Background technology
Along with the development of radio traffic, frequency spectrum resource becomes more and more nervous, thus promoted the development of cognitive radio technology, cognitive radio technology is by real-time perception information frequently, thus distribute flexibly and frequency of utilization resource, enable unauthorized user wait for an opportunity to share the frequency spectrum resource of authorized user, thus substantially increase frequency spectrum service efficiency, therefore, cognitive radio technology has been acknowledged as the technical foundation of next generation wireless communication network.Frequency spectrum perception is then the top priority of cognitive radio, cognitive user detects frequency spectrum cavity-pocket by perceived spectral environment, and can the perception of cognitive user decide cognitive radio system and have stronger adaptability, wider bandwidth, less conflict and the higher availability of frequency spectrum.Cognition technology in cognitive radio, not only requirement can carry out continuous perception in wide-band, unknown spatial domain, and require cognitive user under various circumstances must accurately, around real-time fast perception by information frequently, thus while the availability of frequency spectrum ensureing cognitive radio networks, avoid bringing harmful interference because of the frequency spectrum perception result that perception is inaccurate or postpone to primary user.
Publication number CN 103780317 A discloses a kind of double-threshold cooperative spectrum sensing method based on degree of belief, the method comprises: System Priority utilizes the cognitive nodes meeting double threshold requirement to participate in collaborative sensing, when meeting the cognitive nodes number deficiency of double threshold requirement, increasing the cognitive nodes meeting degree of belief parameter request and participate in collaborative sensing. fusion center stores the detection record of cognitive nodes, and arranges fusion weight as local testing result.The perceptual parameters of the required transmission of the method decreases, and the channel width taken reduces.Meanwhile, due to the minimizing of unreliable cognitive nodes, the detection perform of the method improves further.
The present invention relates to a kind of cooperative frequency spectrum sensing method based on multi-threshold, the method is transparent to wireless device technical approach, implementation method does not have the requirement of particular job standard to wireless device, highly versatile, strong robustness, frequency spectrum perception verification and measurement ratio is high, can be widely used in the radio art needing frequency spectrum perception, as cognitive radio, sensor network and monitoring radio-frequency spectrum, but do not limit to the above scope enumerated.
Summary of the invention
The object of the present invention is to provide that a kind of frequency spectrum detection rate is high, robustness good, the cooperative frequency spectrum sensing method based on multi-threshold of highly versatile.
Based on the cooperative frequency spectrum sensing method of multi-threshold, it is characterized in that, the method sets multiple different energy threshold λ in received energy regional extent
i, solved each threshold value by threshold value method for solving, cognitive user is based on solving the different-energy thresholding λ obtained
idetect primary user's signal, the signal energy value detected drops on different energy area E
iinside then have different detection probability Q
dwith false alarm probability Q
f.When M user carries out collaborative sensing, according to each energy area E
iinterior number of users D
iadd accordingly and advise coefficient w
i, in conjunction with given decision threshold D
t, according to judgement formula
primary user's signal is adjudicated, as D>=D
t, then primary user is adjudicated just in use authority frequency range; Otherwise then adjudicate the non-use authority frequency range of primary user.
The present invention compared with prior art, its remarkable advantage is: 1, in received energy regional extent, set multiple different energy threshold, by threshold value method for solving, each threshold value is solved, cognitive user detects primary user's signal based on solving the different-energy thresholding obtained, and improves the adaptive capacity to environment; 2, when user carries out collaborative sensing, according to number of users in each energy area that multi-threshold divides with add accordingly and advise coefficient, in conjunction with given decision threshold, primary user's signal is adjudicated, improve the data fusion efficiency between collaboration user and the verification and measurement ratio to primary user's signal; 3, utilize the cooperative frequency spectrum sensing method based on multi-threshold, effectively can realize collaborative spectrum sensing, cognitive radio, sensor network and monitoring radio-frequency spectrum field can be widely used in, but do not limit to the above scope enumerated.
Embodiment
For further illustrating the present invention, provide an embodiment below in conjunction with accompanying drawing 1 and accompanying drawing 2, the present embodiment is only limitted to a kind of implementation method of the present invention is described, does not represent the restriction to coverage of the present invention.
Collaborative spectrum sensing based on multi-threshold sets multiple different decision threshold in received energy regional extent, and as shown in Figure 1, the signal energy value that cognitive user detects drops on different energy area E
iin.For three Threshold detection, suppose to define three different decision thresholds, be respectively λ
1, λ
2, λ
3, wherein three decision thresholds are divided into E energy space
0, E
1, E
2and E
3four regions, there is a corresponding weights coefficient in each energy area.The false alarm probability that each decision threshold is corresponding and detection probability are respectively P
f1, P
f2, P
f3and P
d1, P
d2, P
d3, the weights coefficient of each energy area is respectively w
0=0, w
1=1, w
2=L, w
3=L
2, then energy weighted sum is:
In formula: D
irepresent that the energy value of detection signal falls into the number of region i, L is default value.Suppose to define D
t=L
2, by weighted sum D and D
tcompare, if D>=D
t, then primary user is adjudicated just in use authority frequency range; Otherwise then adjudicate the non-use authority frequency range of primary user.
When cognitive user carries out collaborative sensing according to multi-threshold detection model, suppose that the cognitive user number participating in detecting is M, in the testing result of user, 0 energy value falls into region E
3, i energy value falls into region E
2, j-i energy value falls into region E
1, then M-j energy value falls into region E
0if, weighted sum D≤D
t(wherein D=(M-j) w
0+ (j-i) w
1+ iw
2), then judgement is the non-use authority frequency range of primary user; Otherwise, then primary user is adjudicated just in use authority frequency range.
When threshold number expands to N number of, whole observation energy value is divided into N+1 region by N number of thresholding, and as shown in Figure 1, M collaborative sensing user carries out collaborative sensing, and the decision rule that setting data fusion center adopts is:
Wherein, H
1represent that authorized user is just in use authority frequency range, H
0represent the non-use authority frequency range of authorized user, D
ibe expressed as energy value and fall into region E
iin collaboration user number, w
ifor the weights that each energy area is corresponding, its value is set to:
Threshold value λ
ideterministic process as shown in Figure 2, specifically describes as follows:
Step one: the initiation parameter needed for input calculates, comprises sampling number N
1, collaboration user number M, iterations m, step-size in search Δ, target cooperation false alarm probability
Step 2: the initial vector λ of stochastic generation
0, calculate the false alarm probability Q of cooperative node
f, verify whether it meets constraints
if do not meet, then regenerate initial vector λ
0recalculate Q
f; If meet, then obtain the cooperative detection probability Q of network
d(λ
0), and make Q
d max=Q
d(λ
0).Make basic point vector λ
band target cooperative detection maximum probability value Q
d maxcorresponding vectorial λ
optbe λ
0, i.e. λ
b=λ
opt=λ
0, wherein
and
Step 3: set up reference axis, vectorial λ
binterior element is as the point in reference axis, and take Δ as step-length, the positive and negative direction finding along each element meets constraints
new growing point vector λ, obtain corresponding target cooperative detection probability Q
d(λ) and and Q
d maxcompare, target cooperative detection probability is greater than Q
d maxvectorial λ (λ=[λ
1, λ
2..., λ
n], 0 < λ
1< λ
2< ... < λ
n) put into point set.
Step 4: find out the target cooperative detection probability Q in point set in vectorial λ
d(λ) the maximum, and replace λ
opt=λ and Q
d max=Q
d(λ).
Step 5: judge whether to arrive iterations m, if meet, then export Q
d maxwith vectorial λ
opt, calculate and terminate; If do not meet, then return execution step 2.