An unmanned aerial vehicle relay wireless energy transmission and anti-interference transmission method, system and device based on semantic assistance and a storage medium

Abstract

This invention discloses a semantically assisted UAV relay wireless power transmission and anti-interference transmission method, system, device, and storage medium, belonging to the field of information and communication technology. The method includes acquiring communication data from a target dual-hop relay air-to-ground communication system; calculating the semantic communication rate of ground users based on a semantic communication mechanism; constructing a joint optimization model based on the semantic communication rate of ground users; robustly handling non-ideal channel errors in the joint optimization model, and then transforming the problem of solving the optimization objective into a second-form problem; iteratively optimizing the initial solution space of the second-form problem using a monotonic optimization algorithm until the difference between the upper and lower bounds of the objective function value corresponding to the global optimal solution of the second-form problem is less than a preset threshold, thus obtaining the optimization result of the optimization variables. This invention fully leverages the inherent robustness of semantic communication under interference, and can simultaneously cope with malicious interference and energy constraints.

Description

Technical Field

[0001] This invention belongs to the field of information and communication technology, and specifically relates to a method, system, device and storage medium for wireless power transmission and anti-interference transmission of unmanned aerial vehicles based on semantic assistance. Background Technology

[0002] With the rapid development of 6G mobile communication and the low-altitude economy, unmanned aerial vehicles (UAVs) have become an important component of integrated air-to-ground communication networks due to their flexible deployment, superior line-of-sight transmission capabilities, and rapid coverage. By carrying communication relay equipment, UAVs play a crucial role in emergency communication, network enhancement, and coverage in remote areas. Currently, a large amount of research focuses on UAV communication, aiming to improve its communication efficiency and resource scheduling capabilities from dimensions such as relay modes, trajectory planning, and power control. In particular, UAVs, as airborne relays, can overcome the shielding effect of ground obstacles, effectively improving the connection performance of ground users. To improve the capacity and reliability of UAV air links, Multiple Input Multiple Output (MIMO) technology is widely used. By deploying multiple antenna arrays at the UAV end, leveraging their advantages in spatial multiplexing, beamforming, and interference suppression, MIMO technology can support parallel transmission by multiple users, concentrate signal energy, and effectively suppress interference and malicious interference between multiple users, thereby significantly improving system spectral efficiency and anti-interference capabilities. Furthermore, MIMO technology can be combined with physical layer security technologies to enhance the anti-eavesdropping and anti-interference performance of communication links while increasing capacity.

[0003] However, the introduction of MIMO systems has also significantly increased the energy consumption of UAV communication equipment, contradicting the limited battery capacity of traditional UAVs. To alleviate this energy bottleneck, Simultaneous Wireless Information and Power Transfer (SWIPT) technology has become an important research direction. Through power splitting mechanisms, UAVs can simultaneously perform information decoding and energy harvesting while receiving signals, achieving coupled utilization of energy and information. Existing research has improved UAV endurance by combining UAV trajectory optimization with SWIPT. On the other hand, the traditional communication paradigm suffers from information efficiency bottlenecks, giving rise to a new paradigm centered on "conveying the meaning of information"—semantic communication. Semantic communication significantly reduces transmission overhead and improves effective information transmission efficiency under limited bandwidth and energy conditions by extracting and transmitting core semantic features relevant to the task, rather than redundant bit streams. In UAV networks, existing research has used it to reduce transmission latency and optimize resource allocation.

[0004] Despite significant progress in key technologies such as UAV relay communication, MIMO, SWIPT, and semantic communication, existing technologies still suffer from systemic defects and insufficient integration when facing efficient, reliable, and secure air-to-ground integrated transmission scenarios.

[0005] First, while MIMO technology can improve capacity and anti-interference capabilities, the increased overhead of multiple radio frequency links and signal processing significantly exacerbates the already limited energy burden of UAVs, restricting their application in long-duration, high-load missions. Second, SWIPT technology, aimed at alleviating energy bottlenecks, is typically based on traditional bit-level communication models, neglecting the semantic content of information. In bandwidth-constrained or heavily interfered scenarios, it still faces the risks of low information efficiency and resource waste. While emerging semantic communication paradigms can improve efficiency and robustness by transmitting the core semantics of the mission, existing research largely focuses on ideal channel environments, lacking a systematic framework for deep joint optimization with SWIPT energy harvesting and MIMO anti-interference beamforming under energy constraints and interference threats. Especially in highly adversarial environments, most existing solutions still prioritize traditional communication rates, failing to fully leverage the inherent robustness of semantic communication under interference. They also lack a complete resource scheduling and transmission mechanism centered on the coordinated optimization of semantic transmission quality and energy efficiency, capable of simultaneously addressing malicious interference and energy constraints. This makes it difficult for systems to achieve secure, efficient, and sustainable semantic information relay in complex electromagnetic environments. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the prior art and provide a semantically assisted UAV relay wireless power transmission and anti-interference transmission method, system, device and storage medium, which fully leverages the inherent robustness of semantic communication under interference and can simultaneously cope with malicious interference and energy constraints.

[0007] This invention provides the following technical method:

[0008] In the first aspect, a semantically assisted method for UAV relay wireless power transmission and anti-jamming transmission is provided, including: acquiring communication data of a target dual-hop relay air-to-ground communication system, the communication data including communication data between the ground transmitter and the relay UAV, communication data between the relay UAV and the ground user, and jamming data between the jammer and the ground user;

[0009] Based on the communication data of the target dual-hop relay air-to-ground communication system, the semantic communication rate of the ground user is calculated based on the semantic communication mechanism.

[0010] Based on the semantic communication rate of ground users, a joint optimization model is constructed with the goal of maximizing the sum of the semantic communication rates of ground users and with the power splitting ratio of the relay UAV decoding, the transmitting precoding vector of the relay UAV and the receiving precoding vector of the ground users as optimization variables.

[0011] After robustly handling the non-ideal channel error in the joint optimization model, the problem of solving the optimization objective is transformed into a second-form problem.

[0012] The initial solution space of the second form problem is iteratively optimized using a monotonic optimization algorithm until the difference between the upper and lower bounds of the objective function value corresponding to the global optimal solution of the second form problem is less than a preset threshold, thus obtaining the optimization result of the optimization variable. In each iteration, a feasibility verification algorithm is used to verify the feasibility of any region in the initial solution space, and infeasible regions are removed from the initial solution space.

[0013] As an optional technical method of the present invention, the step of calculating the semantic communication rate of the ground user based on the communication data of the target dual-hop relay air-to-ground communication system and the semantic communication mechanism includes:

[0014] The ground transmitter semantically encodes the mission information and sends it to the relay UAV. The received signal from the relay UAV is represented as follows:

[0015] ;

[0016] in, This indicates the received signal of the relay drone. Indicates the power of the transmitting node. This represents the transmission precoding vector of the ground transmitter. This represents the received precoding vector of the relay drone. This represents the additive white Gaussian noise at the relay drone location. This represents the channel vector from the ground transmitter to the relay UAV. This indicates the semantic transmission signal of the ground transmitter. Indicates conjugate transpose;

[0017] The relay drone harvests energy from the received signal. The power harvested by the relay drone is expressed as:

[0018] ;

[0019] ;

[0020] ;

[0021] in, This indicates the power collected by the relay drone. This represents the logistic function related to the input power of the relay UAV's energy harvesting. This represents the maximum power collected when the energy harvesting circuit is saturated. Used to ensure response under zero input or zero output conditions during energy harvesting. and These all represent constants related to resistance and other parameters in a circuit. This indicates the input power for energy harvesting;

[0022] The relay drone performs semantic decoding on the received signal, and calculates the semantic information decoded signal of the relay drone, which is represented as:

[0023] ;

[0024] in, This represents the semantic information decoding signal of the relay drone. This indicates the power splitting ratio for decoding by the relay drone. Indicates noise processing;

[0025] The semantic communication rate of the relay UAV is calculated and expressed as:

[0026] ;

[0027] ;

[0028] ;

[0029] in, This indicates the semantic communication rate of the relay drone. This indicates the bandwidth from the ground transmitter to the drone. This represents the average semantic information of each sentence in the signal received by the relay drone. This represents the average number of semantic symbols required for a relay drone to receive each word in the signal. This represents the average number of words per sentence received by the relay drone. This indicates the semantic similarity between the original sentence and semantic information in the signal received by the relay drone. Indicates the semantic received signal-to-noise ratio. Indicates the left asymptote. Indicates the right asymptote. Represents the logical growth rate. Indicates the center point. , Indicates the power used to process noise. This represents the power of the additive white Gaussian noise at the relay drone. Denotes the Euclidean norm;

[0030] Calculate the first The received signal of a ground user is represented as follows:

[0031] ;

[0032] in, Indicates the first The received signal of each ground user Indicates the first Received precoding for each ground user Indicates the first Noise at ground user locations Indicates the relay drone to the Channel matrix for each ground user This represents the total number of ground users. Indicates that the relay drone is for the first The precoded vector sent by each ground user Indicates the jammer to the number Channel matrix for each ground user This indicates the jamming signal sent by the jammer. This indicates that the relay drone will re-encode the decoded semantic information into a sequence representing the first... Semantic transmission signals for each ground user;

[0033] Calculate the first The semantic communication rate of a ground user is expressed as:

[0034] ;

[0035] ;

[0036] ;

[0037] in, Indicates the first Semantic communication rate of each ground user Indicates the relay drone to the Bandwidth for each ground user Indicates the first The average semantic information of each sentence in the signal received by a ground user. Indicates the first The average number of semantic symbols required per word for a ground user to receive the signal. Indicates the first The average number of words per sentence received by a ground user. Indicates the first Semantic similarity between the original sentence and semantic information in the signals received by each ground user Indicates the first Semantic reception signal-to-noise ratio for each ground user Indicates that the relay drone is for the first The precoded vector sent by each ground user This indicates the jammer's transmission power. Indicates the first Noise power at each ground user location.

[0038] As an optional technical method of the present invention, the joint optimization model is expressed as follows:

[0039] ;

[0040] in, Indicates that the relay drone is for the first The precoded vector sent by each ground user Indicates the first The received precoding vectors of each ground user This indicates the power splitting ratio for decoding by the relay drone. This indicates that through optimization , and To obtain the maximum value, Indicates the first Semantic communication rate of each ground user This represents the total number of ground users. This indicates a non-ideal channel error. This indicates that through optimization To obtain the minimum value, This represents the semantic communication rate threshold for relay drones. Indicates and the The semantic communication rate threshold for each ground user This indicates the semantic communication rate of the relay drone. This indicates the power consumption required for the relay drone to fly. This indicates the power collected by the relay drone. This represents the semantic communication rate constraint for relay drones. This represents the semantic communication rate constraint for ground users. This indicates the power overhead constraint of the relay drone. This represents the normalization constraint of the received precoding vector. Indicates the power splitting ratio constraint;

[0041] The non-ideal channel error Represented as:

[0042] ;

[0043] in, Indicates the jammer to the number Channel matrix for each ground user Indicates the jammer to the number The departure elevation angle of a ground user Indicates the jammer to the number The departure azimuth angle of a ground user Indicates the jammer to the number The minimum departure elevation angle for a ground user. Indicates the jammer to the number The maximum departure elevation angle for a ground user. Indicates the jammer to the number The minimum departure azimuth angle for each ground user. Indicates the jammer to the number The maximum value of the departure azimuth angle for each ground user.

[0044] As an optional technical method of the present invention, after robustly processing the non-ideal channel error in the joint optimization model, the problem of solving the optimization objective is transformed into a second-form problem, including:

[0045] After robustly handling the non-ideal channel error in the joint optimization model, the problem of solving the optimization objective is transformed into a first-form problem, expressed as:

[0046] ;

[0047] Transforming the problem in the first form into the problem in the second form, it can be expressed as:

[0048] ;

[0049] ;

[0050] ;

[0051] in, Represents the new set of optimization variables. Let k be the new optimization variable. Represents the old set of optimization variables. This indicates that through optimization To obtain the maximum value, This indicates the first [unclear] under the old set of optimization variables. Semantic communication rate of each ground user This represents the feasible region of the new optimization variable. This represents the feasible region of the old optimization variables. Represents the new set of optimization variables The objective function value is as follows;

[0052] The set of old optimization variables corresponding to the optimal solution of the new set of optimization variables in the second form of solving the problem is the unique solution of the first form of solving the problem.

[0053] As an optional technical method of the present invention, the step of iteratively optimizing the initial solution space of the second-form problem using a monotonic optimization algorithm until the difference between the upper and lower bounds of the objective function value corresponding to the global optimal solution of the second-form problem is less than a preset threshold, and obtaining the optimization result of the optimization variable, includes:

[0054] For the second form of the problem, let , Construct the initial solution space ,in This represents the lower bound of the initial solution space. This represents the upper bound of the initial solution space. This represents the minimum value of the new set of optimization variables. This represents the maximum value of the new set of optimization variables;

[0055] Randomly select any region space in the initial solution space A feasibility verification algorithm was used to verify the spatial region. The feasibility, if the regional space If not feasible, then the regional space Remove from the initial solution space;

[0056] If the regional space It is feasible to use a binary search method to connect the lines. Find the regional space Intersection vector with Pareto boundary and the interval of intersection connecting wire Represented as:

[0057] ;

[0058] in, Representing regional space The lower bound, Representing regional space The upper realm, This represents the lower bound of a feasible solution for the interval of intersection points. This indicates an upper bound for feasible solutions in the interval of intersection points. This represents an intermediate variable used when searching for Pareto boundary intersections on the connector. Indicates the lower bound The objective function value at that point, Indicates the normalization factor;

[0059] Update the lower bound of the objective function value corresponding to the global optimal solution to , Indicates a feasible solution based on the lower bound. The calculated objective function value;

[0060] regional space Perform orthogonal partitioning to obtain There are three non-overlapping subspaces, and the upper and lower vertices of each subspace are represented as follows:

[0061] ;

[0062] ;

[0063] in, This represents the upper bound vertex vector of the m-th subspace obtained after partitioning the current region space in the n-th iteration. This represents the lower bound vertex vector of the m-th subspace obtained by partitioning the current region space in the n-th iteration. Indicates only the first A unit vector with 1 element and the rest being 0. Represents the intersection vector The coordinates in the m-th dimension This represents the result obtained by partitioning the current region space in the nth iteration. The lower bound vertex vector of each subspace. Indicates only the first A unit vector with 1 element and 0 elements;

[0064] Regional space After partitioning, a new subspace is obtained. Furthermore, the subspaces are orthogonal and do not overlap. The updated new region space is represented as follows:

[0065] ;

[0066] in, Indicates a new regional space;

[0067] For new regional space Remove invalid regions that are below their lower bound;

[0068] For new regional space The objective function value at the upper bound ,like This will create a new regional space. Remove; if Then update the new area space. The lower bound is represented as:

[0069] ;

[0070] in, Indicates the updated new area space The lower bound of the value is denoted by min, which represents the minimum value. Indicates the updated new area space The upper realm, This represents the lower bound of the objective function value corresponding to the global optimal solution.

[0071] As an optional technical method of the present invention, the step of verifying the feasibility of any region in the initial solution space using a feasibility verification algorithm includes:

[0072] Verification area space Does a feasible solution exist for the new set of optimization variables? Feasible solution Satisfying the sixth constraint ;

[0073] Transform the problem from the first form of solution into the third form of solution:

[0074] ;

[0075] For the Received precoding vectors for each ground user The solution is obtained using the linear minimum mean square error algorithm, expressed as:

[0076] ;

[0077] ;

[0078] in, Indicates that the relay drone is for the first The precoded vector sent by each ground user Represents the identity matrix. Indicates the transition amount;

[0079] For relay drones to the first Precoded vectors sent to each ground user ,like Then the relay drone to the first Precoded vectors sent to each ground user It is a trivial stationary point in the third form of problem-solving; let Introducing the first auxiliary variable ,satisfy The problem is transformed from a third-form problem to a fourth-form problem:

[0080] ;

[0081] in, Indicates using replace Semantic communication rate constraints for subsequent ground users Indicates using replace Power overhead constraints for subsequent relay drones Indicates using replace The sixth constraint after;

[0082] Semantic communication rate constraints for relay drones Simplifying, we get:

[0083] ;

[0084] in, Indicates the power used to process noise. This represents the power of the additive white Gaussian noise at the relay drone. Indicates the power of the transmitting node. This represents the transmission precoding vector of the ground transmitter. This represents the received precoding vector of the relay drone. This represents the additive white Gaussian noise at the relay drone location. This represents the channel vector from the ground transmitter to the relay UAV. This indicates the semantic transmission signal of the ground transmitter. , Indicates the left asymptote. Indicates the right asymptote. Represents the logical growth rate. Indicates the center point. Indicates the relay drone to the Bandwidth for each ground user Indicates the first The average semantic information of each sentence in the signal received by a ground user. Indicates the first The average number of semantic symbols required per word for a ground user to receive the signal. Indicates the first The average number of words per sentence received by a ground user;

[0085] By combining the power splitting ratio constraint C5, the power splitting ratio for relay UAV decoding is obtained. The constraint C7 is represented as:

[0086] ;

[0087] will use replace Semantic communication rate constraints for subsequent ground users and use replace The sixth constraint after By merging, we obtain the merged expression:

[0088] ;

[0089] in, , Indicates the first The first auxiliary variable corresponding to each ground user ;

[0090] Introducing a second auxiliary variable Third auxiliary variable Fourth auxiliary variable Fifth auxiliary variable The merged expression is transformed into convex form, with the eighth constraint C8, ninth constraint C9, tenth constraint C10, eleventh constraint C11, and twelfth constraint C12 represented as follows:

[0091] ;

[0092] ;

[0093] ;

[0094] ;

[0095] ;

[0096] in, Describes the Euclidean norm. Indicates transpose. Represents the fourth auxiliary variable The The value of the next iteration. Represents the fifth auxiliary variable The The value of the next iteration, max indicates taking the maximum value;

[0097] Introducing the first slack variable Second relaxation variable , will use replace Power overhead constraints of subsequent relay drones The thirteenth constraint C13, the fourteenth constraint C14, and the fifteenth constraint C15, converted to convex form, are expressed as follows:

[0098] ;

[0099] ;

[0100] ;

[0101] in, and These all represent constants related to resistance and other parameters in a circuit. Indicates energy harvesting efficiency. , Represents the first slack variable The The value of the next iteration;

[0102] The fourth form problem is transformed into a convex optimization problem, expressed as:

[0103] ;

[0104] Solve the convex optimization problem to obtain feasible solutions in the domain space, and use the feasible solution obtained in the last iteration as the optimization result for the optimization variables.

[0105] Secondly, a semantically assisted UAV relay wireless power transmission and anti-jamming transmission system is provided, including: an acquisition module for acquiring communication data of a target dual-hop relay air-to-ground communication system, the communication data including communication data between the ground transmitter and the relay UAV, communication data between the relay UAV and the ground user, and interference data between the jammer and the ground user;

[0106] The rate calculation module is used to calculate the semantic communication rate of ground users based on the communication data of the target dual-hop relay air-to-ground communication system and the semantic communication mechanism.

[0107] The model building module is used to construct a joint optimization model based on the semantic communication rate of ground users, with the optimization objective being to maximize the sum of the semantic communication rates of ground users and the optimization variables being the power splitting ratio of the relay UAV decoding, the transmitting precoding vector of the relay UAV, and the receiving precoding vector of the ground users.

[0108] The conversion module is used to robustly process the non-ideal channel error in the joint optimization model and then convert the problem of solving the optimization objective into a second-form problem.

[0109] The solution module is used to iteratively optimize the initial solution space of the second form problem using a monotonic optimization algorithm until the difference between the upper and lower bounds of the objective function value corresponding to the global optimal solution of the second form problem is less than a preset threshold, thus obtaining the optimization result of the optimization variable. In each iteration, a feasibility verification algorithm is used to verify the feasibility of any region in the initial solution space, and infeasible regions are removed from the initial solution space.

[0110] Thirdly, a semantically assisted UAV relay wireless power transmission and anti-interference transmission device is provided, comprising a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the semantically assisted UAV relay wireless power transmission and anti-interference transmission method described in the first aspect.

[0111] Fourthly, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the semantically assisted UAV relay wireless power transmission and anti-interference transmission method described in the first aspect.

[0112] Compared with the prior art, the beneficial effects of the present invention are:

[0113] This invention provides a semantically assisted UAV relay wireless power transmission and anti-interference transmission method. It introduces a semantic communication mechanism into the communication link and combines it with power splitting to achieve coordinated optimization of information transmission and energy harvesting. By jointly optimizing the power splitting ratio of the relay UAV decoding, the transmitting precoding vector of the relay UAV and the receiving precoding vector of the ground user, the sum of the semantic communication rate of the ground user is maximized. This fully leverages the inherent robustness of semantic communication under interference and can simultaneously cope with malicious interference and energy constraints. Attached Figure Description

[0114] Figure 1 This is a flowchart of a semantically assisted UAV relay wireless power transmission and anti-interference transmission method in an embodiment of the present invention.

[0115] Figure 2 This is a schematic diagram of a dual-hop relay air-to-ground communication system model in an embodiment of the present invention;

[0116] Figure 3 This is a convergence performance graph comparing the method of this invention with other methods in this embodiment of the invention;

[0117] Figure 4 This is a performance comparison chart between the method described in this embodiment of the invention and other methods. Detailed Implementation

[0118] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical methods of the present invention, and should not be construed as limiting the scope of protection of the present invention.

[0119] Example 1

[0120] This embodiment provides a semantically assisted UAV relay wireless power transmission and anti-interference transmission method, such as... Figure 1 As shown, the specific steps include the following:

[0121] Step 1: Acquire the communication data of the target dual-hop relay air-to-ground communication system. The communication data includes the communication data between the ground transmitter and the relay UAV, the communication data between the relay UAV and the ground user, and the jamming data between the jammer and the ground user.

[0122] like Figure 2 As shown, this embodiment considers a system consisting of a ground transmitter, a relay drone, and... A dual-hop relay air-to-ground communication system consisting of one ground user and one enemy-made jamming device. Ground transmitter, relay UAV, and... Each ground user is equipped with a linear surface antenna, the number of which are as follows: , and The jammer transmits jamming signals to all ground users, interfering with their normal signal reception. Each terminal node in the system is equipped with a semantic codec.

[0123] Step 2: Calculate the semantic communication rate of the ground user based on the semantic communication mechanism, according to the communication data of the target dual-hop relay air-to-ground communication system.

[0124] In the first phase, the ground transmitter semantically encodes the mission information and sends it to the relay UAV. The relay UAV uses the SWIPT mechanism to perform energy harvesting and semantic decoding on the received signal, including power splitting ratio. Used for receiving and decoding semantic signals. The ratio is used for energy harvesting.

[0125] In the first phase, the ground transmitter will transmit information Encoded as semantic transmission signals And send it to the relay drone, and meet the following requirements. The received signal of the relay drone is represented as:

[0126] ;

[0127] in, This indicates the received signal of the relay drone. Indicates the power of the transmitting node. This represents the transmission precoding vector of the ground transmitter. This represents the received precoding vector of the relay drone. This represents the additive white Gaussian noise at the relay drone location. This represents the channel vector from the ground transmitter to the relay UAV. This indicates the semantic transmission signal of the ground transmitter. This indicates the conjugate transpose.

[0128] The relay drone harvests energy from the received signal. The power harvested by the relay drone is expressed as:

[0129] ;

[0130] ;

[0131] ;

[0132] in, This indicates the power collected by the relay drone. This represents the logistic function related to the input power of the relay UAV's energy harvesting. This represents the maximum power collected when the energy harvesting circuit is saturated. Used to ensure response under zero input or zero output conditions during energy harvesting. and These all represent constants related to resistance and other parameters in a circuit. This indicates the input power for energy harvesting. , This indicates the energy harvesting efficiency.

[0133] The relay drone performs semantic decoding on the received signal, and calculates the semantic information decoded signal of the relay drone, which is represented as:

[0134] ;

[0135] in, This represents the semantic information decoding signal of the relay drone. This indicates the power splitting ratio for decoding by the relay drone. This indicates noise processing.

[0136] Unlike traditional bit transmission, semantic transmission uses semantic symbols to convey semantic information. The semantic communication rate of a relay drone can be calculated as follows:

[0137] ;

[0138] ;

[0139] Semantic similarity exhibits an S-shaped curve and monotonically increases with signal-to-noise ratio, depending on... and Its expression is:

[0140] ;

[0141] in, This indicates the semantic communication rate of the relay drone. This indicates the bandwidth from the ground transmitter to the drone. This represents the average semantic information of each sentence in the signal received by the relay drone. This represents the average number of semantic symbols required for a relay drone to receive each word in the signal. This represents the average number of words per sentence received by the relay drone. This indicates the semantic similarity between the original sentence and semantic information in the signal received by the relay drone. Indicates the semantic received signal-to-noise ratio. Indicates the left asymptote. Indicates the right asymptote. Represents the logical growth rate. Indicates the center point. , Indicates the power used to process noise. This represents the power of the additive white Gaussian noise at the relay drone. This represents the Euclidean norm.

[0142] The energy harvested by the drone in the first phase is then converted into DC power to support the relay process in the second phase. In the second phase, the relay drone re-encodes the decoded semantic information and uses beamforming technology to transmit the information to multiple ground users. The relay drone re-encodes the decoded semantic information into... In information Before launch, precoding and weighting are required. The transmission channel of the relay UAV is represented as follows: Since the total transmit power of the UAV is obtained from the energy harvesting in the first stage, it satisfies:

[0143] .

[0144] Calculate the first The received signal of a ground user is represented as follows:

[0145] ;

[0146] in, Indicates the first The received signal of each ground user Indicates the first Received precoding for each ground user Indicates the first Noise at ground user locations Indicates the relay drone to the Channel matrix for each ground user This represents the total number of ground users. Indicates that the relay drone is for the first The precoded vector sent by each ground user Indicates the jammer to the number Channel matrix for each ground user This indicates the jamming signal sent by the jammer. , Indicates the normalized interference signal. This indicates that the relay drone will re-encode the decoded semantic information into a sequence representing the first... Semantic transmission signals for each ground user.

[0147] Calculate the first The semantic communication rate of a ground user is expressed as:

[0148] ;

[0149] ;

[0150] ;

[0151] in, Indicates the first Semantic communication rate of each ground user Indicates the relay drone to the Bandwidth for each ground user Indicates the first The average semantic information of each sentence in the signal received by a ground user. Indicates the first The average number of semantic symbols required per word for a ground user to receive the signal. Indicates the first The average number of words per sentence received by a ground user. Indicates the first Semantic similarity between the original sentence and semantic information in the signals received by each ground user Indicates the first Semantic reception signal-to-noise ratio for each ground user Indicates that the relay drone is for the first The precoded vector sent by each ground user This indicates the jammer's transmission power. Indicates the first Noise power at each ground user location.

[0152] Step 3: Based on the semantic communication rate of ground users, construct a joint optimization model with the goal of maximizing the sum of semantic communication rates of ground users and with the power splitting ratio of relay UAV decoding, the transmitting precoding vector of relay UAV and the receiving precoding vector of ground users as optimization variables.

[0153] The joint optimization model is expressed as follows:

[0154] ;

[0155] in, Indicates that the relay drone is for the first The precoded vector sent by each ground user Indicates the first The received precoding vectors of each ground user This indicates the power splitting ratio for decoding by the relay drone. This indicates that through optimization , and To obtain the maximum value, Indicates the first Semantic communication rate of each ground user This represents the total number of ground users. This indicates a non-ideal channel error. This indicates that through optimization To obtain the minimum value, This represents the semantic communication rate threshold for relay drones. Indicates and the The semantic communication rate threshold for each ground user This indicates the semantic communication rate of the relay drone. This indicates the power consumption required for the relay drone to fly. This indicates the power collected by the relay drone. This represents the semantic communication rate constraint for relay drones. This represents the semantic communication rate constraint for ground users. This indicates the power overhead constraint of the relay drone. This represents the normalization constraint of the received precoding vector. This indicates a power splitting ratio constraint.

[0156] The semantic communication rate of each user in the objective function depends on a non-convex fractional expression, which is simultaneously coupled with the UAV transmit beam vector and the user receive beam vector, resulting in a highly coupled non-convex structure among the optimization variables. Secondly, the energy harvesting constraint tightly links the power allocation in the first stage with the beam design in the second stage, further increasing the cross-stage coupling of the optimization problem. Furthermore, the uncertainty error of the non-ideal channel introduces infinitely non-convex characteristics to the objective function and constraints.

[0157] For information about legitimate communication links in an air-to-ground cooperative network, the transceiver can estimate and predict channel state information by sending pilot signals. Therefore, the channel between the relay UAV and the transceiver user can be accurately obtained, including... and However, because the jammer and the legitimate user are not in a cooperative relationship, information about the jamming channel is difficult to obtain through estimation or prediction. Measuring the jamming signal can roughly determine the relative position between the jammer and the legitimate transmitter, but there is a certain estimation error. This is due to non-ideal channel errors. Represented as:

[0158] ;

[0159] in, Indicates the jammer to the number Channel matrix for each ground user Indicates the jammer to the number The departure elevation angle of a ground user Indicates the jammer to the number The departure azimuth angle of a ground user Indicates the jammer to the number The minimum departure elevation angle for a ground user. Indicates the jammer to the number The maximum departure elevation angle for a ground user. Indicates the jammer to the number The minimum departure azimuth angle for each ground user. Indicates the jammer to the number The maximum value of the departure azimuth angle for each ground user.

[0160] Step 4: After robustly processing the non-ideal channel error in the joint optimization model, the problem of solving the optimization objective is transformed into a second-form problem.

[0161] Robust processing of imperfect channel state information is employed to address infinite non-convex uncertainties. Firstly, a generalized discretization method is used to process the non-ideal channel information. and Uniform sampling within the error range is represented as follows:

[0162] ;

[0163] ;

[0164] in, Indicates the first Each sampling elevation angle Indicates the first Each sampling azimuth angle. This represents the total number of sampling points indicating the elevation angle. The number of sampling points indicating the azimuth angle. The sampling interval represents the elevation angle. This indicates the sampling interval for the azimuth angle.

[0165] jammer and the first The worst-case robust channel representation among ground users is:

[0166] ;

[0167] in, , Indicates the discrete angle The channel response vector at that location, Indicates the jammer and the first Robust channel between ground users.

[0168] After robustly handling the non-ideal channel error in the joint optimization model, the problem of solving the optimization objective is transformed into a first-form problem, expressed as:

[0169] ;

[0170] Since the problem in the first form exhibits an S-shaped curve, conventional convex optimization methods cannot solve it. The problem in the first form is transformed into a problem in the second form, expressed as:

[0171] ;

[0172] ;

[0173] ;

[0174] in, Represents the new set of optimization variables. Let k be the new optimization variable. Represents the old set of optimization variables. This indicates that through optimization To obtain the maximum value, This indicates the first [unclear] under the old set of optimization variables. Semantic communication rate of each ground user This represents the feasible region of the new optimization variable. This represents the feasible region of the old optimization variables. Represents the new set of optimization variables The objective function value is given below.

[0175] The set of old optimization variables corresponding to the optimal solution of the new set of optimization variables in the second form of solving the problem is the unique solution of the first form of solving the problem.

[0176] because for It is a monotonically increasing function, and the optimal one. It must exist on the upper boundary. This can be obtained by solving the following K equations. :

[0177] ;

[0178] To find a solution In For example, we need to solve K linear equations:

[0179] ;

[0180] in, Represented as:

[0181] .

[0182] In the above equation, the channel is only related to location; therefore, the K linear equations are independent and linear, and a unique set of solutions can be obtained. For the other two variables and This can also be proven by the same reasoning. Therefore corresponding It is unique. By solving new problems. That way you can find the problem. The solution.

[0183] Step 5: Iteratively optimize the initial solution space of the second form problem using a monotonic optimization algorithm until the difference between the upper and lower bounds of the objective function value corresponding to the global optimal solution of the second form problem is less than a preset threshold, and obtain the optimization result of the optimization variable.

[0184] For the second form of the problem, let , Construct the initial solution space ,in This represents the lower bound of the initial solution space. This represents the upper bound of the initial solution space. This represents the minimum value of the new set of optimization variables. This represents the maximum value of the new set of optimization variables. For the lower bound... Due to constraint C2, the minimum semantic transmission rate for each ground user is expressed as: For the upper realm ,by For example, its relationship with signal-to-interference-plus-noise ratio Proportional, therefore when the drone concentrates all its power on the first The maximum signal-to-noise ratio can be obtained when considering only one ground user and ignoring signals from malicious jammers. .

[0185] Randomly select any region space in the initial solution space A feasibility verification algorithm was used to verify the spatial region. The feasibility, if the regional space If not feasible, then the regional space Remove from the initial solution space.

[0186] If the regional space It is feasible to use a binary search method to connect the lines. Find the regional space Intersection vector with Pareto boundary and the interval of intersection connecting wire Represented as:

[0187] ;

[0188] in, Representing regional space The lower bound, Representing regional space The upper realm, This represents the lower bound of a feasible solution for the interval of intersection points. This indicates an upper bound for feasible solutions in the interval of intersection points. This represents an intermediate variable used when searching for Pareto boundary intersections on the connector. Indicates the lower bound The objective function value at that point, This represents the normalization factor.

[0189] Update the lower bound of the objective function value corresponding to the global optimal solution to , Indicates a feasible solution based on the lower bound. The calculated objective function value.

[0190] regional space Perform orthogonal partitioning to obtain There are three non-overlapping subspaces, and the upper and lower vertices of each subspace are represented as follows:

[0191] ;

[0192] ;

[0193] in, This represents the upper bound vertex vector of the m-th subspace obtained after partitioning the current region space in the n-th iteration. This represents the lower bound vertex vector of the m-th subspace obtained by partitioning the current region space in the n-th iteration. Indicates only the first A unit vector with 1 element and the rest being 0. Represents the intersection vector The coordinates in the m-th dimension This represents the result obtained by partitioning the current region space in the nth iteration. The lower bound vertex vector of each subspace. Indicates only the first A unit vector with one element being 1 and the rest being 0.

[0194] Regional space After partitioning, a new subspace is obtained. Furthermore, the subspaces are orthogonal and do not overlap. The updated new region space is represented as follows:

[0195] ;

[0196] in, Indicates a new regional space;

[0197] For new regional space Remove invalid regions that are below their lower bound;

[0198] For new regional space The objective function value at the upper bound ,like This will create a new regional space. Remove; if Then update the new area space. The lower bound is represented as:

[0199] ;

[0200] in, Indicates the updated new area space The lower bound of the value is denoted by min, which represents the minimum value. Indicates the updated new area space The upper realm, This represents the lower bound of the objective function value corresponding to the global optimal solution.

[0201] In any iteration, if If this indicates that the best point in the subspace is worse than the currently known feasible point, then the optimization result obtained through the feasibility verification algorithm will be used as the final optimization result. The upper bound of the objective function value corresponding to the global optimum indicates that the subspace may contain a better solution, and thus the subspace is updated. The lower bound is used for region shrinkage. This indicates a preset threshold.

[0202] In each iteration, a feasibility verification algorithm is used to verify the feasibility of any region in the initial solution space, and infeasible regions are removed from the initial solution space. Specifically, this includes:

[0203] Verification area space Does a feasible solution exist for the new set of optimization variables? Feasible solution The sixth constraint must be satisfied. .

[0204] Transform the problem from the first form of solution into the third form of solution:

[0205] ;

[0206] For the Received precoding vectors for each ground user The solution is obtained using the linear minimum mean square error algorithm, expressed as:

[0207] ;

[0208] ;

[0209] in, Indicates that the relay drone is for the first The precoded vector sent by each ground user Represents the identity matrix. Indicates a transitional quantity.

[0210] For relay drones to the first Precoded vectors sent to each ground user ,like Then the relay drone to the first Precoded vectors sent to each ground user It is a trivial stationary point in the third form of the problem-solving process. Let ,in According to the properties of low-dimensional subspaces, for any non-stationary point... It must be located in Within the vector space, that is, satisfying , Therefore, optimize the variables. The dimension is And optimization The dimension of the variable is Since the number of transmitting antennas is usually greater than the total number of user antennas, that is... Introducing the first auxiliary variable ,satisfy The problem is transformed from a third-form problem to a fourth-form problem:

[0211] ;

[0212] in, Indicates using replace Semantic communication rate constraints for subsequent ground users Indicates using replace Power overhead constraints for subsequent relay drones Indicates using replace The sixth constraint.

[0213] Semantic communication rate constraints for relay drones Simplifying, we get:

[0214] ;

[0215] in, Indicates the power used to process noise. This represents the power of the additive white Gaussian noise at the relay drone. Indicates the power of the transmitting node. This represents the transmission precoding vector of the ground transmitter. This represents the received precoding vector of the relay drone. This represents the additive white Gaussian noise at the relay drone location. This represents the channel vector from the ground transmitter to the relay UAV. This indicates the semantic transmission signal of the ground transmitter. , Indicates the left asymptote. Indicates the right asymptote. Represents the logical growth rate. Indicates the center point. Indicates the relay drone to the Bandwidth for each ground user Indicates the first The average semantic information of each sentence in the signal received by a ground user. Indicates the first The average number of semantic symbols required per word for a ground user to receive the signal. Indicates the first The average number of words per sentence received by a ground user.

[0216] By combining the power splitting ratio constraint C5, the power splitting ratio for relay UAV decoding is obtained. The constraint C7 is represented as:

[0217] ;

[0218] will use replace Semantic communication rate constraints for subsequent ground users and use replace The sixth constraint after By merging, we obtain the merged expression:

[0219] ;

[0220] in, , Indicates the first The first auxiliary variable corresponding to each ground user .

[0221] Introducing a second auxiliary variable Third auxiliary variable Scaling the merged expression, it is represented as:

[0222] (1);

[0223] (2);

[0224] (3);

[0225] Formula (3) is convex, serving as the eighth constraint C8, while formulas (1) and (2) are non-convex. A fourth auxiliary variable is introduced. Fifth auxiliary variable Convert formula (1) to:

[0226] (4);

[0227] (5);

[0228] Equation (4) is transformed into the ninth constraint C9 and the tenth constraint C10, as follows:

[0229] ;

[0230] .

[0231] Through Performing a first-order Taylor expansion, formula (5) is transformed into the eleventh constraint C11, expressed as:

[0232] ;

[0233] make for The The next iteration value, through Performing a first-order Taylor expansion, the lower bound of the expression on the left side of constraint formula (2) is expressed as:

[0234] ;

[0235] ;

[0236] Equation (2) is transformed into the twelfth constraint C12, which is expressed as:

[0237] ;

[0238] in, Describes the Euclidean norm. Indicates transpose. Represents the fourth auxiliary variable The The value of the next iteration. Represents the fifth auxiliary variable The The value of the next iteration, max represents taking the maximum value.

[0239] right After equivalent transformation, we get:

[0240] .

[0241] Introducing the first slack variable Second relaxation variable After scaling, we get:

[0242] ;

[0243] ;

[0244] Thirteenth constraint C13: .

[0245] formula The equivalent derivation yields the fourteenth constraint C14:

[0246] .

[0247] formula The right side is non-convex, so its lower bound can be obtained by performing a first-order Taylor expansion. The transformation yields the fifteenth constraint, C15. ;

[0248] in, and These all represent constants related to resistance and other parameters in a circuit. Indicates energy harvesting efficiency. , Represents the first slack variable The The value of the next iteration.

[0249] The fourth form problem is transformed into a convex optimization problem, expressed as:

[0250] ;

[0251] Solve the convex optimization problem to obtain feasible solutions in the domain space, and use the feasible solution obtained in the last iteration as the optimization result for the optimization variables.

[0252] Example 2

[0253] This embodiment conducted a simulation experiment to verify the effectiveness of the method proposed in Embodiment 1.

[0254] The simulation parameters are set as follows: It consists of one ground transmitter, one aerial UAV, three ground receivers, and one jammer. The UAV's position is set as follows: The three ground receivers are located at elevation and azimuth angles respectively. The distance between the transmitter and the jammer and the drone is 200 meters. The elevation and azimuth angles of the transmitter and jammer relative to the drone are: and The distances are 100 meters and 500 meters respectively. The channel uncertainty between the jammer and the receiver is... The jamming type is full-band jamming, the jamming range is 1000 meters, the jammer's transmitting power is 100W, the transmission bandwidth is 1MHz, and the flight energy consumption is... The average number of characters per word is 5. Energy harvesting efficiency: The noise power is .

[0255] This embodiment uses the following five benchmark algorithms to verify the effectiveness of the proposed method:

[0256] Monotonic optimization scheme without dimensionality reduction: In the proposed system model, a monotonic optimization algorithm including feasibility verification is adopted, but no dimensionality reduction method is used.

[0257] Traditional bit transmission scheme: In the proposed system model, conventional bit communication methods are used for information transmission, without semantic decoding and encoding.

[0258] Branch pruning and delimitation scheme: In the proposed system model, the branch pruning and delimitation algorithm is used to solve the monotonic optimization problem, and SCA is used for feasibility verification.

[0259] Multi-block approximation scheme: In the proposed system model, a multi-block approximation algorithm is used to solve the monotonic optimization problem, and SCA is used for feasibility verification.

[0260] Random power splitting and beamforming scheme: In the proposed system model, the UAV platform randomly splits power and then randomly generates an airborne transmission beam.

[0261] Figure 3 The relationship between semantic transmission rate and the number of UAV antennas is demonstrated. As the array size increases, the semantic rate continuously improves, indicating that larger arrays provide higher spatial resolution and array gain, thereby achieving more accurate main lobe pointing and stronger interference suppression capabilities. Simultaneously, the proposed scheme achieves nearly the same performance as traditional non-dimensionality reduction optimization schemes while reducing the dimensionality of optimization variables, effectively reducing computational complexity. Compared to traditional bit transmission schemes, the proposed scheme can extract core semantic information and compress redundant bits, achieving a higher semantic transmission rate under the same resource budget. Furthermore, compared to branch pruning and multi-block approximation schemes, the proposed scheme further improves optimization efficiency through subspace pruning and relocation techniques, exhibiting superior overall performance under different antenna configurations.

[0262] Figure 4The relationship between semantic communication rate and jammer transmit power is demonstrated. As jamming power increases, the transmission performance of all schemes gradually decreases. The proposed scheme coordinates communication resources by jointly optimizing transmit and receive beamforming and power combining ratios, accurately pointing the main lobe towards the target ground user while simultaneously pointing the deep null point towards the jamming source, thus maintaining the performance improvement. Overall, the proposed method consistently outperforms all baseline comparison schemes, especially in strong jamming environments, fully demonstrating the synergistic advantages of semantic robustness, spatial null guidance, and the SWIPT mechanism. Without an optimized resource scheduling scheme, the random power splitting and beamforming schemes cannot form a stable beam structure and exhibit the worst performance. Traditional bit transmission schemes aim to achieve accurate and error-free transmission and recovery of each bit, but this mode is extremely sensitive to channel distortion, noise, and hostile interference. Any bit errors caused by the jammer directly reduce the system's effective data throughput or bit error rate performance. Therefore, traditional bit transmission schemes lack semantic compression and similarity-based robustness, and their effective throughput decreases sharply with increasing jamming power.

[0263] Example 3

[0264] This embodiment provides a semantically assisted UAV relay wireless power transmission and anti-interference transmission system, including:

[0265] The acquisition module is used to acquire communication data of the target dual-hop relay air-to-ground communication system. The communication data includes communication data between the ground transmitter and the relay UAV, communication data between the relay UAV and the ground user, and jamming data between the jammer and the ground user.

[0266] The rate calculation module is used to calculate the semantic communication rate of ground users based on the communication data of the target dual-hop relay air-to-ground communication system and the semantic communication mechanism.

[0267] The model building module is used to construct a joint optimization model based on the semantic communication rate of ground users, with the optimization objective being to maximize the sum of the semantic communication rates of ground users and the optimization variables being the power splitting ratio of the relay UAV decoding, the transmitting precoding vector of the relay UAV, and the receiving precoding vector of the ground users.

[0268] The conversion module is used to robustly process the non-ideal channel error in the joint optimization model and then convert the problem of solving the optimization objective into a second-form problem.

[0269] The solution module is used to iteratively optimize the initial solution space of the second form problem using a monotonic optimization algorithm until the difference between the upper and lower bounds of the objective function value corresponding to the global optimal solution of the second form problem is less than a preset threshold, thus obtaining the optimization result of the optimization variable. In each iteration, a feasibility verification algorithm is used to verify the feasibility of any region in the initial solution space, and infeasible regions are removed from the initial solution space.

[0270] Example 4

[0271] This embodiment provides a semantically assisted UAV relay wireless power transmission and anti-interference transmission device, including a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the semantically assisted UAV relay wireless power transmission and anti-interference transmission method described in Embodiment 1.

[0272] Example 5

[0273] This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the semantically assisted UAV relay wireless power transmission and anti-interference transmission method described in Embodiment 1.

[0274] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0275] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0276] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0277] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0278] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A semantically assisted UAV relay wireless power transmission and anti-interference transmission method, characterized in that, include: Acquire communication data from the target's dual-hop relay air-to-ground communication system. The communication data includes communication data between the ground transmitter and the relay UAV, communication data between the relay UAV and the ground user, and jamming data between the jammer and the ground user. Based on the communication data of the target dual-hop relay air-to-ground communication system, the semantic communication rate of the ground user is calculated based on the semantic communication mechanism. Based on the semantic communication rate of ground users, a joint optimization model is constructed with the goal of maximizing the sum of the semantic communication rates of ground users and with the power splitting ratio of the relay UAV decoding, the transmitting precoding vector of the relay UAV and the receiving precoding vector of the ground users as optimization variables. After robustly handling the non-ideal channel error in the joint optimization model, the problem of solving the optimization objective is transformed into a second-form problem. The initial solution space of the second form problem is iteratively optimized using a monotonic optimization algorithm until the difference between the upper and lower bounds of the objective function value corresponding to the global optimal solution of the second form problem is less than a preset threshold, thus obtaining the optimization result of the optimization variable; In each iteration, a feasibility verification algorithm is used to verify the feasibility of any region in the initial solution space, and infeasible regions are removed from the initial solution space.

2. The semantically assisted UAV relay wireless power transmission and anti-interference transmission method according to claim 1, characterized in that, The step of calculating the semantic communication rate of ground users based on the communication data of the target dual-hop relay air-to-ground communication system and using a semantic communication mechanism includes: The ground transmitter semantically encodes the mission information and sends it to the relay UAV. The received signal from the relay UAV is represented as follows: ； in, This indicates the received signal of the relay drone. Indicates the power of the transmitting node. This represents the transmission precoding vector of the ground transmitter. This represents the received precoding vector of the relay drone. This represents the additive white Gaussian noise at the relay drone location. This represents the channel vector from the ground transmitter to the relay UAV. This indicates the semantic transmission signal of the ground transmitter. Indicates conjugate transpose; The relay drone harvests energy from the received signal. The power harvested by the relay drone is expressed as: ； ； ； in, This indicates the power collected by the relay drone. This represents the logistic function related to the input power of the relay UAV's energy harvesting. This represents the maximum power collected when the energy harvesting circuit is saturated. Used to ensure response under zero input or zero output conditions during energy harvesting. and These all represent constants related to resistance and other parameters in a circuit. Indicates the input power for energy harvesting; The relay drone performs semantic decoding on the received signal, and calculates the semantic information decoded signal of the relay drone, which is represented as: ； in, This represents the semantic information decoding signal of the relay drone. This indicates the power splitting ratio for decoding by the relay drone. Indicates noise processing; The semantic communication rate of the relay UAV is calculated and expressed as: ； ； ； in, This indicates the semantic communication rate of the relay drone. This indicates the bandwidth from the ground transmitter to the drone. This represents the average semantic information of each sentence in the signal received by the relay drone. This represents the average number of semantic symbols required for a relay drone to receive each word in the signal. This represents the average number of words per sentence received by the relay drone. This indicates the semantic similarity between the original sentence and semantic information in the signal received by the relay drone. Indicates the semantic received signal-to-noise ratio. Indicates the left asymptote. Indicates the right asymptote. Represents the logical growth rate. Indicates the center point. , Indicates the power used to process noise. This represents the power of the additive white Gaussian noise at the relay drone. Denotes the Euclidean norm; Calculate the first The received signal of a ground user is represented as follows: ； in, Indicates the first The received signal of each ground user Indicates the first Received precoding for each ground user Indicates the first Noise at ground user locations Indicates the relay drone to the Channel matrix for each ground user This represents the total number of ground users. Indicates that the relay drone is for the first The precoded vector sent by each ground user Indicates the jammer to the number Channel matrix for each ground user This indicates the jamming signal sent by the jammer. This indicates that the relay drone will re-encode the decoded semantic information into a sequence representing the first... Semantic transmission signals for each ground user; Calculate the first The semantic communication rate of a ground user is expressed as: ； ； ； in, Indicates the first Semantic communication rate of each ground user Indicates the relay drone to the Bandwidth for each ground user Indicates the first The average semantic information of each sentence in the signal received by a ground user. Indicates the first The average number of semantic symbols required per word for a ground user to receive the signal. Indicates the first The average number of words per sentence received by a ground user. Indicates the first Semantic similarity between the original sentence and semantic information in the signals received by each ground user Indicates the first Semantic reception signal-to-noise ratio for each ground user Indicates that the relay drone is for the first The precoded vector sent by each ground user This indicates the jammer's transmission power. Indicates the first Noise power at each ground user location.

3. The semantically assisted UAV relay wireless power transmission and anti-interference transmission method according to claim 1, characterized in that, The joint optimization model is expressed as follows: ； in, Indicates that the relay drone is for the first The precoded vector sent by each ground user Indicates the first The received precoding vectors of each ground user This indicates the power splitting ratio for decoding by the relay drone. This indicates that through optimization , and To obtain the maximum value, Indicates the first Semantic communication rate of each ground user This represents the total number of ground users. This indicates the non-ideal channel error. This indicates that through optimization To obtain the minimum value, This represents the semantic communication rate threshold for relay drones. Indicates and the The semantic communication rate threshold for each ground user This indicates the semantic communication rate of the relay drone. This indicates the power consumption required for the relay drone to fly. This indicates the power collected by the relay drone. This represents the semantic communication rate constraint for relay drones. This represents the semantic communication rate constraint for ground users. This indicates the power overhead constraint for the relay drone. This represents the normalization constraint of the received precoding vector. Indicates a power splitting ratio constraint; The non-ideal channel error Represented as: ； in, Indicates the jammer to the number Channel matrix for each ground user Indicates the jammer to the number The departure elevation angle of a ground user Indicates the jammer to the number The departure azimuth angle of a ground user Indicates the jammer to the number The minimum departure elevation angle for a ground user. Indicates the jammer to the number The maximum departure elevation angle for a ground user. Indicates the jammer to the number The minimum departure azimuth angle for each ground user. Indicates the jammer to the number The maximum value of the departure azimuth angle for each ground user.

4. The semantically assisted UAV relay wireless power transmission and anti-interference transmission method according to claim 3, characterized in that, After robustly processing the non-ideal channel error in the joint optimization model, the problem of solving the optimization objective is transformed into a second-form problem, including: After robustly handling the non-ideal channel error in the joint optimization model, the problem of solving the optimization objective is transformed into a first-form problem, expressed as: ； Transforming the problem in the first form into the problem in the second form, it can be expressed as: ； ； ； in, Represents the new set of optimization variables. Let k be the new optimization variable. Represents the old set of optimization variables. This indicates that through optimization To obtain the maximum value, This indicates the first [unclear] under the old set of optimization variables. Semantic communication rate of each ground user This represents the feasible region of the new optimization variable. This represents the feasible region of the old optimization variables. Represents the new set of optimization variables The objective function value is as follows; The set of old optimization variables corresponding to the optimal solution of the new set of optimization variables in the second form of solving the problem is the unique solution of the first form of solving the problem.

5. The semantically assisted UAV relay wireless power transmission and anti-interference transmission method according to claim 4, characterized in that, The initial solution space of the second-form problem is iteratively optimized using a monotonic optimization algorithm until the difference between the upper and lower bounds of the objective function value corresponding to the global optimal solution of the second-form problem is less than a preset threshold. The optimization results of the optimization variables are then obtained, including: For the second form of the problem, let , Construct the initial solution space ,in This represents the lower bound of the initial solution space. This represents the upper bound of the initial solution space. This represents the minimum value of the new set of optimization variables. This represents the maximum value of the new set of optimization variables; Randomly select any region space in the initial solution space A feasibility verification algorithm was used to verify the spatial region. The feasibility, if the regional space If not feasible, then the regional space Remove from the initial solution space; If the regional space It is feasible to use a binary search method to connect the lines. Find the regional space Intersection vector with Pareto boundary and the interval of intersection connecting wire Represented as: ； in, Representing regional space The lower bound, Representing regional space The upper realm, This represents the lower bound of a feasible solution for the interval of intersection points. This indicates an upper bound for feasible solutions in the interval of intersection points. This represents an intermediate variable used when searching for Pareto boundary intersections on the connector. Indicates the lower bound The objective function value at that point, Indicates the normalization factor; Update the lower bound of the objective function value corresponding to the global optimal solution to , Indicates a feasible solution based on the lower bound. The calculated objective function value; regional space Perform orthogonal partitioning to obtain There are three non-overlapping subspaces, and the upper and lower vertices of each subspace are represented as follows: ； ； in, This represents the upper bound vertex vector of the m-th subspace obtained after partitioning the current region space in the n-th iteration. This represents the lower bound vertex vector of the m-th subspace obtained by partitioning the current region space in the n-th iteration. Indicates only the first A unit vector with 1 element and the rest being 0. Represents the intersection vector The coordinates in the m-th dimension This represents the result obtained by partitioning the current region space in the nth iteration. The lower bound vertex vector of each subspace. Indicates only the first A unit vector with 1 element and 0 elements; Regional space After partitioning, a new subspace is obtained. Furthermore, the subspaces are orthogonal and do not overlap. The updated new region space is represented as follows: ； in, Indicates a new regional space; For new regional space Remove invalid regions that are below their lower bound; For new regional space The objective function value at the upper bound ,like This will create a new regional space. Remove; if Then update the new area space. The lower bound is represented as: ； in, Indicates the updated new area space The lower bound of the value is denoted by min, which represents the minimum value. Indicates the updated new area space The upper realm, This represents the lower bound of the objective function value corresponding to the global optimal solution.

6. The semantically assisted UAV relay wireless power transmission and anti-interference transmission method according to claim 5, characterized in that, The feasibility verification algorithm used to verify the feasibility of any region in the initial solution space includes: Verification area space Does a feasible solution exist for the new set of optimization variables? Feasible solution The sixth constraint must be satisfied. ; Transform the problem from the first form of solution into the third form of solution: ； For the Received precoding vectors for each ground user The solution is obtained using the linear minimum mean square error algorithm, expressed as: ； ； in, Indicates that the relay drone is for the first The precoded vector sent by each ground user Represents the identity matrix. Indicates the transition amount; For relay drones to the first Precoded vectors sent to each ground user ,like Then the relay drone to the first Precoded vectors sent to each ground user It is a trivial stationary point in the third form of problem-solving; let Introducing the first auxiliary variable ,satisfy The problem is transformed from a third-form problem to a fourth-form problem: ； in, Indicates using replace Semantic communication rate constraints for subsequent ground users Indicates using replace Power overhead constraints for subsequent relay drones Indicates using replace The sixth constraint after; Semantic communication rate constraints for relay drones Simplifying, we get: ； in, Indicates the power used to process noise. This represents the power of the additive white Gaussian noise at the relay drone. Indicates the power of the transmitting node. This represents the transmission precoding vector of the ground transmitter. This represents the received precoding vector of the relay drone. This represents the additive white Gaussian noise at the relay drone location. This represents the channel vector from the ground transmitter to the relay UAV. This indicates the semantic transmission signal of the ground transmitter. , Indicates the left asymptote. Indicates the right asymptote. Represents the logical growth rate. Indicates the center point. Indicates the relay drone to the Bandwidth for each ground user Indicates the first The average semantic information of each sentence in the signal received by a ground user. Indicates the first The average number of semantic symbols required per word for a ground user to receive the signal. Indicates the first The average number of words per sentence received by a ground user; By combining the power splitting ratio constraint C5, the power splitting ratio for relay UAV decoding is obtained. The constraint C7 is represented as: ； will use replace Semantic communication rate constraints for subsequent ground users and use replace The sixth constraint after By merging, we obtain the merged expression: ； in, , Indicates the first The first auxiliary variable corresponding to each ground user ; Introducing a second auxiliary variable Third auxiliary variable Fourth auxiliary variable Fifth auxiliary variable The merged expression is transformed into convex form, with the eighth constraint C8, ninth constraint C9, tenth constraint C10, eleventh constraint C11, and twelfth constraint C12 represented as follows: ； ； ； ； ； in, Describes the Euclidean norm. Indicates transpose. Indicates the fourth auxiliary variable The The value of the next iteration. Represents the fifth auxiliary variable The The value of the next iteration, max indicates taking the maximum value; Introducing the first slack variable Second relaxation variable , will use replace Power overhead constraints of subsequent relay drones The thirteenth constraint C13, the fourteenth constraint C14, and the fifteenth constraint C15, converted to convex form, are expressed as follows: ； ； ； in, and These all represent constants related to resistance and other parameters in a circuit. Indicates energy harvesting efficiency. , Represents the first slack variable The The value of the next iteration; The fourth form problem is transformed into a convex optimization problem, expressed as: ； Solve the convex optimization problem to obtain feasible solutions in the domain space, and use the feasible solution obtained in the last iteration as the optimization result for the optimization variables.

7. A semantically assisted UAV relay wireless power transmission and anti-interference transmission system, characterized in that, include: The acquisition module is used to acquire communication data of the target dual-hop relay air-to-ground communication system. The communication data includes communication data between the ground transmitter and the relay UAV, communication data between the relay UAV and the ground user, and jamming data between the jammer and the ground user. The rate calculation module is used to calculate the semantic communication rate of ground users based on the communication data of the target dual-hop relay air-to-ground communication system and the semantic communication mechanism. The model building module is used to construct a joint optimization model based on the semantic communication rate of ground users, with the optimization objective being to maximize the sum of the semantic communication rates of ground users and the optimization variables being the power splitting ratio of the relay UAV decoding, the transmitting precoding vector of the relay UAV, and the receiving precoding vector of the ground users. The conversion module is used to robustly process the non-ideal channel error in the joint optimization model and then convert the problem of solving the optimization objective into a second-form problem. The solution module is used to iteratively optimize the initial solution space of the second form problem using a monotonic optimization algorithm until the difference between the upper and lower bounds of the objective function value corresponding to the global optimal solution of the second form problem is less than a preset threshold, thereby obtaining the optimization result of the optimization variable; In each iteration, a feasibility verification algorithm is used to verify the feasibility of any region in the initial solution space, and infeasible regions are removed from the initial solution space.

8. A semantically assisted UAV relay wireless power transmission and anti-interference transmission device, characterized in that, Including processor and storage media; The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the semantically assisted UAV relay wireless power transfer and anti-interference transmission method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the semantically assisted UAV relay wireless power transmission and anti-interference transmission method as described in any one of claims 1 to 6.

Images