Offshore hidden monitoring method based on assistance of unmanned aerial vehicle-mounted RIS

Abstract

The invention provides a maritime hidden monitoring method based on unmanned aerial vehicle-mounted RIS assistance. The maritime hidden monitoring method comprises the following steps: step 1, establishing a system channel model; 2, calculating the reachable information rate of the system; step 3, constructing a system security optimization problem; and step 4, respectively solving the three sub-optimization problems. According to the scheme, the RIS is deployed on the interference unmanned aerial vehicle, the flight path and the RIS phase are jointly optimized, the target can be followed to move, an additional reflection channel is constructed, and the problems that the offshore long-distance transmission loss is large and the sight distance is blocked are effectively solved.

Description

Technical Field

[0001] This invention relates to the field of covert surveillance, and more particularly to a method for covert surveillance at sea based on UAV-borne RIS assistance. Background Technology

[0002] In recent years, with the surge in various maritime activities, unmanned aerial vehicles (UAVs) have been widely used in low-altitude maritime communication networks due to their advantages of on-demand deployment and flexible scheduling. While maritime UAVs offer various benefits, their malicious use by criminals could pose a serious threat to maritime public safety. Therefore, effective monitoring of low-altitude maritime UAV communication networks is crucial.

[0003] Although existing active eavesdropping technologies have attracted widespread discussion, some key issues remain unresolved: On the one hand, the impact of stealth on eavesdropping performance needs to be considered. If a legitimate eavesdropper gets too close to the target, they are easily detected by visual or radar means, thus exposing the eavesdropping mission. Therefore, strictly introducing safe distance constraints is a necessary prerequisite for eavesdropping missions. However, if the eavesdropper moves away from suspicious transmissions, the eavesdropping performance will degrade. On the other hand, in real-world scenarios, suspicious receivers usually have interference detection capabilities. If artificial noise is blindly injected to assist eavesdropping, it is easy to trigger the alarm mechanism due to abnormal power, thereby forcing the suspicious transmission to interrupt communication and causing the eavesdropping mission to fail.

[0004] Reconfigurable Smart Surfaces (RIS) technology, proposed in recent years, offers a new approach to addressing these challenges by enhancing signal reception through reconfiguration of the wireless propagation environment. However, existing RIS-assisted listening research is mostly focused on terrestrial scenarios with fixed infrastructure. The physically constrained marine environment not only limits the deployment of traditional fixed RISs but also places higher demands on the energy efficiency of UAVs due to resupply difficulties. Summary of the Invention

[0005] Purpose of the invention: The technical problem to be solved by the present invention is to address the shortcomings of the existing technology by providing a method for covert maritime surveillance based on UAV-borne RIS (Radio Router Assurance), comprising the following steps: Step 1: Establish the system channel model; Step 2: Calculate the system's achievable information rate; Step 3: Construct system security optimization issues; Step 4: Solve the three sub-optimization problems separately.

[0006] Step 1 includes: Establish a legal covert maritime surveillance system assisted by UAV-S and a suspicious vessel D, where a suspicious UAV-S and a suspicious vessel D travel along a fixed route and their communication may contain illegal information. A surveillance UAV-E is used to monitor the suspicious communication link between the suspicious UAV-S and the suspicious vessel D. A jamming UAV-J sends jamming signals to the suspicious vessel D and also provides a reflection link for the surveillance UAV-E through the RIS mounted on the top of the jamming UAV-J. All devices are configured to be equipped with a single antenna. Focusing on a UAV flight time T, and discretizing T at equal intervals... There are 1 time slot, and the length of each time slot is 1. and use To represent a set of time slots; Considering a 3D Cartesian coordinate system, the coordinates of the listening UAV-E and the jamming UAV-J are respectively expressed as: and ,in Represents the horizontal coordinate. ; The coordinates of the suspicious UAV-S and the suspicious vessel D are respectively determined by and To indicate; The mobility constraints for monitoring UAV-E and jamming UAV-J are expressed as follows: (1), (2), (3), in, , and These are the upper limits for horizontal and vertical speeds of the UAV-E for monitoring and the UAV-J for jamming. and These are the starting and ending coordinates of the monitoring system, respectively. and These represent the maximum and minimum safe flight altitudes for monitoring UAV-E and jamming UAV-J, respectively. The following minimum safe distance constraints are introduced to avoid drone collisions: (4), in It is the defined minimum safe distance for collision avoidance, and From point Time The Euclidean distance, where ; The jamming UAV-J carries a chip made of A RIS (Resonance Analysis System) consisting of a uniform linear array of reflective elements is used to enhance listening capabilities, and the phase of each element can be continuously changed by a controller. Defined as follows: (5), For the first The element in the first The phase shift at each time slot is defined as follows: (6), For RIS in the The diagonal phase shift matrix for each time slot, where diag is a diagonal matrix; e is the natural constant, and j is an imaginary number; The altitudes for monitoring UAV-E and jamming UAV-J should also meet the following constraints: (7), in It is the minimum height difference constraint; All channels are assumed to be dominated by a typical composite channel, which contains both large-scale and small-scale fading links. Therefore, in time slots... From point Time Channel power gain It conforms to the free space path loss model and is defined as: (8), in, This is the channel power gain at a reference distance of 1m. It is the Rice factor. Represents random scattering components; Channel gain from suspected UAV-S to interfering UAV-J Defined as: (9), in Represented as the set of complex numbers, It is the spacing between the antennas. It is the carrier wavelength of the transmitted signal. Indicates in time slot The cosine of the angle of arrival of the signal from the suspected UAV-S to the interfering UAV-J; The channel gain from interfering UAV-J to monitoring UAV-E is Defined as: (10) in Indicates in time slot The cosine of the reflection angle from the interfering UAV-J to the listening UAV-E.

[0007] Step 2 includes: monitoring the signals received by the UAV-E. and the signals received by the suspicious vessel D They are represented as follows: (11), (12) in and These are the transmitted signals from the suspicious UAV-S and the interfering UAV-J, respectively, and H represents the conjugate transpose of the matrix; The artificial noise (AN) transmitting antenna for interfering with UAV-J is positioned below the interfering UAV-J; and These are additive white Gaussian noises located at the suspicious vessel D and the listening UAV-E, respectively. and The average power is respectively and ; Formula (12) can be rewritten as: (13) Channel capacity of suspicious vessel D and the channel capacity of UAV-E for monitoring The following are given by Shannon's formula: (14) (15) in and These are the transmission power of the suspicious UAV-S and the transmission power of the interfering UAV-J, respectively, and the suspicious UAV-S is configured to adaptively adjust its information rate according to the channel conditions.

[0008] Step 3 includes: Interference power The following constraints must be met: (16) (17) in This refers to all the energy remaining in the jamming UAV-J after deducting the energy used for flight. This is the maximum peak power of the jamming UAV-J. Monitoring UAV-E and jamming UAV-J should maintain a certain safe distance from suspected systems to ensure the concealment of monitoring activities. Specific constraints are as follows: (18) (19) in , and These are the minimum concealed distances from the listening system to the suspicious UAV-S and the minimum concealed distances from the listening system to the suspicious vessel D, respectively. Assuming the suspicious vessel D possesses noise detection capabilities, it will collect channel state information and analyze the ambient noise level to determine whether it is being interfered with by AN. When the detected ambient noise power exceeds a decision threshold, it will terminate communication with the suspicious UAV-S. Therefore, the noise power interfering with UAV-J should also satisfy the following: (20), in This is the noise decision threshold for the suspicious vessel D; Constraint (17) is rewritten as: (twenty one), (twenty two), If and only if the listening channel capacity Not less than the suspicious communication rate Only when the listening party has a sufficiently small error probability can the information be decoded; otherwise, it cannot obtain all the information without distortion. Therefore, an effective listening rate is defined. for: (twenty three), The objective is to jointly optimize the 3D trajectories of the listening UAV-E and the jamming UAV-J, the diagonal phase shift matrix of the RIS, and the AN jamming power, under the constraints of mobility, safety, stealth, and energy limitations of dual UAVs, to maximize effective listening and speed. This can be expressed as: (twenty four); Formulas (1)~(4); (5), (7), (18), (19); (16), (21).

[0009] Step 4 includes: Initial problem equivalent transformation: First, introduce indicator functions. To indicate whether the monitoring was successful, the optimization issues after conversion are as follows: (25) (26) Formulas (1)~(4); (5), (7), (18), (19); (16), (21); The initial problem is decomposed into three sub-problems: trajectory optimization of the monitoring UAV-E, joint optimization of the trajectory of the jamming UAV-J and the diagonal phase shift matrix of RIS, and jamming power optimization of the jamming UAV-J. Finally, the suboptimal solution of the initial problem is obtained by rotating the optimization of the three sub-problems.

[0010] In step 4, the trajectory optimization of the monitoring UAV-E includes: for a given trajectory of the interfering UAV-J Interference power and diagonal phase shift matrix Problem (25) is rephrased as:

[0011] (1)~(4); (7), (18), (19); (26); (27); Replace the objective function with or Problem (27) is rewritten as: (28); (1) ~ (4); (7), (18), (19); First, define the given diagonal phase shift matrix components. This is always optimal, yielding the following result: (29); Signal-to-interference-plus-noise ratio of UAV-E receiver The lower bound is obtained by averaging over a time slot. As shown below:

[0012]

[0013] (30) in, , It is the lower bound of the small-scale fading between point i and point j; Lower bound of achievable eavesdropping rate Written as:

[0014]

[0015] (31), intermediate parameters intermediate parameters intermediate parameters ; By introducing slack variables and Defined as: (32), (33), Given a constant , and In the case of equation Compared to and It is a convex function, for Expand to obtain the lower bound. for:

[0016] (34), intermediate parameters ; and It is in the The feasible solution obtained in the next iteration; according to formulas (30)~(34), the problem defined in (28) is approximately transformed into: (35), (1) ~ (4); (7), (18), (19); (32), (33).

[0017] In step 4, the joint optimization of the trajectory of the interfering UAV-J and the diagonal phase shift matrix of RIS includes: given the trajectory of the monitoring UAV-E and interference power In this case, problem (25) is transformed into: (36) (1)~(4); (5), (7), (18), (19), (23); Replace the objective function with Additional weights are added here. This causes the objective function to further change to The new questions that arise are as follows: (37) (1) ~ (4); (5), (7), (18), (19); Problem (37) can be solved in two parts: the phase shift matrix. Optimization and interference of UAV-J trajectory The optimization, given the disturbed UAV-J trajectory, yields: (38) By aligning the phase of the signal received by the interfering UAV-J, the energy of the received signal can be maximized. Therefore, the following settings are made:

[0018] (39) Or it can be expressed as: (40), (41), in ; The following results were obtained: (42), Then, based on formula (31), when the variable is transformed into the trajectory of the interfering UAV-J, the achievable monitoring rate of the monitoring UAV-E is determined. Represented as: (43), intermediate parameters intermediate parameters ; Then, slack variables are introduced. and Defined as: (44), (45) And on Performing a first-order Taylor expansion yields the lower bound. for:

[0019] (46) intermediate parameters ; and It is in the The feasible solution obtained in the next iteration; right The upper bound is obtained by averaging over a time slot. for:

[0020]

[0021] (47) in , It is the upper bound of the small-scale fading between point i and point j; a new slack variable is introduced. Defined as: (48) Problem (37) is refactored as follows: (49) (1)~(4); (5), (7), (18), (19); (44), (45), (48).

[0022] In step 4, the interference power optimization of the interfering UAV-J includes: given the trajectories of the monitoring UAV-E and the interfering UAV-J and the phase shift matrix, problem (25) is rewritten as: (50), (16), (21), (23); Derive the optimal interference power exist Obtained at that time, according to The value may fall into one of the following three categories. Further processing is required. , and ; when At that time, the actual optimal interference power was set to ; when At that time, the optimal interference power is set to ; when At that time, the optimal interference power is set to ; Optimal interference power for problem (50) The derivation is as follows: (51), in: (52), (53), (54), The energy efficiency calculation will be reset to: (55).

[0023] The present invention also provides an electronic device, including a processor and a memory, the memory storing program code that, when executed by the processor, causes the processor to perform the steps of the method.

[0024] The present invention also provides a storage medium storing a computer program or instructions that, when the computer program or instructions are run on a computer, execute the steps of the method described.

[0025] Beneficial Effects: With the increasing demand for maritime communication and the development of the low-altitude economy, the maritime low-altitude legal eavesdropping scheme with RIS assistance for UAVs under energy-constrained conditions proposed in this invention has significant performance improvements in the following aspects compared to traditional fixed facilities or single jamming technologies: (1) Constructing a dynamic reflection channel: Compared with the traditional fixed RIS which is limited by the maritime geographical environment, this scheme deploys the RIS on the jamming UAV. By jointly optimizing the flight trajectory and the RIS phase, it can follow the target and construct an additional reflection channel, effectively solving the problems of large transmission loss and obstructed line of sight at sea.

[0026] (2) Achieving covert jamming and monitoring: For suspicious targets with jamming detection capabilities, this solution intelligently and collaboratively plans the trajectory and power of jamming UAVs, dynamically adjusting the strategy while meeting the covert detection threshold. This enables the system to effectively jam while preventing the target from noticing and interrupting communication, maintaining efficient monitoring in dynamic adversarial environments.

[0027] (3) Extending the operational endurance at sea: To address the pain point of difficult resupply of UAVs at sea, this solution designs an optimal interference power allocation algorithm under energy constraints. Compared with traditional fixed power or continuous interference schemes, this strategy can intelligently adjust energy consumption according to channel conditions, significantly improving energy efficiency while ensuring monitoring performance and reducing the cost of operations at sea.

[0028] In summary, this solution significantly outperforms existing technologies in terms of surveillance concealment, environmental adaptability, and energy efficiency, providing an efficient and reliable solution for low-altitude safety management in complex marine environments. Attached Figure Description

[0029] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, and the advantages of the present invention in the above and / or other aspects will become clearer.

[0030] Figure 1 This is a schematic diagram of a legal maritime surveillance system assisted by a drone-borne RIS.

[0031] Figure 2 This is a schematic diagram (3D trajectory) illustrating the impact of different UAV horizontal speed limits on the trajectories of the monitoring UAV-E and the interfering UAV-J.

[0032] Figure 3 This is a schematic diagram (top view) illustrating the impact of different UAV horizontal speed limits on the trajectories of the monitoring UAV-E and the interfering UAV-J.

[0033] Figure 4 This is a schematic diagram (height map) illustrating the impact of different UAV horizontal speed limits on the trajectories of the monitoring UAV-E and the interfering UAV-J.

[0034] Figure 5 This is a schematic diagram illustrating the impact of the number of RIS components on monitoring UAV-E trajectories.

[0035] Figure 6 This is a schematic diagram illustrating the impact of RIS deployment and the horizontal speed limit of dual UAVs on the effective listening rate.

[0036] Figure 7 This is a schematic diagram illustrating the impact of RIS deployment and the horizontal speed limit of dual UAVs on power allocation.

[0037] Figure 8 Different total power limits and noise detection threshold A schematic diagram illustrating the impact on power distribution.

[0038] Figure 9 This is a diagram comparing the performance of different solutions. Detailed Implementation

[0039] This embodiment provides a method for covert maritime surveillance based on UAV-borne RIS assistance, including the following steps: Step 1: Establish the system channel model; This method considers a legal, covert maritime surveillance system assisted by an unmanned aerial vehicle (UAV)-borne RIS (Reference Signal). Suspicious UAV-S and suspicious vessel D travel along a fixed route, and their communications may contain illegal information. Therefore, UAV-E is used to monitor its downlink, while UAV-J, while sending jamming signals to vessel D, also provides a reflection link to UAV-E via its top-mounted RIS, increasing its monitoring rate. For ease of analysis, all devices are configured with a single antenna. This method primarily focuses on a specific UAV flight period. and discretize it at equal intervals into There are 1 time slot, and the length of each time slot is 1. and use To represent the set of time slots, when When the coordinate trajectories are sufficiently small, the coordinate trajectories of adjacent time slots can be considered approximately continuous. Further considering a 3D Cartesian coordinate system, the coordinates of the listening UAV-E and the jamming UAV-J can be expressed as follows: and ,in The coordinates of the RIS are represented by its horizontal coordinates. Since the RIS is fixed on top of the interfering UAV-J, its coordinates can be represented by the coordinates of the interfering UAV-J. Correspondingly, the coordinates of the suspicious UAV-S and the suspicious vessel D are represented by... and To represent. Based on the above settings, the mobility constraints of the monitoring UAV-E and the jamming UAV-J are expressed as follows: (1), (2), (3), in , , and These are the upper limits of the horizontal and vertical speeds for monitoring UAV-E and jamming UAV-J, respectively. and These are the starting and ending coordinates of the monitoring system, respectively. and These represent the highest and lowest safe flight altitudes for monitoring UAV-E and jamming UAV-J, respectively. Furthermore, considering UAV flight safety, the following minimum safe distance constraints are introduced to prevent UAV collisions: (4), in It is the defined minimum safe distance for collision avoidance, and From point Time The Euclidean distance, where .

[0040] To counter the eavesdropping noise detection capabilities of the suspected vessel D, the jamming UAV-J carried a device consisting of... A RIS (Resonance Analysis System) consisting of a uniform linear array of reflective elements is used to enhance listening capabilities, and the phase of each element can be continuously changed by a controller. Defined as follows: (5), For the first The element in the first The phase shift at each time slot is defined as follows: (6), For RIS in the The diagonal phase shift matrix for each time slot, considering that the RIS is horizontally deployed on top of the jamming UAV-J, and the signal source of the RIS and the user should be on the same side of the RIS, therefore the heights of the monitoring UAV-E and the jamming UAV-J should also satisfy the following constraints: (7), in This is the minimum height difference constraint. Because maritime channels are susceptible to the effects of the marine environment, all channels are assumed to be dominated by a typical composite channel, which simultaneously contains both large-scale and small-scale fading links. Therefore, in time slots... From point Time The channel power gain conforms to the free-space path loss model and is defined as: (8), in, This is the channel power gain at a reference distance of 1m. It is the Rice factor, which is usually a constant. Let represent the random scattering component, which is a zero-mean, unit-variance, circularly symmetric complex Gaussian random variable. Therefore, the channel gain from the suspected UAV-S to the interfering UAV-J (RIS) can be expressed as: Defined as: (9), The right half of the right side of equation (9) is The array response of each element. It is the spacing between the antennas. It is the carrier wavelength of the transmitted signal. Indicates in time slot Let $\frac{ ... Defined as: (10) in Indicates in time slot The cosine of the reflection angle from the interfering UAV-J to the listening UAV-E.

[0041] Step 2: Calculate the system's achievable information rate; Based on the channel description in step 1, the signals received by the monitored UAV-E and the suspicious vessel D can now be represented as follows: (11), (12) here and These are the transmitted signals from the suspicious UAV-S and the interfering UAV-J, respectively. Because the suspicious vessel D is at a lower altitude than the interfering UAV-J (RIS), it cannot receive the reflected signal from RIS. Furthermore, the AN transmitting antenna of the interfering UAV-J is positioned directly below it. and The additive white Gaussian noise at the suspicious vessel D and the listening UAV-E, respectively, has an average power of and It is worth noting that, This is completely unknown to the suspected vessel D, and therefore can be treated as noise. and channel gain For UAV-E monitoring, this is completely known. For ease of analysis, it is assumed that it can be completely eliminated upon reception. Therefore, formula (12) can be rewritten as: (13) In summary, the channel capacities of the suspicious vessel D and the eavesdropping UAV-E can be given by Shannon's formulas respectively: (14) (15) in and These are the transmission powers of the suspicious UAV-S and the interfering UAV-J, respectively, and it is assumed that the suspicious UAV-S will adaptively adjust its information rate according to the channel conditions.

[0042] Step 3: Construct system security optimization issues; Because the jamming UAV-J is powered by a limited onboard battery, and it cannot land on the sea to recharge its energy, the jamming power is limited. The following constraints must be met: (16) (17) here This refers to all the energy remaining in the jamming UAV-J after deducting the energy used for flight. This represents the maximum peak power of the interfering UAV-J. Furthermore, this method considers that the suspicious vessel D may consciously detect whether it is being monitored. This detection includes checking for the deliberate approach of other equipment; equipment that gets too close will be identified as potential monitoring devices. Therefore, when the monitoring UAV-E or the interfering UAV-J gets too close to the suspicious communication system, the suspicious communication system will terminate its communication activities. To avoid this situation, the monitoring UAV-E and the interfering UAV-J should maintain a certain safe distance from the suspicious system to ensure the concealment of the monitoring activities. The corresponding constraints are as follows: (18) (19) in , and These are the minimum covert distances from the monitoring system to the suspected UAV-S and the suspected vessel D, respectively. Furthermore, the suspected vessel D is assumed to have noise detection capabilities; it will collect channel state information and analyze the level of ambient noise to determine whether it is being interfered with by the AN. When the detected ambient noise power exceeds its decision threshold, communication with the suspected UAV-S will be terminated. Therefore, the noise power interfering with UAV-J should also satisfy: (20), here This is the noise decision threshold for the suspicious vessel D. Therefore, constraint (17) is rewritten as: (twenty one), (twenty two), If and only if the listening channel capacity Not less than the suspicious communication rate Only when the listening party has a sufficiently small error probability can it decode the information; otherwise, it cannot obtain all the information without distortion. Therefore, an effective listening rate is defined. for: (twenty three), The aim of this method is to jointly optimize the 3D trajectories of the listening UAV-E and the jamming UAV-J, the diagonal phase shift matrix of the RIS (Rapid Recognition Matrix), and the AN (Anti-Negative Array) jamming power, under the constraints of mobility, safety, stealth, and energy limitations of dual UAVs, to maximize effective listening and data rate. This problem can be formulated as: (twenty four); Formulas (1)~(4); (5), (7), (18), (19); (16), (21); Problem (24) is a nonconvex optimization problem because it involves a highly nonlinear objective function. Depends on optimization variables Obtaining the global optimal solution is very challenging. Therefore, in the next step, we will introduce an efficient iterative algorithm to obtain a feasible solution to problem (24).

[0043] Step 4: Solve the three sub-optimization problems separately; (1) Initial problem equivalent transformation. First, introduce the indicator function. To indicate whether the monitoring was successful, the optimization issues after conversion are as follows: (25) (26) Formulas (1)~(4); (5), (7), (18), (19); (16), (21); In the following three sections, the initial problem is decomposed into three subproblems: trajectory optimization for the monitoring UAV-E, joint optimization of the trajectory of the jamming UAV-J and the diagonal phase shift matrix of the RIS, and jamming power optimization for the jamming UAV-J. Finally, a suboptimal solution to the initial problem is obtained through the rotational optimization of the three subproblems.

[0044] (2) Monitoring UAV trajectory optimization. For a given trajectory of the interfering UAV-J... Interference power and diagonal phase shift matrix Problem (25) can be restated as:

[0045] (1)~(4); (7), (18), (19); (26); (27); suspicious communication rate Given a specific target, optimizing the tracking path for UAV-E can result in a better listening rate. The probability of successful listening will also increase accordingly. Therefore, in order to successfully listen to more time slots and obtain higher effective listening and speed, this section replaces the objective function with... or Thus, problem (27) is rewritten as: (28); (1) ~ (4); (7), (18), (19); because The molecule contains diagonal phase shift matrix components. Problem (28) is difficult to solve directly. Here, this section first assumes the given diagonal phase shift matrix components. This is always optimal, and we can obtain the following result: (29); because It includes small-scale fading, and the problem (28) remains highly nonlinear, thus affecting the received signal-to-interference-plus-noise ratio of the UAV-E. Taking the average value over a time slot yields its lower bound as shown below:

[0046]

[0047] (30) in, , This is the lower bound of the small-scale fading between points i and j. Therefore, the lower bound of the achievable eavesdropping rate is written as:

[0048]

[0049] (31), in , , .

[0050] In addition, by introducing slack variables and Defined as: (32), (33), In a given , and In the case of equation Compared to and It is a convex function. Based on the above description and the fact that the first-order Taylor expansion approximation of a convex function is a global estimate, for... By expanding it, we can obtain its lower bound. for:

[0051] (34), in ,here, and It is in the The feasible solution obtained in the next iteration. According to formulas (30)~(34), the problem defined in (28) is approximately transformed into: (35), (1)~(4); (7), (18), (19); (32), (33); At this point, problem (35) has been transformed from the SCA method into a solvable form, which can be solved using existing standard convex optimization techniques. The obtained trajectory of the listening UAV-E will be incorporated into the calculations in the next subsection.

[0052] (3) Joint optimization of interference UAV trajectory and RIS phase shift. Given the trajectory of the monitoring UAV-E... and interference power In this case, problem (25) is transformed into: (36) (1)~(4); (5), (7), (18), (19), (23); By optimizing the trajectory of the interfering UAV-J and the diagonal phase shift matrix of RIS Listening rate and suspicious communication rates Everything will change. It is foreseeable that the closer the interfering UAV-J is to the suspicious vessel D, the lower the suspicious communication rate will be. The lower the value, the better, but this phenomenon can be offset by optimizing the AN interference power. Similarly, in order to achieve successful listening in more time slots, this section replaces the objective function with... Additionally, to adjust the target bias of the interfering UAV-J, extra weights are added here. This causes the objective function to further change to By changing The value of this value can make the jamming UAV-J more inclined to provide better listening performance for the listening UAV-E or to approach the suspicious vessel D to reduce power consumption. The resulting new problem is as follows: (37) (1) ~ (4); (5), (7), (18), (19); Problem (37) can be solved in two parts: the phase shift matrix. Optimization and interference of UAV-J trajectory Optimization. This section first presents the optimization scheme for the phase shift matrix. Given the interference UAV-J trajectory, we obtain: (38) Here, by aligning the phase of the signal received by the interfering UAV-J, the energy of the received signal can be maximized. Therefore, the following settings are made:

[0053] (39) Or it can be expressed as: (40), (41), in This completes the phase alignment, allowing the interfering UAV-J to obtain maximum signal reception energy, resulting in the following outcome: (42), Then, based on formula (31), when the variable is transformed into the trajectory of the interfering UAV-J, the achievable monitoring rate of the monitoring UAV-E is expressed as: (43), in , .

[0054] Then, slack variables are introduced. and Defined as: (44), (45) And on Performing a first-order Taylor expansion yields the lower bound. for:

[0055] (46) in ,here and It is in the The feasible solution obtained in the next iteration. Similarly, for The upper bound is obtained by averaging over a time slot. for:

[0056]

[0057] (47) in , This is the upper bound of the small-scale fading between point i and point j. However, even using an upper bound for the questionable communication rate... to replace Problem (37) remains difficult to solve, so new slack variables are introduced. Defined as: (48) Thus, problem (37) is restructured as follows: (49) (1)~(4); (5), (7), (18), (19); (44), (45), (48); Thus, problem (49) can now be conveniently solved using the interior point method, and the obtained trajectory of the interfering UAV-J will be incorporated into the calculations in the next subsection. Simultaneously, the optimal diagonal phase shift matrix of RIS... It can also be obtained through formula (41). It is worth noting that the optimal... It depends not only on the trajectory of the jamming UAV-J, but also closely on the trajectory of the monitoring UAV-E. Therefore, after the trajectories of both the monitoring UAV-E and the jamming UAV-J are determined, the optimal diagonal phase shift matrix is ​​then calculated.

[0058] (4) Optimization of jamming UAV power. Given the trajectories and phase shift matrices of the listening UAV-E and jamming UAV-J, problem (25) is rewritten as: (50), (16), (21), (23); Derive the optimal interference power exist It was obtained at that time, but what was obtained at that time It doesn't necessarily mean that the power constraint of the interfering UAV-J is not met, therefore it is also necessary to consider... Three possible scenarios for the value: Further processing is required. That is: , and .

[0059] when This means that the channel quality of the monitoring ship E is inherently superior to that of the suspicious communication, allowing successful monitoring and acquisition of all information from the suspicious communication even without AN jamming. Therefore, the optimal jamming power should be set to... .

[0060] when This means that in the current time slot, the power... Sending AN satisfies the conditions for successful interception; therefore, the optimal interference power is set to... .

[0061] when This indicates that even when interfering with UAV-J at maximum interference power... Injecting AN into the suspected vessel D could not achieve successful interception. Therefore, to conserve power, the optimal jamming power for this time slot was set to... .

[0062] Therefore, the optimal interference power for problem (50) The derivation is as follows: (51), in: (52), (53), (54), In practical applications, since UAVs are powered by onboard batteries with limited energy, the jamming UAV-J may not be able to allocate appropriate jamming power in every time slot. Therefore, the energy efficiency calculation will be reset to: (55), Through the above derivation, it is possible to obtain the optimal AN jamming power that maximizes effective listening and rate while satisfying power and concealment constraints.

[0063] Step 5: Set system simulation parameters; This section uses simulation results to confirm the performance of the proposed solution. The relevant parameters are as follows: the coordinates of the initial and final positions of the monitoring UAV-E and the interfering UAV-J are respectively: , , and Furthermore, the maximum and minimum flight altitudes of the drone are... and Furthermore, unless otherwise specified, the maximum horizontal and vertical speeds of the monitoring UAV-E and the jamming UAV-J are respectively... and The transmission power of the suspected UAV-S is The peak power of the interfering UAV-J is The total power of the interfering UAV-J is The number of components in the RIS is The minimum collision avoidance distance for drones is The minimum altitude difference constraint for the drone is The concealment distance to the suspicious system is and The channel power gain per unit distance is The average power of additive white Gaussian noise is The noise detection threshold for suspicious vessel D is The value of Rice factor is Therefore, the upper and lower bounds of small-scale fading are respectively and Considering the power-saving requirements of interfering with UAV-J, the weights of the multi-objective function are set to... .

[0064] In this embodiment, a schematic diagram of a UAV-borne RIS-assisted maritime lawful surveillance system is shown below. Figure 1 As shown, a jamming drone can secretly send jamming signals to a suspicious receiver, while a deployed RIS can potentially create a listening channel against a legitimate drone. The goal of this method is to jointly design the three-dimensional flight trajectory and power allocation of the jamming drone, the three-dimensional flight trajectory of the legitimate drone, and the reflection phase shift of the RIS to maximize the sum and eavesdropping rate across all time slots. The eavesdropping method includes the following steps: Step 1: Establish a system cascaded channel model; Step 2: Calculate the system's achievable information rate; Step 3: Construct system security optimization issues; Step 4: Solve the three sub-optimization problems separately; Step 5: Set system simulation parameters; Step 6. Please refer to the simulation results. Figures 2-9 .

[0065] Figures 2-4 The impact of different maximum horizontal velocity limits of the two UAVs on the trajectories of the listening UAV-E and the interfering UAV-J is presented. Indicates the starting position. Indicates the endpoint location. From Figure 3 As can be seen from this, when the upper limit of horizontal speed is relatively small (i.e. The listening UAV-E and jamming UAV-J will first approach their respective targets as close as possible, and then fly towards their destination due to insufficient flight time. When the horizontal speed limit is high (i.e., After the monitoring system approaches the suspicious communication system, it will maintain a follow-flying pattern with its respective target for a period of time. This is because the monitoring UAV-E needs to get close to the suspicious UAV-S to obtain more information, while the jamming UAV-J needs to get close to the suspicious vessel D to reduce power consumption. Figure 2 and Figure 4 As can be seen, in the initial few time slots, the listening UAV-E did not show a tendency to approach the suspicious UAV-S in terms of altitude. Instead, it chose to approach the interfering UAV-J. This is because when the listening system is far from the suspicious communication system, the signal component power from the suspicious UAV-S is low, while the reflected signal component provided by the interfering UAV-J is strong. Therefore, the listening UAV-E tends to approach the interfering UAV-J in the initial few time slots. However, when the listening system is closer to the suspicious communication system, the signal component from the suspicious UAV-S becomes stronger, and the listening UAV-E will find the position with the highest overall signal reception power between the suspicious UAV-S and the interfering UAV-J for listening. In addition, when the horizontal speed limit of the two UAVs is high (i.e., The jamming UAV-J will adjust its altitude to meet the concealment distance constraint while keeping following the suspicious vessel D, while the listening UAV-E will adjust its altitude in sync to avoid the collision avoidance distance constraint.

[0066] Figure 5 The impact of the number of components in the RIS (Reflection Link System) on the trajectory of the eavesdropping UAV-E is depicted. Because the jamming UAV-J is equipped with a RIS that provides a reflection link for the eavesdropping UAV-E, the trajectory of the eavesdropping UAV-E shows a tendency to approach the jamming UAV-J (RIS). This is because the reflection link component is stronger than the direct link component when the eavesdropping system is far from the suspected communication system. Figure 5 It can be seen that when the number of components in the RIS is At that time, the listening UAV-E showed results that were closer to the interfering UAV-J. However, when the number of components in the RIS decreased (i.e., The listening UAV-E will fly directly towards the suspected UAV-S, and after reaching the concealment distance constraint, it will choose to descend to approach the interfering UAV-J. This is because the signal components from the interfering UAV-J (RIS) are weakened, and the listening UAV-E will tend to fly towards the source of the direct link. When it cannot get closer to the suspected UAV-S, it will choose to descend to approach the interfering UAV-J to find the location with the best overall signal reception power.

[0067] Figure 6 This paper depicts the effective listening rate over time slots with and without RIS (Real-Inspection System) and with varying horizontal speed limits for dual UAVs. Overall, the effective listening rate initially increases and then decreases with increasing time slots, due to the listening system's approach and distance from the suspected communication system. As the horizontal speed limit of the listening system gradually increases, the effective listening rate per time slot also increases because the listening UAV-E and jamming UAV-J can approach the suspected communication system more quickly, thus achieving a higher listening rate. However, when the speed limit is sufficiently high, the listening system cannot approach the suspected communication system further due to covert distance constraints, resulting in a theoretical upper limit for the effective listening rate. (Comparison) Figure 6 The changes in the effective listening rate with and without RIS show that installing RIS can not only effectively increase the effective listening rate per time slot, but also increase its theoretical upper limit.

[0068] Figure 7 The paper presents the effects of different horizontal speed limits for dual UAVs and the presence or absence of a RIS (Rapid Interference System) on power allocation. Overall, as the number of time slots increases, the power allocation per time slot initially decreases and then increases. This is because when the listening system is far from the suspected communication system, the listening rate is low, requiring higher jamming power for successful listening. Conversely, when the distance is close, the listening rate is high, and the required AN (Anti-Aggressive Power) is lower. When the horizontal speed limit of the listening system is low, the listening system cannot quickly approach the suspected communication system, thus requiring higher AN power. Increasing the speed limit of the jamming UAV-J allows it to approach the suspected vessel D more closely, resulting in lower jamming power consumption. When the horizontal speed limit of the listening UAV-J also increases, the overall system's achievable listening rate... This will also increase, thus further reducing the interference power required per time slot. (Comparison) Figure 7 The interference power allocation with and without RIS shows that, at the same speed limit, the AN power required per time slot is lower when RIS is installed. This is because RIS can provide additional monitoring performance gain for the monitoring UAV-E, thereby reducing the consumption of interference power.

[0069] Figure 8Depicting different total power and noise detection threshold Impact on interference power allocation. Figure 8 The two black dashed lines represent the noise monitoring thresholds for the suspected vessel D. and The corresponding concealment power limit When the noise detection threshold for the suspected vessel D is relatively lenient (i.e. The optimal interference power for each time slot did not exceed the concealment power constraint. At this point, as long as the total power Sufficient power allows for the allocation of appropriate interference power to each time slot (i.e., When the total power is finite (i.e.) When interfering with UAV-J, some time slots with lower energy efficiency will be removed and no longer allocated interference power, instead allocating the limited power to time slots with higher energy efficiency. When the noise detection threshold of the suspected vessel D is relatively strict (i.e., Even if the interfering UAV-J has sufficient total power (i.e. In some time slots, AN transmissions also had to be cancelled because continued transmission of noise interference would cause suspicious communication systems to cease communication, thus causing the eavesdropping task to fail. When the total power further decreased (i.e.... The interference UAV-J will first meet the requirements of concealment power constraints, and then allocate interference power according to the energy efficiency level.

[0070] Figure 9 The paper depicts a performance comparison of various schemes. The total power on the horizontal axis is [data missing]. From a constraint perspective, under the same total power limit, the UAV-borne RIS-assisted 3D joint optimization scheme proposed in this paper achieves the highest average effective listening rate. The average effective listening rate is represented by the vertical axis. To achieve the same effective listening rate, the proposed scheme requires the least total power. In the 2D joint optimization scheme, the fixed height of the two UAVs prevents the listening UAV-E from approaching the interfering UAV-J by changing its height, thus weakening the reflection link from the RIS and resulting in a loss of effective listening rate. In the 3D joint optimization scheme, reducing the number of reflection elements in the RIS also reduces the listening performance per time slot. Without the RIS, the listening performance drops to its lower limit. In the fixed power allocation scheme, the interfering UAV-J distributes all the interference power evenly across each time slot. As the total power increases, more new time slots can be successfully listened to, and the average effective listening rate increases. However, once the increase in total power fails to generate new listenable time slots, the average effective listening rate decreases due to the increase in interference power.

[0071] This embodiment also proposes a novel maritime low-altitude communication monitoring system using a jamming UAV equipped with a RIS (Reference Signal Processing) system. In this system, a legitimate monitoring UAV, with the aid of a jamming UAV equipped with a RIS, collaboratively monitors suspicious UAV-ship communication links. Specifically, this system considers a practical and challenging scenario where the suspicious receiver has artificial noise detection capabilities, and the UAV performing the jamming mission faces severe energy constraints. To address these challenges, this solution jointly optimizes the three-dimensional flight trajectories of the two UAVs, the RIS reflection phase shift, and the jamming power allocation, enabling the jamming UAV to covertly transmit jamming signals to the suspicious receiver while simultaneously utilizing the airborne RIS to construct an additional reflection monitoring channel to the legitimate UAV. This design not only solves the deployment challenges of RIS at sea but also significantly improves the system's effective monitoring rate while meeting concealment and energy constraints.

[0072] This invention provides a method for covert maritime surveillance based on UAV-borne RIS (Radio Router Assurance). Many methods and approaches exist for implementing this technical solution; the above description is merely a preferred embodiment. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention. All components not explicitly stated in this embodiment can be implemented using existing technologies.

Claims

1. A method for covert maritime surveillance based on UAV-borne RIS assistance, characterized in that, Includes the following steps: Step 1: Establish the system channel model; Step 2: Calculate the system's achievable information rate; Step 3: Construct system security optimization issues; Step 4: Solve the three sub-optimization problems separately.

2. The method according to claim 1, characterized in that, Step 1 includes: Establish a legal covert maritime surveillance system assisted by a reconfigurable intelligent surface (RIS) carried by an unmanned aerial vehicle (UAV). A suspicious UAV-S and a suspicious vessel D travel along a fixed route, and their communication may contain illegal information. A surveillance UAV-E is used to monitor the suspicious communication link between the suspicious UAV-S and the suspicious vessel D. A jamming UAV-J sends jamming signals to the suspicious vessel D and also provides a reflection link for the surveillance UAV-E through the RIS mounted on the top of the jamming UAV-J. All devices are configured to be equipped with a single antenna. Focusing on a UAV flight time T, and discretizing T at equal intervals... There are 1 time slot, and the length of each time slot is 1. and use To represent a set of time slots; Considering a 3D Cartesian coordinate system, the coordinates of the listening UAV-E and the jamming UAV-J are respectively expressed as: and ,in Represents the horizontal coordinate. ; The coordinates of the suspicious UAV-S and the suspicious vessel D are respectively determined by and To indicate; The mobility constraints for monitoring UAV-E and jamming UAV-J are expressed as follows: （1）， （2）， （3）， in, , and These are the upper limits for horizontal and vertical speeds of the UAV-E for monitoring and the UAV-J for jamming. and These are the starting and ending coordinates of the monitoring system, respectively. and These represent the maximum and minimum safe flight altitudes for monitoring UAV-E and jamming UAV-J, respectively. The following minimum safe distance constraints are introduced to avoid drone collisions: （4）， in It is the defined minimum safe distance for collision avoidance, and From point Time The Euclidean distance, where ; The jamming UAV-J carries a chip made of A RIS (Resonance Analysis System) consisting of a uniform linear array of reflective elements is used to enhance listening capabilities, and the phase of each element can be continuously changed by a controller. Defined as follows: （5）， For the first The element in the first The phase shift at each time slot is defined as follows: （6）， For RIS in the The diagonal phase shift matrix for each time slot, where diag is a diagonal matrix; e is the natural constant, and j is an imaginary number; The altitudes for monitoring UAV-E and jamming UAV-J should also meet the following constraints: （7）， in It is the minimum height difference constraint; All channels are assumed to be dominated by a typical composite channel, which contains both large-scale and small-scale fading links. Therefore, in time slots... From point Time Channel power gain It conforms to the free space path loss model and is defined as: （8）， in, This is the channel power gain at a reference distance of 1m. It is the Rice factor. Represents random scattering components; Channel gain from suspected UAV-S to interfering UAV-J Defined as: （9）， in Represented as the set of complex numbers, It is the spacing between the antennas. It is the carrier wavelength of the transmitted signal. Indicates in time slot The cosine of the angle of arrival of the signal from the suspected UAV-S to the interfering UAV-J; The channel gain from interfering UAV-J to monitoring UAV-E is Defined as: （10）， in Indicates in time slot The cosine of the reflection angle from the interfering UAV-J to the listening UAV-E.

3. The method according to claim 2, characterized in that, Step 2 includes: monitoring the signals received by the UAV-E. and the signals received by the suspicious vessel D They are represented as follows: （11）， （12）， in and These are the transmitted signals from the suspicious UAV-S and the interfering UAV-J, respectively, and H represents the conjugate transpose of the matrix; The artificial noise AN transmitting antenna for interfering with UAV-J is positioned below the interfering UAV-J; and These are additive white Gaussian noises located at the suspicious vessel D and the listening UAV-E, respectively. and The average power is respectively and ; Formula (12) can be rewritten as: （13）， Channel capacity of suspicious vessel D and the channel capacity of UAV-E for monitoring The following are given by Shannon's formula: （14）， （15）， in and These are the transmission power of the suspicious UAV-S and the transmission power of the interfering UAV-J, respectively, and the suspicious UAV-S is configured to adaptively adjust its information rate according to the channel conditions.

4. The method according to claim 3, characterized in that, Step 3 includes: Interference power The following constraints must be met: （16）， （17）， in This refers to all the energy remaining in the jamming UAV-J after deducting the energy used for flight. This is the maximum peak power of the jamming UAV-J. Monitoring UAV-E and jamming UAV-J should maintain a certain safe distance from suspected systems to ensure the concealment of monitoring activities. Specific constraints are as follows: （18）， （19）， in , and These are the minimum concealed distances from the listening system to the suspicious UAV-S and the minimum concealed distances from the listening system to the suspicious vessel D, respectively. Assuming the suspicious vessel D possesses noise detection capabilities, it will collect channel state information and analyze the ambient noise level to determine whether it is being interfered with by AN. When the detected ambient noise power exceeds a decision threshold, it will terminate communication with the suspicious UAV-S. Therefore, the noise power interfering with UAV-J should also satisfy the following: （20）， in This is the noise decision threshold for the suspicious vessel D; Constraint (17) is rewritten as: （21）， （22）， If and only if the listening channel capacity Not less than the suspicious communication rate Only when the listening party has a sufficiently small error probability can the information be decoded; otherwise, it cannot obtain all the information without distortion. Therefore, an effective listening rate is defined. for: （23）， The objective is to jointly optimize the 3D trajectories of the listening UAV-E and the jamming UAV-J, the diagonal phase shift matrix of the RIS, and the AN jamming power, under the constraints of mobility, safety, stealth, and energy limitations of dual UAVs, to maximize effective listening and speed. This can be expressed as: （24）； Formulas (1) to (4); (5), (7), (18), (19); (16), (21).

5. The method according to claim 4, characterized in that, Step 4 includes: Initial problem equivalent transformation: First, introduce indicator functions. To indicate whether the monitoring was successful, the optimization issues after conversion are as follows: （25）， （26）， Formulas (1)~(4); (5), (7), (18), (19); (16), (21); The initial problem is decomposed into three sub-problems: trajectory optimization of the monitoring UAV-E, joint optimization of the trajectory of the jamming UAV-J and the diagonal phase shift matrix of RIS, and jamming power optimization of the jamming UAV-J. Finally, the suboptimal solution of the initial problem is obtained by rotating the optimization of the three sub-problems.

6. The method according to claim 5, characterized in that, In step 4, the trajectory optimization of the monitoring UAV-E includes: for a given trajectory of the interfering UAV-J Interference power and diagonal phase shift matrix Problem (25) is rephrased as: ， （1）~（4）；（7），（18），（19）；（26）； （27）； Replace the objective function with or Problem (27) is rewritten as: （28）； （1）~（4）；（7），（18），（19）； First, define the given diagonal phase shift matrix components. This is always optimal, yielding the following result: （29）； Signal-to-interference-plus-noise ratio of UAV-E receiver The lower bound is obtained by averaging over a time slot. As shown below: （30）， in, , It is the lower bound of the small-scale fading between point i and point j; Lower bound of achievable eavesdropping rate Written as: （31）， intermediate parameters intermediate parameters intermediate parameters ; By introducing slack variables and Defined as: （32）， （33）， Given a constant , and In the case of equation Compared to and It is a convex function, for Expand to obtain the lower bound. for: （34）， intermediate parameters ; and It is in the The feasible solution obtained in the next iteration; according to formulas (30)~(34), the problem defined in (28) is approximately transformed into: （35）， （1）~（4）；（7），（18），（19）；（32），（33）。 7. The method according to claim 6, characterized in that, In step 4, the joint optimization of the trajectory of the interfering UAV-J and the diagonal phase shift matrix of RIS includes: given the trajectory of the monitoring UAV-E and interference power In this case, problem (25) is transformed into: （36）， （1）~（4）；（5），（7），（18），（19），（23）； Replace the objective function with Additional weights are added here. This causes the objective function to further change to The new questions that arise are as follows: （37）， （1）~（4）；（5），（7），（18），（19）； Problem (37) can be solved in two parts: the phase shift matrix. Optimization and interference of UAV-J trajectory The optimization, given the disturbed UAV-J trajectory, yields: （38）， By aligning the phase of the signal received by the interfering UAV-J, the energy of the received signal can be maximized. Therefore, the following settings are made: （39）， Or it can be expressed as: （40）， （41）， in ; The following results were obtained: （42）， Then, based on formula (31), when the variable is transformed into the trajectory of the interfering UAV-J, the achievable monitoring rate of the monitoring UAV-E is determined. Represented as: （43）， intermediate parameters intermediate parameters ; Then, slack variables are introduced. and Defined as: （44）， （45）， And on Performing a first-order Taylor expansion yields the lower bound. for: （46）， intermediate parameters ; and It is in the The feasible solution obtained in the next iteration; right The upper bound is obtained by averaging over a time slot. for: （47）， in , It is the upper bound of the small-scale fading between point i and point j; a new slack variable is introduced. Defined as: （48）， Problem (37) is refactored as follows: （49）， （1）~（4）；（5），（7），（18），（19）；（44），（45），（48）。 8. The method according to claim 7, characterized in that, In step 4, the interference power optimization of the interfering UAV-J includes: given the trajectories of the monitoring UAV-E and the interfering UAV-J and the phase shift matrix, problem (25) is rewritten as: （50）， （16），（21），（23）； Derive the optimal interference power exist Obtained at that time, according to The value may fall into one of the following three categories. Further processing is required. , and ; when At that time, the actual optimal interference power was set to ; when At that time, the optimal interference power is set to ; when At that time, the optimal interference power is set to ; Optimal interference power for problem (50) The derivation is as follows: （51）， in: （52）， （53）， （54）， The energy efficiency calculation will be reset to: （55）。 9. An electronic device, characterized in that, It includes a processor and a memory, the memory storing program code that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 8.

10. A storage medium, characterized in that, It stores a computer program or instructions that, when run on a computer, perform the steps of the method as described in any one of claims 1 to 8.

Images