A Multi-UAV Cooperative Spectrum Sensing Method Based on the Quantum Sooty Tern Mechanism

Abstract

The present invention provides a multi-UAV collaborative spectrum sensing method based on the quantum sooty tern mechanism, establishes a multi-UAV spectrum sensing model based on the quantum sooty tern mechanism, aims at the detection probability of spectrum sensing technology, designs a sooty tern quantum position evolution mechanism with quantum coding, obtains a new quantum sooty tern mechanism method, uses the position of the quantum sooty tern as the weight vector of cognitive UAV users, and finally calculates the optimal weight vector. The quantum sooty tern mechanism overcomes the disadvantages of poor convergence performance of previous classical algorithms and improves the optimization rate. The present invention designs a multi-UAV collaborative spectrum sensing method based on the quantum sooty tern mechanism, and the purpose of this method is to obtain the optimal weight vector of cognitive UAV users to determine the contribution of cognitive UAV users to global sensing, providing advantageous conditions for subsequent spectrum allocation for UAVs.

Description

Technical Field

[0001] The present invention relates to a multi-UAV collaborative spectrum sensing method based on a quantum sooty tern mechanism, and belongs to the field of UAV spectrum sensing. Background Technique

[0002] Unmanned aerial vehicles (UAVs) are an emerging technology and have been widely used in military, civilian and other fields. On the one hand, it is because the cost of UAVs is relatively low, and on the other hand, the reliability of collaborative work is high. In a UAV system, communication between UAVs is an important part of the system. Therefore, the research on UAV networks is particularly important. UAV networks generally have the following characteristics: no central node, the UAV network belongs to a distributed architecture, and communication can be carried out between UAVs; self-organization, the UAV network does not require fixed network communication facilities; self-healing and dynamic changes in the network topology, etc.

[0003] Spectrum sensing is a key technology of cognitive radio. Its main function is to detect spectrum holes available for cognitive users and monitor the activities of primary user signals at the same time. This technology enables cognitive UAV users to opportunistically utilize authorized or unauthorized frequency bands. Therefore, cognitive UAV users can use the idle spectrum to work continuously without affecting the communication quality of primary users. The currently proposed spectrum sensing methods mainly include matched filter detection, cyclostationary feature detection, and multi-resolution spectrum sensing, etc. These spectrum sensing methods are all single-node sensing methods. However, in the case of shadow and deep fading, the sensing results of a single node are unreliable. Therefore, the present invention uses a multi-UAV collaborative spectrum sensing technology, which can overcome the above disadvantages.

[0004] According to the existing literature, Zhang Hongwei et al. in the "Research on Multi-UAV Cooperative Spectrum Sensing in Cognitive UAV Networks" published in the Journal of Air Force Engineering University (2020, 21(1): 92-98), combined with the problem of shortage of spectrum resources, established a cognitive UAV network model, and proposed an optimal fusion criterion to study the performance of multi-UAV cooperative spectrum sensing. Finally, the performance of this method was proved through certain experimental simulations. However, this article did not use an optimization method to obtain the total error rate, and did not discuss the weights of cognitive UAV users, which will make it difficult to determine the contribution of each UAV to the global sensing.

[0005] In summary, the above literature has made certain contributions to the spectrum sensing of multiple UAVs, but does not discuss the weights of cognitive UAV users. Therefore, the present invention establishes a multi-UAV spectrum sensing model based on the quantum sooty tern mechanism. With the detection probability of spectrum sensing technology as the goal, a sooty tern quantum position evolution mechanism with quantum coding is designed to obtain a new quantum sooty tern mechanism method. Using the position of the quantum sooty tern as the weight vector of cognitive UAV users, the optimal weight vector is finally calculated. The quantum sooty tern mechanism overcomes the disadvantages of poor convergence performance of previous classical algorithms and improves the optimization rate. Summary of the Invention

[0006] The present invention designs a multi-UAV cooperative spectrum sensing method based on the quantum sooty tern mechanism. The purpose of this method is to obtain the optimal weight vector of cognitive UAV users to determine the contribution of cognitive UAV users to global sensing, providing advantageous conditions for subsequent spectrum allocation to UAVs.

[0007] The purpose of the present invention is achieved as follows: The steps are as follows:

[0008] Step 1: Establish a multi-UAV cooperative spectrum sensing model.

[0009] Assume that there are a total of N cognitive UAV users, and the set of cognitive UAV users is U = {U1, U2, …, U N}. The number of samples of each cognitive UAV user is T, and the signals sampled from N cognitive UAV users can form an N×T-dimensional vector matrix X N×T . At the t-th moment, the binary hypothesis testing model of the cognitive UAV user U n can be expressed as where n = 1, 2…, N, t = 0, 1, … T-1, s(t) is the signal transmitted by the primary user, which can be received by all cognitive UAV users, v n (t) is the additive white Gaussian noise, and the variance vector of v n (t) is (·) T is the transpose of the matrix, α n represents the channel attenuation, which can be considered a constant throughout the sensing process, x n (t) is the signal received by the UAV user U n at the t-th moment, and H1 represents the occupied spectrum, while H0 represents the unoccupied spectrum.

[0010] The present invention uses energy detection to achieve the sensing of each cognitive UAV user. The received signal is sampled. That is, after T-point sampling is completed within the sampling interval period, the decision statistic of the cognitive UAV user U n is The decision statistic is affected by the control channel noise during the transmission process. At this time, at the fusion center, the decision statistic received by cognitive UAV user U n is where n = 1, 2, …, N, is the noise introduced by the control channel. It is assumed that the channel noise follows a Gaussian distribution with a mean of 0, and the variance vector of the control channel noise is At the fusion center, the global decision statistic is where w n is the weight vector of the decision statistic y n , representing the contribution degree of cognitive UAV user U n to the global perception.

[0011] The fusion center compares the global statistical decision quantity Y with the decision threshold λ. If the global statistical decision quantity Y is greater than the decision threshold λ, the signal of the authorized cognitive UAV user exists; otherwise, the signal does not exist. In the multi-UAV spectrum sensing model, the false alarm probability is The detection probability is where W = [w1, w2, …, w N , α = [|α1| 2 , |α2| 2 , …, |α N | 2 T , A = 2Tdiag 2 (σ) + diag(δ), diag(·) is a diagonal matrix, B = 2Tdiag 2 (σ) + diag(δ) + 4Ediag(α)diag(σ).

[0012] When the false alarm probability is known, the maximum detection probability can be obtained through the spectrum sensing method, and the decision threshold can be expressed by the false alarm probability as At this time, the detection probability is

[0013] According to the detection probability, establish the minimum value optimization objective function of the multi-UAV cooperative spectrum sensing model Since Q is a monotonically decreasing function, finding the minimum value of the objective function is to calculate the maximum value of the detection probability. W = [w1, w2, …, w N is the weight vector of cognitive UAV users, w n is the contribution degree of cognitive UAV user U n , and it satisfies

[0014] Step 2: Initialize the quantum position of the initial quantum tern and set parameters.​

[0015] Set the population size to K1 and the maximum number of iterations to K2. In the initial population, randomly initialize the quantum positions of the quantum sooty terns. The initial quantum position of the i-th quantum sooty tern in the first generation is where S is the maximum dimension of the quantum position vector, and any dimension of all quantum positions is a random number between [0, 1]. The position of the quantum sooty tern can be obtained by mapping the quantum position. The quantum position of the i-th quantum sooty tern in the k-th iteration is

[0016] The position of the i-th quantum sooty tern in the k-th iteration is It can be mapped from its quantum position The mapping rule is:

[0017] Step 3: Calculate the fitness function value of the quantum sooty tern position.

[0018] Correspond the position of the i-th quantum sooty tern in the k-th iteration to the weight vector corresponding to the cognitive UAV user. Let At this time, the fitness function of the quantum sooty tern is It can be known that the fitness function value of the i-th quantum sooty tern in the k-th iteration. After comparing the fitness function values of all quantum sooty terns, find the quantum position

[0019] of the optimal quantum sooty tern in the k-th iteration.

[0020] In the migration strategy, the quantum sooty tern has two behaviors: anti-collision behavior and convergence behavior. The quantum sooty tern updates its own quantum position relying on these two behaviors. In the anti-collision behavior, the h-th dimensional quantum rotation angle of the i-th quantum sooty tern is C(k) is the anti-collision behavior attenuation factor, and Then use the quantum rotation gate to update the h-th dimensional quantum position of the i-th quantum sooty tern in the anti-collision behavior: Map the quantum position to obtain the position Then calculate the fitness function value of

[0021] After the anti-collision behavior of the quantum sooty tern, it will converge to the optimal quantum sooty tern. In the convergence behavior, the h-th dimensional quantum rotation angle of the i-th quantum sooty tern is is a random number between [0, 1]. Then, the quantum rotation gate is used to update the h - dimensional quantum position of the i - th quantum sooty tern in the convergence behavior: The quantum position is mapped to obtain the position Then calculate the fitness function value of

[0022] After obtaining and without collision, the i - th quantum sooty tern continues to converge to the optimal quantum sooty tern. In the continuous convergence behavior, the h - dimensional quantum position of the i - th quantum sooty tern is The quantum position is mapped to obtain the position Then calculate the fitness function value of

[0023] Step 5: Update the quantum position of the quantum sooty tern using the attack strategy.

[0024] When the quantum sooty tern attacks prey, it will generate a spiral behavior in the air, which is a unique attack characteristic of the quantum sooty tern. In the attack strategy, the h - dimensional quantum rotation angle of the i - th quantum sooty tern is is a random number between [0, 2π], and R i is the spiral radius when the i - th quantum sooty tern attacks prey. The quantum rotation gate is used to update the h - dimensional quantum position of the i - th quantum sooty tern in the attack strategy: Then calculate the fitness function value of Assign a value to the quantum position of the i - th quantum sooty tern in the (k + 1)-th iteration. The assignment rule is as follows: Among select the optimal quantum position among the four and assign it to the quantum position

[0025] Step 6: Determine whether the maximum iteration number K2 of the quantum sooty tern is reached. If so, terminate the iteration, map the position of the optimal quantum sooty tern to the weight vector of the cognitive UAV user and output; otherwise, let k = k + 1 and continue to execute Step 4.

[0026] Compared with the prior art, the beneficial effects of the present invention are:

[0027] (1) The currently proposed spectrum sensing methods, such as the matched filter detection, energy detection and other methods, are single-node sensing methods. However, in the case of shadow and deep fading, the sensing results of a single node are unreliable. Therefore, the present invention uses a cooperative spectrum sensing method, which fuses the sensing results of multiple nodes to achieve the global sensing of the multi-UAV cooperative sensing model.

[0028] (2) The present invention uses the sooty tern method, which has two update strategies, the migration strategy and the attack strategy. Each quantum sooty tern can generate four quantum positions according to these two strategies. These four quantum positions are generated by different update methods, which greatly improves the optimization rate compared with the existing methods.

[0029] (3) The present invention designs a quantum-encoded sooty tern quantum position evolution mechanism to obtain a new sooty tern mechanism method, which overcomes the drawback that the previous algorithms are prone to fall into local convergence and has better convergence performance compared with the existing methods. And a mapping strategy between quantum positions is designed. This mapping strategy makes the position of each sooty tern correspond to a weight vector of a cognitive UAV user, and satisfies the condition that the sum of the elements in the weight vector is 1. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 Schematic diagram of the multi-UAV cooperative spectrum sensing method based on the sooty tern mechanism designed by the present invention;

[0031] Figure 2 Curve showing the relationship between the objective function value and the number of iterations when there are 15 cognitive UAV users;

[0032] Figure 3 Curve showing the relationship between the detection probability and the number of iterations when there are 15 cognitive UAV users;

[0033] Figure 4 Curve showing the relationship between the objective function value and the false alarm probability when there are 15 cognitive UAV users;

[0034] Figure 5 Curve showing the relationship between the detection probability and the false alarm probability when there are 15 cognitive UAV users;

[0035] Figure 6 Curve showing the relationship between the objective function value and the number of iterations when there are 50 cognitive UAV users;

[0036] Figure 7 Curve showing the relationship between the detection probability and the number of iterations when there are 50 cognitive UAV users;

[0037] Figure 8 Curve showing the relationship between the objective function value and the false alarm probability when there are 50 cognitive UAV users;

[0038] Figure 9 It represents the relationship curve between the detection probability and the false alarm probability when there are 50 cognitive UAV users. Specific implementation manners

[0039] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0040] Step 1: Establish a multi-UAV collaborative spectrum sensing model.

[0041] Assume that there are a total of N cognitive UAV users, and the set of cognitive UAV users is U = {U1, U2,..., U N}. The number of samples of each cognitive UAV user is T, and the signals sampled from N cognitive UAV users can form an N×T-dimensional vector matrix X N×T . At the t-th moment, the binary hypothesis testing model of the cognitive UAV user U n can be expressed as where n = 1, 2,..., N, t = 0, 1,..., T - 1, s(t) is the signal transmitted by the primary user, which can be received by all cognitive UAV users, v n (t) is the additive white Gaussian noise, and the variance vector of v n (t) is (·) T is the transpose of the matrix, α n represents the channel attenuation, which can be considered as a constant during the entire sensing process, x n (t) is the signal received by the UAV user U n at the t-th moment, and H1 represents the occupied spectrum, and H0 represents the unoccupied spectrum.

[0042] The present invention uses energy detection to realize the sensing of each cognitive UAV user. The received signal is sampled. That is, after T-point sampling is completed within the sampling interval period, the decision statistic of the cognitive UAV user U n is The decision statistic will be affected by the control channel noise during the transmission process. At this time, the decision statistic received by the fusion center from the cognitive UAV user U n is where n = 1, 2,..., N, is the noise introduced by the control channel. Assume that the channel noise follows a Gaussian distribution with a mean of 0, and the variance vector of the control channel noise is In the fusion center, the global decision statistic is where w n is the weight vector of the decision statistic y n , indicating the contribution degree of the cognitive UAV user U n to the global sensing.

[0043] The fusion center compares the global statistical decision quantity Y with the decision threshold λ. If the global statistical decision quantity Y is greater than the decision threshold λ, the signal of the authorized cognitive UAV user exists; otherwise, the signal does not exist. In the multi-UAV spectrum sensing model, the false alarm probability is The detection probability is where W = [w1, w2, …, w N , α = [|α1| 2 , |α2| 2 , …, |α N | 2 T , A = 2Tdiag 2 (σ) + diag(δ), diag(·) is a diagonal matrix, B = 2Tdiag 2 (σ) + diag(δ) + 4Ediag(α)diag(σ).

[0044] When the false alarm probability is known, the maximum detection probability can be obtained through the spectrum sensing method, and the decision threshold can be expressed by the false alarm probability as At this time, the detection probability is

[0045] Establish the minimum value optimization objective function of the multi-UAV cooperative spectrum sensing model according to the detection probability Since Q is a monotonically decreasing function, finding the minimum value of the objective function is to calculate the maximum value of the detection probability. W = [w1, w2, …, w N is the weight vector of the cognitive UAV user, w n is the contribution degree of the cognitive UAV user U n , and satisfies

[0046] Step 2: Initialize the quantum position of the quantum sooty tern and set parameters.

[0047] Set the population size as K1 and the maximum number of iterations as K2. In the initial population, randomly initialize the quantum position of the quantum sooty tern. The initial quantum position of the i-th quantum sooty tern in the first generation is where S is the maximum dimension of the quantum position vector, and any dimension of all quantum positions is a random number between [0, 1]. The position of the quantum sooty tern can be obtained through the mapping of the quantum position. The quantum position of the i-th quantum sooty tern in the k-th iteration is

[0048] The position of the i-th quantum sooty tern in the k-th iteration is ​ can be obtained by its quantum position through mapping, and the mapping rule is:

[0049] Step 3: Calculate the fitness function value of the quantum sooty tern position.

[0050] Correspond the position of the i-th quantum sooty tern in the k-th iteration to the weight vector of the cognitive UAV user Let At this time, the fitness function of the quantum sooty tern is It can be known that the fitness function value of the i-th quantum sooty tern in the k-th iteration After comparing the fitness function values of all quantum sooty terns, find the quantum position of the optimal quantum sooty tern in the k-th iteration

[0051] Step 4: Update the quantum position of the quantum sooty tern using the migration strategy.

[0052] In the migration strategy of the quantum sooty tern, there are two behaviors in total: anti-collision behavior and convergence behavior. The quantum sooty tern updates its own quantum position relying on these two behaviors. In the anti-collision behavior, the h-th dimensional quantum rotation angle of the i-th quantum sooty tern is C(k) is the anti-collision behavior attenuation factor, and Then use the quantum rotation gate to update the h-th dimensional quantum position of the i-th quantum sooty tern in the anti-collision behavior: Map the quantum position to obtain the position Then calculate the fitness function value of

[0053] After the anti-collision behavior of the quantum sooty tern, it will converge to the optimal quantum sooty tern. In the convergence behavior, the h-th dimensional quantum rotation angle of the i-th quantum sooty tern is is a random number between [0, 1]. Then use the quantum rotation gate to update the h-th dimensional quantum position of the i-th quantum sooty tern in the convergence behavior: Map the quantum position to obtain the position Then calculate the fitness function value of

[0054] After obtaining and and without collision, the i-th quantum sooty tern continues to converge to the optimal quantum sooty tern, and the h-th dimensional quantum position of the i-th quantum sooty tern in the continuous convergence behavior is Map the quantum positions to obtain positions Then calculate the fitness function value of

[0055] Step 5: Update the quantum positions of the quantum sooty terns using the attack strategy.

[0056] When the quantum sooty tern attacks prey, it will generate a spiral behavior in the air, which is a unique attack characteristic of the quantum sooty tern. In the attack behavior, the h - dimensional quantum rotation angle of the i - th quantum sooty tern is a random number between [0, 2π], and R i is the spiral radius when the i - th quantum sooty tern attacks prey. Update the h - dimensional quantum position of the i - th quantum sooty tern in the attack behavior using the quantum rotation gate: Then calculate the fitness function value of Assign a value to the quantum position of the i - th quantum sooty tern in the (k + 1)-th iteration The assignment rule is as follows: Among select the optimal quantum position among the four and assign it to the quantum position of the i - th quantum sooty tern in the (k + 1)-th iteration

[0057] Step 6: Determine whether the maximum iteration number K2 of the quantum sooty tern is reached. If so, terminate the iteration, map the position of the optimal quantum sooty tern to the weight vector of the cognitive UAV user and output it; otherwise, let k = k + 1 and continue to execute Step 4.

[0058] Denote the quantum sooty tern optimization method as QSTO and the particle swarm optimization method as PSO. To verify the performance of the multi - UAV cooperative spectrum sensing method based on the quantum sooty tern mechanism, two groups of simulation tests were carried out in this invention. Set the population size as K1 = 100, the iteration number as K2 = 200, and the simulation results take the average value of 200 independent repeated experiments. The learning factor sizes of the PSO method are both 2, and the search radii of all quantum sooty terns in the QSTO method are all R i= 2, there are 15 cognitive drone users in the first group of experiments, and the values of other relevant parameters are: N = 15, s(t) = 1 is satisfied at any time, the sampling point T = 20, σ = [2, 2.5, 0.9, 2.7, 1.3, 3.3, 2, 2.5, 0.9, 2.7, 2, 2.5, 0.9, 2.7, 1.3], δ = [1.3, 0.8, 2, 3.8, 2.3, 0.4, 1.3, 0.8, 2, 3.1, 1.3, 0.8, 2, 3.1, 1.3], α = [0.4, 0.5, 0.7, 0.3, 0.4, 0.3, 0.6, 0.5, 0.2, 0.3, 0.4, 0.5, 0.7, 0.3, 0.4], and the false alarm probability is P = 0.08. The relationship between the objective function value, the detection probability, and the number of iterations is as Figure 2 , 3 shown. The simulation results show that the convergence speed and convergence accuracy of the QSTO method designed by the present invention are significantly better than those of the PSO method.

[0059] To illustrate the relationship between the detection probability and the false alarm probability, the relationship between the objective function value, the detection probability, and the false alarm probability when there are 15 cognitive drone users is obtained by simulation as Figure 4 , 5 shown. Compared with the existing spectrum sensing methods, the performance of the multi-drone cooperative spectrum sensing method based on the quantum sooty tern mechanism causes less interference to the primary users. And in solving the problem of spectrum sensing, the performance of the QSTO method is better than that of the PSO method.

[0060] In the second group of experiments, there are 50 cognitive UAV users, and the values of other relevant parameters are: N = 50, s(t) = 1 is satisfied at any time, the sampling point T = 20, σ = [2, 2.5, 0.9, 2.7, 1.3, 3.3, 2.0, 2.5, 0.9, 2.7, 2, 2.5, 0.9, 2.7, 1.3, 3.8, 1.2, 2.5, 0.7, 1.5, 1.9, 2.1, 2.4, 3.3, 0.4, 0.6, 1.2, 1.4, 1.7, 0.8, 2.5, 2.7, 2.9, 0.3, 1.4, 3.1, 2.0, 1.8, 1.0, 0.4, 1.0, 2.0, 0.7, 1.5, 1.3, 2.4, 1.0, 0.5, 0.8, 2.0], δ = [1.3, 0.8, 2.0, 3.8, 2.3, 0.4, 1.3, 0.8, 2.0, 3.1, 1.3, 0.8, 2, 3.1, 1.3, 0.3, 1.5, 1.8, 0.7, 0.9, 2.2, 2.4, 2.0, 1.0, 0.5, 2.0, 2.1, 2.3, 1.6, 1.2, 1.8, 0.5, 0.9, 2.3, 1.5, 0.7, 1.3, 0.4, 2.1, 0.9, 1.0, 2.0, 2.4, 2.9, 1.2, 1.1, 2.6, 0.9, 1.2, 0.8], α = [0.4, 0.5, 0.7, 0.3, 0.4, 0.3, 0.6, 0.5, 0.2, 0.3, 0.4, 0.5, 0.7, 0.3, 0.4, 0.5, 0.8, 0.7, 0.5, 0.9, 0.4, 0.7, 0.2, 0.6, 0.7, 0.4, 0.6, 0.6, 0.7, 0.4, 0.6, 0.7, 0.5, 0.9, 0.4, 0.7, 0.3, 0.8, 0.2, 0.5, 0.9, 0.4, 0.7, 0.7, 0.4, 0.6, 0.7, 0.5, 0.7, 0.6], and the false alarm probability is P = 0.15. The relationship between the objective function value, the detection probability, and the number of iterations is as shown in Figure 6 , 7 shown. The simulation results show that after the number of cognitive UAV users increases, the convergence speed and convergence accuracy of the QSTO method designed by the present invention are still better than those of the PSO method, and the gap is more obvious.

[0061] To further illustrate the relationship between the detection probability and the false alarm probability, the relationship between the objective function value, the detection probability, and the false alarm probability when there are 50 cognitive UAV users is obtained by simulation as shown in Figure 8 , 9 shown. It can be seen that in solving the spectrum sensing problem, the PSO method is prone to falling into local optimum and has poor performance. Compared with the PSO method, the detection probability of the QSTO method can reach more than 0.999 for all 6 false alarm probabilities, which shows that the QSTO method has excellent performance in solving the spectrum sensing problem.

Claims

1. A multi-UAV collaborative spectrum sensing method based on the quantum sooty tern mechanism, characterized in that The steps are as follows: Step 1, establish a multi-UAV collaborative spectrum sensing model; There are N cognitive UAV users, and the set of cognitive UAV users is U = {U1, U2, …, U N}, the number of samples for each cognitive UAV user is T, and the signals sampled from N cognitive UAV users form an N×T-dimensional vector matrix X N×T , at the t-th moment, the binary hypothesis testing model of the cognitive UAV user U n is as follows: Where n = 1, 2…, N, t = 0, 1,… T-1, s(t) is the signal transmitted by the primary user, and all cognitive drone users can receive the signal, v n (t) is the additive Gaussian white noise, and v n The variance vector of (t) is α n Indicates channel attenuation. In the entire perception process, the channel attenuation can be considered as a constant. n (t) is the number of drone users U at the tth moment n The received signal, where H1 indicates occupied spectrum and H0 indicates unoccupied spectrum; Cognitive UAV user U n The decision statistic is At the fusion center, the decision statistic received from cognitive UAV user U n is The noise introduced to the control channel, where the channel noise follows a Gaussian distribution with a mean of 0, and the variance vector of the control channel noise is At the fusion center, the global decision statistic is w n is the weight vector of the decision statistic y n , representing the contribution degree of cognitive UAV user U n to the global perception; the fusion center compares the global decision statistic Y with the decision threshold λ; if the global decision statistic Y is greater than the decision threshold λ, the signal of the authorized cognitive UAV user exists; otherwise, the signal does not exist; in the multi-UAV spectrum sensing model, the false alarm probability is The detection probability is Among them, W = [w1, w2, …, w N , α = [|α1| 2 , |α2| 2 , …, |α N | 2 T , A = 2Tdiag 2 (σ) + diag(δ), diag(·) is a diagonal matrix, B = 2Tdiag 2 (σ) + diag(δ) + 4Ediag(α)diag(σ);​ Step 2, initialize the quantum positions of quantum sooty terns and set parameters; Step 3, calculate the fitness function values of the quantum positions of quantum sooty terns; The position of the $i$-th quantum sooty tern in the $k$-th iteration corresponds to the weight vector $W$ of the cognitive UAV user i k , let the fitness function value of the position of the $i$-th quantum sooty tern be: Step 4, update the quantum positions of quantum sooty terns using the migration strategy; Step 5, update the quantum positions of quantum sooty terns using the attack strategy; Step 6, determine whether the maximum iteration number K2 of the quantum sooty terns is reached. If so, terminate the iteration, map the position of the optimal quantum sooty tern to the weight vector of the cognitive UAV user and output it; otherwise, let k = k + 1 and continue to execute Step 4.

2. The multi-UAV collaborative spectrum sensing method based on the quantum sooty tern mechanism according to claim 1, wherein Step 2 specifically includes: setting the population size as K1 and the maximum number of iterations as K2; in the initial population, randomly initialize the quantum positions of the quantum sooty terns, and the initial quantum position of the i-th quantum sooty tern in the first generation is where S is the maximum dimension of the quantum position vector, and any dimension of all quantum positions is a random number between [0, 1]. The position of the quantum sooty tern can be obtained through the mapping of the quantum position; the quantum position of the i-th quantum sooty tern in the k-th iteration is The position of the i-th quantum sooty tern in the k-th iteration is can be obtained from its quantum position through mapping, and the mapping rule is:

3. The multi-UAV collaborative spectrum sensing method based on the quantum sooty tern mechanism according to claim 2, wherein After comparing the fitness function values of all quantum black terns in step 3, find the quantum position of the optimal quantum black tern in the kth iteration 4. The multi-UAV collaborative spectrum sensing method based on the quantum sooty tern mechanism according to claim 3, characterized in that, Step 4 specifically includes: in the anti-collision behavior, the h-th dimensional quantum rotation angle of the i-th quantum sooty tern is C(k) is the anti-collision behavior attenuation factor, and Update the h-th dimensional quantum position of the i-th quantum sooty tern in the anti-collision behavior by using the quantum rotation gate: Map the quantum position to obtain the position Then calculate the fitness function value of After the anti-collision behavior, the quantum sooty tern converges to the optimal quantum sooty tern. During the convergence behavior, the quantum rotation angle of the $h$-th dimension of the $i$-th quantum sooty tern is a random number between [0, 1]; update the quantum position of the $h$-th dimension of the $i$-th quantum sooty tern during the convergence behavior using the quantum rotation gate: Map the quantum position to obtain the position Then calculate the fitness function value of In getting and After that, the i-th quantum sooty tern continues to converge to the optimal quantum sooty tern under the premise of no collision. In the continued convergence behavior, the h-dimensional quantum position of the i-th quantum sooty tern is h=1,2,…,S, the quantum position Mapping to get the location Then calculate The fitness function value of 5. The multi-UAV collaborative spectrum sensing method based on the quantum sooty tern mechanism according to claim 4, characterized in that, Step 5 specifically includes: in the attack strategy, the h-dimensional quantum rotation angle of the i-th quantum sooty tern is is a random number between [0,2π], R i is the spiral radius of the i-th quantum sooty tern when attacking its prey; the h-th dimensional quantum position of the i-th quantum sooty tern in the attack strategy is updated using the quantum revolving gate: Then calculate The fitness function value of The quantum position of the i-th quantum sooty tern in the k+1th iteration is Assign values, and the assignment rules are as follows: Among them, the best quantum position is selected and assigned to the quantum position of the i-th quantum sooty tern in the k+1th iteration.

Images