291498A CN 108718223 B CN108718223 B CN 108718223B
- Authority
- CN
- China
- Prior art keywords
- signal
- spectrum sensing
- baseband signal
- cooperative
- detection threshold
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/382—Monitoring; Testing of propagation channels for resource allocation, admission control or handover
Abstract
The invention discloses a blind spectrum sensing method of non-cooperative signals, which provides a blind spectrum sensing method for decomposing a characteristic value of a fourth-order cumulant of baseband signals aiming at zero priori knowledge of the non-cooperative signals, calculates a maximum characteristic value and a detection threshold, and judges whether the signals exist or not by judging the maximum characteristic value and the detection threshold to realize spectrum sensing.
Description
Technical Field
The invention relates to a blind spectrum sensing method of a non-cooperative signal, belonging to the technical field of signal sensing.
Background
The spectrum sensing of non-cooperative signals (such as signals in systems of electronic reconnaissance, anti-interference, passive radar and the like) is to sense the spectrum and detect the existence of the signals under the condition that the prior knowledge is zero. The spectrum sensing of non-cooperative signals does not have any prior information on transmitted and received signals in the sensing process, which brings certain difficulty for signal sensing, faces very serious examination, and becomes a research focus in the spectrum sensing problem in recent years.
The existing spectrum sensing methods mainly comprise matched filter detection, energy detection, cyclostationary feature detection and the like, and the detection methods respectively have advantages and disadvantages. The matched filter detection needs to know the prior information of the signal, if the prior information of the signal is inaccurate, the detection result is greatly influenced, and it is difficult to accurately acquire the prior information of the signal. Cyclostationary feature detection can distinguish useful signals from noise, so the detection effect is good, but the method has high calculation complexity and requires long observation time. Energy detection is a relatively simple non-coherent detection method that can effectively detect observed signals without prior information. Therefore, energy detection technology is one of the hot spots in the research of the spectrum detection technology. But since energy detection is a signal energy as an important parameter for detection, it is susceptible to noise power uncertainty.
There are several factors that affect the perceptual performance of spectrum sensing. Firstly, due to factors such as multipath fading, hidden terminal and shadow effect, the signal-to-noise ratio of the signal to be detected reaching the receiver is very low. In addition, the noise power may vary with time, making it difficult to obtain accurate noise power. To overcome the above disadvantages, many blind spectrum sensing methods have been proposed, which are mostly based on eigenvalue decomposition of autocorrelation matrices. The estimated autocorrelation matrix has larger calculation amount and is more susceptible to the influence of Gaussian color noise, especially signals with low signal-to-noise ratio are contained in the Gaussian color noise, and the detection performance of the blind spectrum sensing method is greatly reduced.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the invention can effectively overcome the influence of Gaussian noise (including Gaussian color noise and Gaussian white noise), can improve the spectrum sensing performance when the signal-to-noise ratio is lower, and does not need to consider the signal form of the signal to be detected.
The technical scheme of the invention is as follows:
a blind spectrum sensing method of a non-cooperative signal comprises the following steps:
A. performing band-pass filtering and down-conversion processing on the received radio frequency signal to obtain a baseband signal;
B. performing time domain sampling processing on the baseband signal to obtain a baseband signal with discrete time domain;
C. obtaining a fourth-order cumulant matrix of the baseband signal according to the baseband signal with discrete time domain;
D. performing eigenvalue decomposition on the fourth-order cumulant matrix to obtain a maximum eigenvalue;
E. calculating a detection threshold according to the false alarm probability;
F. judging whether the received radio frequency signal contains a non-cooperative signal or not according to the maximum characteristic value and a detection threshold, and judging that the received radio frequency signal contains the non-cooperative signal when the maximum characteristic value is larger than the detection threshold; and when the maximum characteristic value is less than or equal to the detection threshold, judging that the received radio frequency signal does not contain a non-cooperative signal, and finishing blind spectrum sensing.
The fourth-order cumulant matrix C in the step CxThe method specifically comprises the following steps:
X＝X(n),Xl＝X(n-l)，X(n)＝[x(n),x(n-1),…x(n-L+1)]T,
wherein cum (·) is a joint cumulant, x (N) is an nth sampling value in the time-domain discrete baseband signal, and N is more than or equal to 1 and less than or equal to Ns，NsNumber of sampled values of baseband signal, N, being time-domain discretesIs a positive integer, L is 0,1, …, L-1 is a time delay, and L is a positive integer.
The detection threshold γ in the step E is specifically:
wherein N issNumber of sampled values of baseband signal, p, being time-domain discretefaIs false alarm probability
H0Where x (n) is η (n), where x (n) is the sample value of the baseband signal and η (n) represents the additive gaussian noise in an independent and identically distributed manner.
Compared with the prior art, the invention has the beneficial effects that:
(1) compared with the existing spectrum sensing hotspot technology energy detection technology, the method does not need information about noise power, is not influenced by uncertain factors of noise, and can effectively improve the detection performance of the signal under the condition of low signal-to-noise ratio.
(2) Compared with the existing blind spectrum sensing technology based on autocorrelation matrix eigenvalue decomposition, the method has the advantages that the fourth-order cumulant matrix can not be influenced by Gaussian color noise, especially, signals with low signal-to-noise ratio are submerged in the Gaussian color noise, the detection performance of the blind spectrum sensing method is greatly reduced, and the performance of spectrum sensing can still be improved.
Drawings
FIG. 1 is a block flow diagram of a spectrum sensing method of the present invention;
FIG. 2 is a schematic diagram showing the comparison of the detection rate of the method of the present invention with the energy detection method with noise uncertainty and the spectrum sensing method with autocorrelation matrix eigenvalue decomposition varying with the signal-to-noise ratio under Gaussian white noise;
fig. 3 is a schematic diagram showing the comparison of the detection rate of the method of the present invention with the energy detection method with noise uncertainty and the spectrum sensing method with autocorrelation matrix eigenvalue decomposition as the signal-to-noise ratio changes under gaussian noise.
Detailed Description
The invention provides a spectrum sensing method based on a fourth-order cumulant maximum characteristic value of a baseband signal, which is a blind spectrum sensing method and can not be influenced by Gaussian color noise. The fourth-order cumulant of the Gaussian signal is generally 0, but the non-Gaussian signal is not 0, so the fourth-order cumulant of the baseband signal is considered to be not influenced by Gaussian color noise, the effect of distinguishing the signal from the noise can be achieved by using the fourth-order cumulant, the existence of the signal to be detected can be directly judged during spectrum sensing, and the detection performance is greatly improved. Therefore, the fourth-order cumulant is used as statistical measurement to judge whether the signal to be detected exists or not in the spectrum sensing, the influence of Gaussian noise can be effectively overcome, and the spectrum sensing performance can be improved under a very low SNR.
The invention is described in further detail below with reference to the accompanying examples.
The invention provides a blind spectrum sensing method of non-cooperative signals, the flow block diagram of which is shown in figure 1, and the method mainly comprises the following steps:
A. in the system, an antenna is used for receiving a radio frequency signal with continuous time domain in a channel;
B. performing band-pass filtering and down-conversion processing on the radio-frequency signal with continuous time domain received by the antenna to obtain a corresponding baseband signal with continuous time domain; then, time domain sampling processing is carried out on the baseband signals with continuous time domains to obtain corresponding signals containing NsThe time domain discrete baseband signal of the sampling values marks the nth sampling value in the time domain discrete baseband signal as x (N), wherein N is more than or equal to 1 and less than or equal to Ns；
C. Calculating a fourth-order cumulant matrix;
the following receive matrix is defined:
X(n)＝[x(n),x(n-1),…x(n-L+1)]T，
the autocorrelation matrix of the baseband signal can be expressed as:
where L ═ 0,1, …, L-1 is time delay, L is a positive integer, and X ═ X (n), XlX (n-l), a special definition of fourth order cumulant is introduced:
wherein, cum (·) is a combined cumulant, the normalized cumulant is:
then the fourth order cumulant matrix:
Cxthe characteristic values of the symmetric matrix and the Toeplitz matrix are decomposed to obtain L characteristic values. Let λ1≥λ2≥…≥λLIs CxThe characteristic value of (2). C can be determined due to the natural blindness of the fourth order cumulant to Gaussian noisexMaximum eigenvalue λ of1As a statistical measure to determine the presence or absence of a signal to be detected, if lambda1No signal to be detected, λ, when equal to 01>0 then the signal to be detected is present. In practice, because the length of the acquired signal is limited, the value of the fourth-order cumulant of gaussian noise is not zero but is a very small value, and when detecting the signal, the corresponding threshold value γ is usually set according to the number of sampling points, the false alarm probability and the smoothing factor to perform the determination, that is, λ is1>Gamma is the presence of the signal to be detected, whereas the signal to be detected is absent.
D. For fourth-order cumulant matrix CxCalculating maximum eigenvalue lambda by eigenvalue decomposition1And judging whether the signal to be detected exists or not as the detection statistic.
E. According to a given false alarm probability pfaCalculating a detection threshold gamma;
according to false alarm probability pfaDefinition of (1)
H0:x(n)＝η(n)，
H1:x(n)＝h(n)s(n)+η(n)，
Wherein N issIs the number of sampling points, where the Q function is defined as:
H0indicating that only noise exists and the signal to be detected does not exist; h1Indicating presence of a signal to be detectedWhere x (n) and s (n) are samples of the baseband signal and the signal to be detected, respectively, η (n) represents the additive gaussian noise that is independently and identically distributed, assuming that the noise is uncorrelated with the signal, and h (n) is the channel gain. Spectrum sensing or signal detection is based on the baseband signal x (n) to determine whether there is a problem with the signal to be detected.
For a given false alarm probability pfaComprises the following steps:
detection threshold γ:
F. according to the statistic lambda1Judging whether the main signal exists or not in relation to a detection threshold gamma when the main signal is lambda1>When gamma is greater than gamma, the received radio frequency signal is judged to contain non-cooperative signal, and when lambda is greater than gamma1And when the value is less than or equal to gamma, judging that the received radio frequency signal does not contain a non-cooperative signal.
The feasibility and effectiveness of the spectrum sensing method of the present invention is further illustrated by the following simulations.
And carrying out corresponding simulation analysis on the spectrum sensing performance of the blind spectrum sensing method of the non-cooperative signal by utilizing matlab software, and comparing the simulation analysis with other spectrum sensing methods. In the ideal energy detection situation, the gaussian noise variance is generally assumed to be a constant value, however, in practical application, the gaussian noise variance is not determined to be constant and has a certain uncertainty, and the larger the noise uncertainty is, the more unstable the energy detection is. The uncertainty of noise in practical systems is typically between 1 and 2 dB. The spectrum sensing performance of the spectrum sensing method is compared with that of an ideal energy detection method, an energy detection method with 1.5dB noise uncertainty and an autocorrelation matrix eigenvalue decomposition spectrum sensing method.
FIG. 2 is a graph showing the variation of detection probability with signal-to-noise ratio (signal-to-noise ratio is from-22 dB to 0dB) of different algorithms under Gaussian white noise, and it can be seen that the detection probability of the four methods is increased along with the increase of the signal-to-noise ratio. Under the same signal-to-noise ratio, the detection probability of the spectrum sensing method is lower than that of an ideal energy detection method and higher than that of an energy detection and autocorrelation matrix eigenvalue decomposition spectrum sensing method with 1.5dB noise uncertainty. FIG. 3 is a graph of the detection probability of different algorithms under Gaussian color noise as a function of the signal-to-noise ratio (the signal-to-noise ratio is from-22 dB to 0dB), and it can be seen that the detection probability of the four methods is increased along with the increase of the signal-to-noise ratio. Under the same signal-to-noise ratio, the detection probability of the spectrum sensing method is higher than that of an ideal energy detection method, energy detection with 1.5dB of noise uncertainty and an autocorrelation matrix eigenvalue decomposition spectrum sensing method. As can be seen from fig. 2 and 3, the detection performance of the spectrum sensing method of the present invention is better than the spectrum sensing performance of the energy detection and autocorrelation matrix eigenvalue decomposition spectrum sensing method with 1.5dB noise uncertainty both under gaussian white noise and gaussian color noise, and the spectrum sensing performance of the spectrum sensing method of the present invention is also obviously better than the spectrum sensing performance of the ideal energy detection method under gaussian color noise. It can be seen that the advantages of the spectrum sensing method of the invention are more obvious in the gaussian color noise environment.
Those skilled in the art will appreciate that the details of the invention not described in detail in the specification are within the skill of those skilled in the art.
Claims (2)
1. A blind spectrum sensing method of a non-cooperative signal is characterized by comprising the following steps:
A. performing band-pass filtering and down-conversion processing on the received radio frequency signal to obtain a baseband signal;
B. performing time domain sampling processing on the baseband signal to obtain a baseband signal with discrete time domain;
C. obtaining a fourth-order cumulant matrix of the baseband signal according to the baseband signal with discrete time domain;
D. performing eigenvalue decomposition on the fourth-order cumulant matrix to obtain a maximum eigenvalue;
E. calculating a detection threshold according to the false alarm probability;
F. judging whether the received radio frequency signal contains a non-cooperative signal or not according to the maximum characteristic value and a detection threshold, and judging that the received radio frequency signal contains the non-cooperative signal when the maximum characteristic value is larger than the detection threshold; when the maximum characteristic value is less than or equal to the detection threshold, judging that the received radio frequency signal does not contain a non-cooperative signal, and finishing blind spectrum sensing;
the fourth-order cumulant matrix C in the step CxThe method specifically comprises the following steps:
X＝X(n),Xl＝X(n-l)，X(n)＝[x(n),x(n-1),…x(n-L+1)]T
wherein cum (·) is a joint cumulant, x (N) is an nth sampling value in the time-domain discrete baseband signal, and N is more than or equal to 1 and less than or equal to Ns，NsNumber of sampled values of baseband signal, N, being time-domain discretesIs a positive integer, L is 0,1, …, L-1 is a time delay, and L is a positive integer.
2. The method according to claim 1, wherein the detection threshold γ in step E specifically is:
wherein the content of the first and second substances,
Nsas the time domainNumber of discrete baseband signal samples, pfaIs false alarm probability
H0Where x (n) is η (n), where x (n) is the sample value of the baseband signal and η (n) represents the additive gaussian noise in an independent and identically distributed manner.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810291498.XA CN108718223B (en) | 2018-04-03 | 2018-04-03 | Blind spectrum sensing method for non-cooperative signals |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810291498.XA CN108718223B (en) | 2018-04-03 | 2018-04-03 | Blind spectrum sensing method for non-cooperative signals |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108718223A CN108718223A (en) | 2018-10-30 |
| CN108718223B true CN108718223B (en) | 2020-11-10 |
Family
ID=63898655
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810291498.XA Active CN108718223B (en) | 2018-04-03 | 2018-04-03 | Blind spectrum sensing method for non-cooperative signals |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108718223B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111447163B (en) * | 2020-03-30 | 2021-12-14 | 电子科技大学 | Time slot blind synchronization method for WCDMA structure signal |
| CN114070437A (en) * | 2021-11-19 | 2022-02-18 | 中国人民武装警察部队工程大学 | Joint spectrum sensing method based on energy and eigenvalue variance |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104660356B (en) * | 2015-03-13 | 2017-11-10 | 湖南大学 | A kind of collaborative frequency spectrum sensing method of half-blindness with reliable false-alarm performance |
| CN106453174B (en) * | 2016-09-18 | 2019-04-19 | 西安电子科技大学 | Cognition wireless network resource allocation methods based on signal modulate |
-
2018
- 2018-04-03 CN CN201810291498.XA patent/CN108718223B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN108718223A (en) | 2018-10-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108038419B (en) | Wi-Fi-based indoor personnel passive detection method | |
| US8135054B2 (en) | Processing of interference on a radiofrequency signal by power inversion | |
| US10972141B2 (en) | Method for estimating arrival time based on noise cancellation | |
| CN106291487A (en) | A kind of radar based on AGC voltage and echo data receives power and RCS method of estimation | |
| CN112684251B (en) | Target signal frequency domain detection method based on power spectrum template | |
| CN109474306B (en) | Detection and judgment method based on direct sequence spread spectrum signal capture | |
| CN105429719B (en) | Based on power spectrum and multi-scale wavelet transformation analysis high reject signal detection method | |
| CN109274438B (en) | Self-adaptive double-threshold cooperative interference detection method and device | |
| CN108718223B (en) | Blind spectrum sensing method for non-cooperative signals | |
| CN106100762B (en) | A kind of weak signal of communication detection method of cyclo-stationary spectrum analysis | |
| CN110398722A (en) | Extension target echo detection method based on the limited spectrum of random matrix | |
| CN107607920B (en) | Composite modulation signal analysis result verification method based on GP distribution fitting test | |
| CN109655794B (en) | Detection and identification method for suppressing interference by narrow-band self-defense noise | |
| CN104253659B (en) | Spectrum sensing method and device | |
| Mohamed et al. | Performance assessment of transient signal detection methods and superiority of energy criterion (EC) method | |
| CN113702703A (en) | Weak signal detection and identification method and system | |
| CN110299926B (en) | Underwater acoustic signal detection method oriented to low signal-to-noise ratio environment | |
| CN110568415B (en) | Signal detection method based on Arctan function under Gaussian mixture model | |
| CN110602009B (en) | BPSK blind analysis result credibility evaluation method based on CFAR criterion | |
| CN105656511B (en) | Differential correlation acquisition method suitable for environment with frequency offset and low signal-to-noise ratio | |
| CN111490956A (en) | MFSK modulation identification method based on first-order cyclostationarity | |
| CN114584227B (en) | Automatic burst signal detection method | |
| CN113259027B (en) | Method for calculating hostile-my identification signal code rate based on Haar transform | |
| CN105959035B (en) | A kind of direct sequence signal intercepts and captures detection method | |
| Kolchev et al. | Use of hazard function for signal detection on ionograms |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |