Unmanned aerial vehicle monitoring method for spectrum aliasing wireless signal identification

Abstract

The application discloses a UAV monitoring method for spectrum aliasing wireless signal identification, and comprises the following steps: performing short-time Fourier transform on a spectrum mixed signal received by a monitoring node in a certain spectrum sampling time slot to obtain a spectrum signal; performing category identification through non-negative matrix decomposition, constructing a target function Q1 and solving, to obtain an estimated value of an unknown signal source number; splicing a coefficient matrix containing category information to be solved under different visual angles with a signal fading template matrix of the monitoring node, constructing a target function Q2 and solving, to obtain a signal category and a transmission power indication matrix representing the position and transmission power of the unknown signal. The application solves three tasks of signal identification, signal position estimation and transmission power estimation through a deep identification method; meanwhile, the UAV is moved to the vicinity of the centroid to obtain better monitoring effect.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of deep recognition of spectrum mixed signals, and particularly relates to a UAV monitoring method for spectrum mixed wireless signal recognition. BACKGROUND

[0002] The application of unmanned aerial vehicles (UAVs) in the field of monitoring is rapidly developing, providing innovative solutions for real-time monitoring, data collection, and extensive intelligent analysis. In the context of wireless signal monitoring, deploying UAVs equipped with advanced spectrum analyzers can dynamically analyze and evaluate the radio frequency spectrum within a specific area. These UAVs not only capture precise data in complex environments but also assist in spectrum management, interference detection, and overall communication security evaluation. Furthermore, the use of UAVs for spectrum monitoring offers high flexibility, extensive coverage, and rapid response, as UAVs can quickly respond to sudden signal changes and adjust their monitoring positions, ensuring the acquisition of critical information in the shortest time. This makes UAVs play a crucial role in ensuring the security and stability of communication networks, especially in the face of increasingly complex electromagnetic environments. UAVs, with their maneuverability and extensive coverage, can provide strong support for the optimal allocation of spectrum resources and the timely detection of communication interference.

[0003] In a radio monitoring system, traditional static monitoring nodes have some inherent limitations due to their fixed positions. The monitoring range of these nodes is often affected by their physical location and the surrounding environment, resulting in limited signal reception coverage, which affects the overall monitoring efficiency and accuracy. In addition, static nodes cannot flexibly respond to dynamic changes in signal sources, especially in complex urban environments or areas with large terrain changes, where their monitoring effectiveness is more easily hindered. The introduction of UAVs brings significant advantages to the radio monitoring system. With their flexible flight capabilities and high maneuverability, UAVs can break through the limitations of traditional static nodes, enabling dynamic monitoring of extensive areas and significantly improving signal reception coverage and monitoring accuracy. However, in the case of blind source signals, due to the lack of prior position information of the signal source, UAVs still face challenges in practical applications. How to optimize the flight strategy of UAVs in the absence of signal source location information to maximize their efficiency in the radio monitoring system is an important topic in current research. This not only involves flight path planning but also involves the application of various signal processing and optimization algorithms to ensure that UAVs can effectively respond to monitoring needs in complex electromagnetic environments.

[0004] In the field of UAV monitoring, the prior art [1] detects and identifies other UAVs by equipping the UAV with a radio frequency receiver and transmitting radio frequency signals. In the prior art [2], wireless data is transmitted in real time by a UAV swarm to achieve automatic modulation recognition in V2X (vehicle-to-everything) radio monitoring, optimize the flight trajectory of the UAV, and obtain better network performance and monitoring effect. In the prior art [3], resource allocation and data collection of a UAV-assisted wireless sensor network can effectively monitor the state of the wireless sensor network.

[0005] In the existing UAV-assisted monitoring system (such as prior arts [1] to [3]), the main focus is on optimizing the flight trajectory of the UAV to obtain better network performance and less resource consumption, without considering how to better identify the radio signals in the monitoring area. The identification of spectrum aliasing signals currently focuses on the identification of signal types, and the position and transmission power characteristics of unknown signals are not fully identified.

[0006] Prior art [1]: Y. Xie, P. Jiang, Y. Gu and X. Xiao, “Dual-Source Detection and Identification System Based on UAV Radio Frequency Signal,” IEEE Transactions on Instrumentation and Measurement, vol. 70, pp. 1-15, 2021.

[0007] Prior art [2]: Q. Zhou, S. Wu, C. Jiang, R. Zhang and X. Jing, “Over-the-Air Federated Transfer Learning Over UAV Swarm for Automatic Modulation Recognition in V2X Radio Monitoring,” IEEE Transactions on Vehicular Technology, vol. 73, no. 3, pp. 3597-3607, March 2024.

[0008] Prior art [3]: T.D.P. Perera, S. Panic, D.N.K. Jayakody, P. Muthuchidambaranathan, and J. Li, "A WPT-enabled UAV-assisted condition monitoring scheme for wireless sensor networks," IEEE Transactions on Intelligent Transportation Systems, vol. 22, no. 8, pp. 5112-5126, 2020. SUMMARY

[0009] The purpose of the present application is to overcome the shortcomings and deficiencies of the prior art, and to provide a UAV monitoring method for spectrum-mixed wireless signal identification, which utilizes the flexibility of UAVs to monitor the radio, deeply identifies the types of spectrum-mixed wireless signals, estimates the positions and transmission powers of the signals, and adjusts the positions of UAVs to effectively improve the identification effect of spectrum-mixed signals through UAV-assisted radio monitoring.

[0010] The purpose of the present application is achieved by the following technical solutions:

[0011] The UAV monitoring method for spectrum-mixed wireless signal identification comprises the following steps:

[0012] S1, performing short-time Fourier transform on the spectrum mixed signals received by the monitoring nodes in a certain spectrum sampling time slot to obtain spectrum signals; wherein, one UAV serves as a monitoring node;

[0013] S2, performing type identification on the spectrum signals through non-negative matrix decomposition to construct a target function Q1;

[0014] S3, solving the target function Q1, respectively deriving the update rule of the coefficient matrix containing type information to be solved and the signal type consensus matrix, and finally obtaining the estimated value of the number of unknown signal sources;

[0015] S4, splicing the coefficient matrix containing type information to be solved under different perspectives and the signal fading template matrix of the monitoring nodes to construct a target function Q2;

[0016] S5, solving the target function Q2 through proximal alternating linearization minimization algorithm to obtain the update rule of the signal type and transmission power indication matrix to be solved, and solving to obtain the signal type and transmission power indication matrix representing the positions and transmission powers of the unknown signals.

[0017] In step S1, the channel gain coefficient of the spectrum mixed signal received by the mth monitoring node in the tth spectrum sampling time slot is defined as follows: m (t) is:

[0018]

[0019] wherein, J m is the number of frequency point samples in the frequency band for the mth monitoring node; is the standard power spectral density function of the nth0 unknown signal at the J m frequency points, and B s is the standard spectrum dictionary of the identifiable signal by the edge node; is the transmission power of the nth0 unknown signal source in time slot t, is the channel gain coefficient of the nth0 unknown signal source to the mth monitoring node, z m (t) is the noise at the mth monitoring node and obeys the Gaussian distribution of .

[0020] In step S1, the channel gain coefficient of the spectrum mixed signal is defined as follows:

[0021] Considering the large-scale fading of wireless propagation in an outdoor rich scattering environment, the channel gain coefficient is defined as:

[0022]

[0023] wherein, the channel constant f is the working center frequency of the unknown signal, c is the speed of light, E r,m and are the antenna gains of the mth monitoring node and the nth0 unknown signal, respectively; and are the channel response and distance between the mth monitoring node and the nth0 unknown signal, respectively;

[0024] and have:

[0025]

[0026] wherein, and (x m , y m , h m ) are the coordinates of the nth0 unknown signal and the mth unmanned aerial vehicle in the coordinate system with the edge node as the origin, respectively; the maximum horizontal coordinate and the maximum vertical coordinate of the signal monitoring range of the monitoring network are ±x max , ±y max , respectively. In T s he gth grid to the mth monitoring node is kept unchanged; the link loss of the gth grid to the mth monitoring node d g,m is the distance from the gth grid to the mth monitoring node, and the signal fading dictionary in the area is defined W=[W1,W2…W M ] G×M , and W m =[q 1,m ,q 2,m …q G,m ], wherein W m is the signal fading template matrix of the mth monitoring node.

[0027] The M monitored spectrum mixed signals s m (t) are subjected to short-time Fourier transform to obtain spectrum signals R m , species identification is performed through non-negative matrix decomposition, a target function Q1 is constructed, and Q1 is solved;

[0028]

[0029] s.t.A m ≥0,A * ≥0;

[0030] wherein R m is the spectrum mixed signal monitored by the mth UAV, B s is a standard spectrum dictionary of signals identifiable by the edge node, A m is a coefficient matrix to be solved containing species information, is a signal species determination matrix, which indicates whether it is the signal of the species by 0 and 1, A * is a signal species consensus matrix;

[0031] Since the problem Q1 is convex in the alternately solved A m and A * , the update rules are obtained by derivation of A m and A * :

[0032]

[0033]

[0034] The estimated value of the unknown signal source number is obtained as follows:

[0035]

[0036] wherein ​It is matrix A * The nth element.

[0037] A from different perspectives m With W m By splicing the images together, the connections between different perspectives are strengthened, A = [A1, A2…A…]. M ] N×M W = [W1, W2, ... W M ] G×M Construct the objective function Q2:

[0038]

[0039] stU≥0,∥(U) n ∥1=∥(U) n ∥2;

[0040] Where U is the signal type and transmit power indication matrix to be solved;

[0041] Since Q2 satisfies the KL condition, the update rule for U is obtained by solving the problem using the proximal alternating linearization minimization algorithm:

[0042]

[0043] The obtained U matrix also represents the position and transmission power of the unknown signal. The nth row of U represents the position and transmission power of the nth signal. The index of the largest element in this row is the grid number where the nth signal is located, and the largest element is the transmission power of the nth signal.

[0044] The flight strategy of the drone is as follows:

[0045] First, based on the grid number of the estimated unknown signal... Calculate the corresponding two-dimensional coordinates The two-dimensional coordinates of grid number g are represented as follows:

[0046]

[0047] Where X max With Y max The farthest horizontal and vertical coordinates of the monitoring range after gridding, where g is the number of grids and d0 is the grid spacing; the centroid coordinates (x, y) of N0 unknown signals are calculated using their two-dimensional coordinates. c ,y c ):

[0048]

[0049]

[0050] Construct a (x)c , y c ) is the center, and the radius is r c , the circle is divided into M equal parts, and the mth equal point is (x m , y m ):

[0051]

[0052] Each UAV moves to the nearest equal point, and if there is a UAV at the equal point, it moves to the next nearest equal point; for M UAVs, the distance matrix of the UAV to the M equal points is calculated, wherein the distance matrix of the mth UAV is D m = [d m,1 , d m,2 …d m,M ];

[0053] If there is no UAV at the m' position, the mth UAV moves to the m' position, wherein m' is:

[0054]

[0055] If there is a UAV at the m' position, d m′,M is deleted from D m ;

[0056] After the M UAVs move to the corresponding equal points, the grid number of the unknown signal is monitored and estimated

[0057] The moving method is repeatedly executed until the estimation result is close to stable.

[0058] Meanwhile, the present application provides:

[0059] A server, comprising a processor and a memory, wherein the memory stores at least one program, and the program is loaded and executed by the processor to realize the above-mentioned UAV-based spectrum superimposed wireless signal identification method.

[0060] A computer readable storage medium, wherein the storage medium stores at least one program, and the program is loaded and executed by a processor to realize the above-mentioned UAV-based spectrum superimposed wireless signal identification method.

[0061] Compared with the prior art, the present application has the following advantages and beneficial effects:

[0062] 1. The present application solves the three tasks of signal identification, signal position estimation and transmission power estimation through a deep identification method.

[0063] 2. The application considers the distribution of unknown signals when formulating a flight method, and controls the movement of the UAV through the centroid. Since the centroid of the unknown signal position represents the distribution of uniform unknown signals around the position, moving the UAV to the vicinity of the centroid can achieve better monitoring effect. BRIEF DESCRIPTION OF DRAWINGS

[0064] Figure 1 A monitoring architecture diagram for the unmanned aerial vehicle monitoring method for spectrum mixed wireless signal identification.

[0065] Figure 2 A comparison chart of the effects of different methods on signal type identification tasks.

[0066] Figure 3 A comparison chart of the effects of different methods on signal position estimation tasks.

[0067] Figure 4 A chart showing the running effect of the unmanned aerial vehicle flight strategy on the signal type identification task.

[0068] Figure 5 A chart showing the running effect of the unmanned aerial vehicle flight strategy on the signal position estimation task.

[0069] Figure 6 A chart showing the running effect of the unmanned aerial vehicle flight strategy on the signal transmission power estimation task. DETAILED DESCRIPTION

[0070] The application will be described in further detail below with reference to the embodiments and drawings, but the embodiments of the application are not limited thereto.

[0071] As Figure 1 , the unmanned aerial vehicle monitoring method for spectrum mixed wireless signal identification includes the following steps:

[0072] S1, performing short-time Fourier transform on the spectrum mixed signals received by the monitoring nodes in a certain spectrum sampling time slot to obtain spectrum signals; wherein an unmanned aerial vehicle serves as a monitoring node;

[0073] S2, performing type identification on the spectrum signals through non-negative matrix factorization to construct an objective function Q1;

[0074] S3, solving the objective function Q1, respectively deriving the update rule for the coefficient matrix containing type information to be solved and the signal type consensus matrix, and finally obtaining the estimated value of the number of unknown signal sources;

[0075] S4, splicing the coefficient matrix containing type information to be solved under different perspectives and the signal fading template matrix of the monitoring nodes to construct an objective function Q2;

[0076] S5, solving the target function Q2 through a proximal alternating linearized minimization algorithm to obtain an update rule of a signal type and a transmit power indication matrix to be solved, and solving the signal type and the transmit power indication matrix to represent positions and transmit powers of unknown signals.

[0077] Specifically,

[0078] The unmanned aerial vehicle monitoring method for spectrum mixed wireless signal identification comprises the following steps:

[0079] Suppose that m unmanned aerial vehicles as m monitoring nodes are uniformly distributed in a communication range, and spectrum data generated by spectrum scanning of each monitoring node is returned to an edge computing node through a wireless manner;

[0080] A detection duration of each monitoring node for a frequency band is T s , and a spectrum data return time of the frequency band is T d , so that a total duration for completing a frequency band monitoring task is T0=T s +T d .

[0081] Suppose that there are N0 unknown wireless signal sources in the same frequency band in the monitoring area, and the monitoring nodes have no prior information about signal types, positions and transmit powers of the unknown wireless signal sources;

[0082] A spectrum mixed signal s m (t) received by the mth monitoring node in the tth spectrum sampling time slot is:

[0083]

[0084] wherein, J m is a frequency point sampling number of the mth monitoring node in the frequency band; is a standard power spectral density function of the nth0 unknown signal on the J m frequency points, and B s is a standard spectrum dictionary of signals identifiable by the edge node; is a transmit power of the nth0 unknown signal source in the time slot t, is a channel gain coefficient of the nth0 unknown signal source to the mth monitoring node, z m (t) is noise at the mth monitoring node and obeys a Gaussian distribution of .

[0085] Considering large-scale fading of wireless propagation in an outdoor rich scattering environment, the channel gain coefficient is defined as:

[0086]

[0087] Wherein, the channel constant f is the operating center frequency of the unknown signal, c is the speed of light, E r,m And The antenna gain of the mth monitoring node and the nth0 unknown signal respectively; And The channel response and distance between the mth monitoring node and the nth0 unknown signal respectively;

[0088] And:

[0089]

[0090] Wherein, And (x m ,y m ,h m ) are the coordinates of the nth0 unknown signal and the mth UAV in the coordinate system with the edge node as the origin; the maximum horizontal coordinate and the maximum vertical coordinate of the signal monitoring range of the monitoring network are ±x max ,±y max ; In T s , it remains unchanged; the link loss of the gth grid to the mth monitoring node d g,m The distance from the gth grid to the mth monitoring node, define the signal fading dictionary in the area W=[W1,W2…W M ] G×M , and W m =[q 1,m ,q 2,m ……q G,m ], wherein W m Is the signal fading template matrix of the mth monitoring node.

[0091] The present application proposes joint multi-layer NMF recognition (JMNR) for deep recognition of spectrum aliasing signals, and the M spectrum aliasing signals s m (t) monitored by the monitoring network is transformed into a frequency spectrum signal R m Through short-time Fourier transform, and the species is identified through non-negative matrix factorization, a target function Q1 is constructed, and Q1 is solved.

[0092]

[0093] Wherein R m Is the spectrum aliasing signal monitored by the mth UAV, B s Is the standard spectrum dictionary of the identifiable signal of the edge node, and Am is the coefficient matrix to be solved containing the category information, is the signal category determination matrix, indicating whether it is the signal of this category by 0 and 1, A * is the signal category consensus matrix.

[0094] Since the problem Q1 is solved alternately A m and A * is convex, respectively A m and A * derivation to get the update rule:

[0095]

[0096]

[0097] get the estimate of the number of unknown signal sources is:

[0098]

[0099] where is the nth element in the matrix A * .

[0100] Splice A m and W m under different perspectives to strengthen the connection under each perspective, A=[A1,A2…A M ] N×M , W=[W1,W2…W M ] G×M . Construct the objective function Q2:

[0101]

[0102] s.t.U≥0,∥(U) n ∥1=∥(U) n ∥2;

[0103] where U is the signal category and transmission power indication matrix to be solved.

[0104] Since Q2 satisfies the KL (Kurdyka-Lojasiewicz) condition, it is solved by the proximal alternating linearization minimization (PLAM) algorithm, and the update rule of U is obtained:

[0105]

[0106] The obtained U matrix also represents the positions and transmitting powers of the unknown signals, the nth row of U represents the position and transmitting power of the nth signal, wherein the subscript corresponding to the maximum element of the row is the grid number where the nth signal is located, and the maximum element is the transmitting power of the nth signal.

[0107] The application also designs a flight method of the unmanned aerial vehicle, which can obtain better monitoring effect by adjusting the position of the unmanned aerial vehicle, first, according to the estimated grid number of the unknown signal The corresponding two-dimensional coordinates are calculated The two-dimensional coordinates of the grid number g are represented as:

[0108]

[0109] Wherein X max and Y max are the farthest horizontal and vertical coordinates of the monitoring range after gridding, g is the number of grids, and d0 is the grid spacing; the centroid coordinates (x c ,y c ) of the N0 unknown signals are calculated as follows:

[0110]

[0111] A circle with (x c ,y c ) as the center and r c as the radius is constructed, the mth division point of the circle is (x m ,y m ):

[0112]

[0113] Each unmanned aerial vehicle moves to the nearest division point, if there is an unmanned aerial vehicle at the division point, then moves to the next nearest division point; for M unmanned aerial vehicles, the distance matrix of the unmanned aerial vehicle to the M division points is calculated, wherein the distance matrix of the mth unmanned aerial vehicle is D m =[d m,1 ,d m,2 …d m,M ];

[0114] If there is no unmanned aerial vehicle at the m' position, the mth unmanned aerial vehicle moves to the m' position, wherein m' is:

[0115]

[0116] If there is an unmanned aerial vehicle at the m' position, d m′,M is deleted from D m ;

[0117] The M-frames UAVs all move to the corresponding good equidistant points, monitor the signals and estimate the grid number of the unknown signals

[0118] The moving method is repeatedly executed until the estimation result approaches stability.

[0119] Meanwhile, the application provides:

[0120] A server, comprising a processor and a memory, wherein at least one program is stored in the memory, and the program is loaded and executed by the processor to implement the above-mentioned UAV-based spectrum superimposed wireless signal identification method.

[0121] A computer readable storage medium, wherein at least one program is stored in the storage medium, and the program is loaded and executed by a processor to implement the above-mentioned UAV-based spectrum superimposed wireless signal identification method.

[0122] The experimental results show that the flight method can effectively improve the signal type identification, position estimation and transmission power estimation of the spectrum superimposed signal.

[0123] As Figure 2 In the signal type identification task, the method proposed by the application is superior to the traditional blind source separation method.

[0124] As Figure 3 In the signal position estimation task, the method proposed by the application is superior to the traditional trilateration algorithm and PSO method.

[0125] As Figure 4 With the increase of the number of UAV flight strategy operation, the signal type identification effect of the method proposed by the application is gradually improved, and when the monitoring range D ＝ 140 m , the signal type identification effect is improved by 5.2%.

[0126] As Figure 5 With the increase of the number of UAV flight strategy operation, the signal position estimation result of the method proposed by the application is comprehensively improved, and when the monitoring range D=60m, the signal position estimation result is improved by 12.3%; when the monitoring range D=140m, the signal position estimation result is improved by 20.9%;

[0127] As Figure 6 With the increase of the number of UAV flight strategy operation, the signal transmission power estimation result of the method proposed by the application is gradually improved, and when the monitoring range D=140m, the signal transmission power estimation result is improved by 25.3%.

[0128] The above embodiments are the preferred embodiments of the present application, but the embodiments of the present application are not limited to the above embodiments, and any changes, modifications, substitutions, combinations, simplifications, etc. made without departing from the spirit and principles of the present application should be equivalent replacement manners and should be included in the protection scope of the present application.

Claims

1. A method for drone monitoring oriented to identification of wireless signals with spectral aliasing, characterized by, The method comprises the following steps: S1, performing short-time Fourier transform on a spectrum mixing signal received by a monitoring node in a certain spectrum sampling time slot to obtain a spectrum signal; wherein, one unmanned aerial vehicle serves as one monitoring node; S2, performing category identification on the spectrum signal through non-negative matrix factorization to construct a target function Q1; S3, solving the target function Q1, respectively deriving an updating rule for a coefficient matrix containing category information to be solved and a signal category consensus matrix, and finally obtaining an estimated value of the number of unknown signal sources; S4, splicing the coefficient matrix containing category information to be solved under different perspectives and a signal fading template matrix of the monitoring node to construct a target function Q2; S5, solving the target function Q2 through a proximal alternating linearized minimization algorithm to obtain an updating rule for the signal category and the transmit power indication matrix to be solved, and solving to obtain a signal category and a transmit power indication matrix representing the position and transmit power of the unknown signal.

2. The method of claim 1, wherein, In step S1, the mth monitoring node receives a spectral mixing signal s m (t) is: in, J m This represents the number of frequency points sampled by the m-th monitoring node within the frequency band. For the n0th unknown signal in J m The standard power spectral density function at each frequency point, and B s A standard spectrum dictionary for signals that edge nodes can recognize; Let n0 be the transmission power of the unknown signal source in time slot t. Let z be the channel gain coefficient from the n0th unknown signal source to the mth monitoring node. m (t) represents the noise at the m-th monitoring node and follows the rules. The Gaussian distribution.

3. The method of claim 1, wherein, In step S1, the channel gain coefficient of the spectrum mixing signal is defined as follows: The large-scale fading of the radio propagation in an outdoor rich scattering environment is taken into account, so that the channel gain coefficient is defined as: wherein the channel constant f is the operating center frequency of the unknown signal, c is the speed of light, E r,m and are the antenna gains of the mthmonitoring node and the nth0unknown signal, respectively; and are the channel response and distance between the mthmonitoring node and the nth0unknown signal, respectively. And there is: wherein, and (x m ,y m ,h m ) are the coordinates of the nth0unknown signal and the mthdrone in the coordinate system with the edge node as the origin respectively; the maximum horizontal coordinate and the maximum vertical coordinate of the signal monitoring range of the monitoring network are ±x max ,±y max ; remains unchanged within T s ; the link loss of the gthgrid to the mthmonitoring node d g,m is the distance from the gthgrid to the mthmonitoring node, and the signal fading dictionary in the defined area is defined as W=[W1,W2…W M ] G×M , and W m =[q 1,m ,q 2,m ……q G,m ], wherein W m is the signal fading template matrix of the mthmonitoring node.

4. The method of claim 1, wherein, for M monitored spectral aliasing signals s m (t) obtaining a spectral signal R by short-time Fourier transform m , by non-negative matrix factorization for species identification, constructing an objective function Q1, solving Q1; s.t.A m ≥0,A * ≥0; wherein R m is the spectrum aliasing signal monitored by the mth UAV, B s is the standard spectrum dictionary of the edge node identifiable signal, A m is the coefficient matrix containing the category information to be solved, A is the signal category determination matrix, indicating whether it is the signal by 0 and 1, A * is the signal category consensus matrix; Since problem Q1 is solved alternately for A m and A * is convex, the update rules are obtained by taking the derivative of A m and A * with respect to A obtaining an estimate of the number of unknown signal sources is: wherein is the matrix A * the nth element in the matrix A.

5. The method of claim 4, wherein, The A under different perspectives m With W m Splicing, strengthening the contact under each perspective, A = [A1, A2…A M ] N×M , W = [W1, W2…W M ] G×M ; Construct the objective function Q2: s.t. U ≥ 0, ||(U) n ||1=||(U) n ||2; Wherein, U is a signal category and transmit power indication matrix to be solved; Since Q2 satisfies the KL condition, the proximal alternating linearized minimization algorithm is used for solving to obtain an updating rule for U: The solved U matrix also represents the position and transmit power of the unknown signal, the nth row of U represents the position and transmit power of the nth signal, wherein the subscript of the maximum element in the row is the grid number where the nth signal is located, and the maximum element is the transmit power of the nth signal.

6. The method of claim 1, wherein, The flight strategy of the unmanned aerial vehicle is as follows: First, the grid number of the estimated unknown signal is determined The corresponding two-dimensional coordinates are calculated The two-dimensional coordinates of the grid number g are represented as where X max and Y max are the coordinates of the center of the monitoring range, g is the number of grids, and d0is the grid spacing; the barycentric coordinates (x c ,y c ) of the N0unknown signals are calculated from their two-dimensional coordinates. Construct a circle with (x c ,y c ) as the center and r c as the radius, divide the circle into M equal parts, and the mth point of division is (x m ,y m ): Each drone moves to the closest equidistant point to itself, and if there is a drone at the equidistant point, the drone moves to the next closest equidistant point to itself; for M drones, a distance matrix of the drones to M equidistant points is calculated, wherein the distance matrix of the mth drone is D m = [d m,1 ,d m,2 …d m,M ] ; If there is no unmanned aerial vehicle at the m' position, the mth unmanned aerial vehicle moves to the m' position, wherein m' is: If there is a drone at the m' position, then delete d from D m m′,M ;​ The M-frame unmanned aerial vehicles all move to the corresponding good equidivision points, monitor the signal and estimate the grid number of the unknown signal The moving method is repeatedly executed until the estimation result approaches stability.

7. A server comprising a processor and a memory, said memory having stored therein at least one program segment, characterized in that, The program is loaded and executed by the processor to realize the unmanned aerial vehicle monitoring method for spectrum aliasing wireless signal identification according to any one of claims 1 to 6.

8. A computer-readable storage medium having stored therein at least one segment of a program, characterized in that, The program is loaded and executed by the processor to realize the unmanned aerial vehicle monitoring method for spectrum aliasing wireless signal identification according to any one of claims 1 to 6.

Images