Resource optimization method and apparatus for UAV-assisted semantic communication

Abstract

This application discloses a resource optimization method and apparatus for using UAV-assisted semantic communication, comprising: determining the original text and generating corresponding semantic information based on the original text; recovering the semantic information and generating recovered text; determining a semantic similarity maximization problem based on the original text and the recovered text, wherein semantic similarity is used to indicate the semantic accuracy and semantic completeness between the original text and the recovered text, and the semantic similarity maximization problem includes a resource allocation optimization problem and a semantic information transmission strategy optimization problem; creating an initial strategy and updating the parameters of the optimized initial strategy using the standard gradient ascent method; determining the relationship between the importance distribution of semantic information and semantic similarity using the updated parameters; and determining the optimal solutions to the resource allocation optimization problem and the semantic information transmission strategy optimization problem based on the determined relationship, thereby maximizing the semantic similarity between the original text and the recovered text.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) communication technology, and in particular to a resource optimization method and apparatus for using UAV-assisted semantic communication. Background Technology

[0002] Currently, information and communication science faces another critical juncture: communication technologies guided by classical information theory are constraining the future development of communication. Meanwhile, people's multi-dimensional and multi-channel communication needs will lead to the diversification of data modalities and task requirements, all of which will pose challenges to existing architectures. These challenges currently require solutions through a new paradigm that combines perceptual and semantic communication.

[0003] In existing communication methods, drones are often used as intermediaries to connect users and base stations for semantic data transmission. That is, the semantic data transmission process is: user (terminal device) – drone – base station. Specifically, the terminal device encodes the original text data to generate corresponding semantic data and transmits this semantic data to the drone. The drone, acting as a relay device, transmits the received semantic data to the base station. The base station then decodes the semantic data and generates the recovered text. In this process, the semantic similarity between the recovered text generated by the base station and the original text is crucial to measuring the quality of this semantic communication process; therefore, it is necessary to maximize semantic similarity. Semantic similarity includes semantic accuracy and semantic completeness.

[0004] However, existing optimization algorithms cannot solve the problem of maximizing semantic similarity, thus failing to improve the quality of the semantic communication process.

[0005] To address the technical problem that traditional optimization algorithms in the existing technologies cannot solve the problem of maximizing semantic similarity, thus failing to improve the quality of the semantic communication process, no effective solution has yet been proposed. Summary of the Invention

[0006] The embodiments of this disclosure provide a resource optimization method and apparatus for using unmanned aerial vehicles (UAVs) to assist semantic communication, thereby at least solving the technical problem that traditional optimization algorithms in the prior art cannot solve the problem of maximizing semantic similarity, thus failing to improve the quality of the semantic communication process.

[0007] According to one aspect of the present disclosure, a resource optimization method for using unmanned aerial vehicle (UAV)-assisted semantic communication is provided, comprising: determining original text and generating corresponding semantic information based on the original text; recovering the semantic information and generating recovered text; determining a semantic similarity maximization problem based on the original text and the recovered text, wherein semantic similarity is used to indicate the semantic accuracy and semantic integrity between the original text and the recovered text, and the semantic similarity maximization problem includes a resource allocation optimization problem and a semantic information transmission strategy optimization problem; creating an initial strategy and updating the parameters of the optimized initial strategy using the standard gradient ascent method; determining the relationship between the importance distribution of semantic information and semantic similarity using the updated parameters; and determining the optimal solutions to the resource allocation optimization problem and the semantic information transmission strategy optimization problem based on the determined relationship, thereby maximizing the semantic similarity between the original text and the recovered text.

[0008] According to another aspect of the present disclosure, a storage medium is also provided, the storage medium including a stored program, wherein, when the program is executed, a processor performs any of the methods described above.

[0009] According to another aspect of the present disclosure, a resource optimization apparatus for using unmanned aerial vehicle (UAV)-assisted semantic communication is also provided, comprising: a semantic information generation module for determining original text and generating corresponding semantic information based on the original text; a restored text generation module for restoring the semantic information and generating restored text; a maximization problem determination module for determining a semantic similarity maximization problem based on the original text and the restored text, wherein semantic similarity is used to indicate the semantic accuracy and semantic integrity between the original text and the restored text, and the semantic similarity maximization problem includes a resource allocation optimization problem and a semantic information transmission strategy optimization problem; an update module for creating an initial strategy and updating the parameters of the optimized initial strategy using the standard gradient ascent method; a relationship determination module for determining the relationship between the importance distribution of semantic information and semantic similarity using the updated parameters; and a solution module for determining the optimal solution of the resource allocation optimization problem and the optimal solution of the semantic information transmission strategy optimization problem based on the determined relationship, thereby maximizing the semantic similarity between the original text and the restored text.

[0010] This application discloses a resource optimization method utilizing UAV-assisted semantic communication. First, a processor determines the original text and generates corresponding semantic information based on it. Then, the processor recovers the semantic information and generates recovered text. Further, the processor determines a semantic similarity maximization problem based on the original and recovered text. Next, the processor creates an initial policy and updates the pre-created initial policy parameters using the standard gradient ascent method. Then, the processor uses the updated parameters to determine the relationship between the importance distribution of semantic information and semantic similarity. Finally, based on the determined relationship, the processor determines the optimal solution to the resource allocation optimization problem and the optimal solution to the semantic information transmission optimization policy, thereby maximizing the semantic similarity between the original and recovered text.

[0011] Since this application identifies the semantic similarity maximization problem based on the original text and the recovered text, it can be further decomposed into a resource allocation optimization problem and a semantic information transmission strategy optimization problem. The resource allocation optimization problem is related to the uplink channel capacity allocated by the UAV to the terminal equipment and the uplink channel capacity allocated by the UAV to the base station; the semantic information transmission strategy optimization problem is related to the selection of semantic information.

[0012] Furthermore, since this application creates an initial strategy, updates the parameters of the optimized initial strategy using the standard gradient ascent method, and uses the updated parameters to determine the relationship between the importance distribution of semantic information and semantic similarity, it can determine the optimal solution to both the resource allocation optimization problem and the semantic information transmission strategy optimization problem based on the determined relationship. This allows for maximizing the semantic similarity between the original text and the recovered text.

[0013] This achieves the technical effect of maximizing the semantic similarity between the original text and the recovered text, thereby improving the quality of the semantic communication process. It also solves the technical problem in existing technologies where traditional optimization algorithms cannot solve the semantic similarity maximization problem, thus failing to improve the quality of the semantic communication process. Attached Figure Description

[0014] The accompanying drawings, which are included to provide a further understanding of this disclosure and form part of this application, illustrate exemplary embodiments of this disclosure and are used to explain this disclosure, but do not constitute an undue limitation of this disclosure. In the drawings:

[0015] Figure 1 This is a hardware structure block diagram of a computing device for implementing the method described in Embodiment 1 of this disclosure;

[0016] Figure 2 This is a schematic diagram of a system model for semantic communication according to Embodiment 1 of this disclosure;

[0017] Figure 3 This is a flowchart illustrating a resource optimization method utilizing unmanned aerial vehicle-assisted semantic communication according to the first aspect of Embodiment 1 of this disclosure; and

[0018] Figure 4 This is a schematic diagram of a resource optimization apparatus for unmanned aerial vehicle-assisted semantic communication according to the first aspect of Embodiment 2 of this disclosure. Detailed Implementation

[0019] To enable those skilled in the art to better understand the technical solutions of this disclosure, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this disclosure.

[0020] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0021] Example 1

[0022] According to this embodiment, a method embodiment for resource optimization using UAV-assisted semantic communication is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0023] The method embodiments provided in this example can be executed on mobile terminals, computer terminals, servers, or similar computing devices. Figure 1 A hardware block diagram of a computing device for implementing a resource optimization method using UAV-assisted semantic communication is shown. Figure 1As shown, a computing device may include one or more processors (processors may include, but are not limited to, microprocessors such as MCUs or programmable logic devices such as FPGAs), memory for storing data, transmission devices for communication functions, and input / output interfaces. The memory, transmission devices, and input / output interfaces are connected to the processor via a bus. In addition, it may also include a display, keyboard, and cursor control device connected to the input / output interfaces. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, a computing device may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0024] It should be noted that the aforementioned one or more processors and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element in a computing device. As involved in the embodiments of this disclosure, the data processing circuits serve as processor control (e.g., selection of a variable resistor termination path connected to an interface).

[0025] The memory can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the resource optimization method using UAV-assisted semantic communication in the embodiments of this disclosure. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby realizing the above-mentioned resource optimization method using UAV-assisted semantic communication for the application. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the computing device via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0026] The transmission device is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the computing device's communications provider. In one example, the transmission device includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0027] The display can be, for example, a touchscreen liquid crystal display (LCD), which allows users to interact with the user interface of the computing device.

[0028] It should be noted here that, in some optional embodiments, the above... Figure 1 The computing device shown may include hardware elements (including circuitry), software elements (including computer code stored on a computer-readable medium), or a combination of both hardware and software elements. It should be noted that... Figure 1 This is only one instance of a specific particular instance, and is intended to illustrate the types of components that may exist in the aforementioned computing devices.

[0029] Figure 2 This is a schematic diagram of the system model for semantic communication according to this embodiment. (Refer to...) Figure 2 As shown, the system model includes N users, M drones, and a base station. Each of the N users has a corresponding terminal device.

[0030] User i can use the corresponding terminal device to process the original text L i Data extraction is performed, and the extracted data is modeled using a knowledge graph to obtain a model that corresponds to the original text L. i The corresponding semantic information G i Furthermore, user i can also transmit semantic information G through the corresponding terminal device. i Send to drone j.

[0031] The drone j is used to transmit the received semantic information G i Send to base station k.

[0032] Base station k is used to process the received semantic information G i Perform the recovery to generate the recovery text L. i (a i ,G).

[0033] It should be noted that the terminal devices, drones j, and base stations k in the system can all use the hardware structure described above.

[0034] Under the aforementioned operating environment, according to the first aspect of this embodiment, a resource optimization method utilizing unmanned aerial vehicle (UAV)-assisted semantic communication is provided. This method comprises... Figure 2 The processor implementation shown. Figure 3 A flowchart illustrating the method is shown below. (Refer to...) Figure 3 As shown, the method includes:

[0035] S302: Determine the original text and generate corresponding semantic information based on the original text;

[0036] S304: Recover semantic information and generate recovered text;

[0037] S306: Based on the original text and the restored text, determine the semantic similarity maximization problem, where semantic similarity is used to indicate the semantic accuracy and semantic integrity between the original text and the restored text. The semantic similarity maximization problem includes the resource allocation optimization problem and the semantic information transmission strategy optimization problem.

[0038] S308: Create an initial policy and update the parameters of the optimized initial policy using the standard gradient ascent method;

[0039] S310: Determine the relationship between the importance distribution of semantic information and semantic similarity using the updated parameters; and

[0040] S312: Based on the determined relationship, determine the optimal solutions to the resource allocation optimization problem and the semantic information transmission strategy optimization problem, thereby maximizing the semantic similarity between the original text and the recovered text. Specifically, refer to... Figure 2 As shown, firstly, the processor in the terminal device determines the original text and generates corresponding semantic information based on the original text (S302). Specifically, the processor in the terminal device can, for example, receive the original text L sent by the user. i Then, for the original text L i Data extraction is performed, and the extracted data is modeled using a knowledge graph to obtain semantic information G. i Among them, semantic information in, And H i It is G′ i The number of semantic triples in G. This semantic information G i The following constraints should be met:

[0041]

[0042] Among them, c i (t) indicates that drone j is assigned to user i to transmit semantic information G. i uplink channel capacity. j(t) represents the semantic information G assigned to base station k by drone j for transmitting the semantic information. i Uplink channel capacity. Z(G′) i R represents the semantic information G being transmitted. i The amount of data.

[0043] Semantic information G extracted by the processor in the terminal device i The amount of data will be much smaller than the original text L i The amount of data is due to the semantic information G. i The dual-token relationship between entity pairs can reduce the amount of raw text L i Redundant context in.

[0044] Then, user i uses the processor in the terminal device to process the determined semantic information G. i The information is transmitted to drone j, where the processor then processes the semantic information G. i Transmitted to base station k.

[0045] Finally, base station k recovers the semantic information and generates the recovered text (S304). Specifically, the processor in base station k uses a text generation model to extract the received semantic information G. i To recover coherent multi-sentence text. The recovered text is... And among them, w′ i,m This represents the restored sentence text.

[0046] Then, the processor in UAV k determines the semantic similarity maximization problem (S306) based on the original text and the restored text. Semantic similarity is used to indicate the semantic accuracy and semantic completeness between the original text and the restored text. The semantic accuracy between the original text and the restored text is:

[0047]

[0048] Where, σ(L′) i (a i (t),G′ i ),w′ i,m ) indicates the recovery of text L′ i (α i (t),G′ i In the restored sentence text w′ i,m The number of times it appears; σ(L) i ,w′ i,m ) indicates in the original text L i In the middle, the restored sentence text w′ i,m Number of times it appears.

[0049] The semantic integrity between the original text and the recovered text is as follows:

[0050]

[0051] Where, σ(L′) i (a i (t),G′ i ),w′ i,m ) indicates the recovery of text L′ i (α i (t),G′ i In the restored sentence text w′ i,m The number of times σ(L) appears i ,w′ i,m ) indicates in the original text L i In the middle, the restored sentence text w′ i,m Number of times it appears.

[0052] Based on Formulas 2 and 3 above, the text L′ is restored. i (α i (t),G′ i Semantic similarity can be represented as:

[0053]

[0054] Here, φ∈(0,1) is a parameter used to adjust the contribution of semantic accuracy and semantic completeness to semantic similarity. That is, φ is used to balance the two. Increasing the value of φ will increase the impact of semantic accuracy on semantic similarity. In other words, as φ increases, semantic accuracy becomes more important for semantic similarity. θ i This is an additional penalty for short texts, which can be expressed by the following formula:

[0055]

[0056] When the recovered text is greater than or equal to the original text, it indicates that the recovered text is relatively complete, and θ i =1; When the recovered text is smaller than the original text, it indicates that the recovered text is incomplete.

[0057] The goal of this scheme is to maximize the semantic similarity between the original text and the recovered text while meeting transmission latency requirements. This maximization problem includes resource allocation optimization and semantic information delivery strategy optimization. The semantic similarity maximization problem can be expressed by the following formula:

[0058]

[0059] sta i,q (t)∈{0,1},i∈U,q∈Q,t∈T (Formula 6b)

[0060]

[0061]

[0062]

[0063] ||q(t+1)-q(t)|| 2 ≤(ν max Δ T ) 2 (Formula 6f)

[0064]

[0065] Formulas 6b, 6c, and 6d ensure that each user i can only occupy one uplink orthogonal resource block, and each uplink orthogonal resource block can only be allocated to one user i for semantic information transmission. Formula 6e represents the latency requirement for semantic information transmission. Formula 6f represents the constraint that the movement of UAV j within adjacent time slots does not exceed the maximum movement distance. Formula 6g represents the distance constraint between two UAVs. As can be seen from Formula 6a, semantic similarity depends on the selected semantic information subset G′. i Allocation of uplink orthogonal resource blocks a i Among them, a i This represents the correlation variable between the user and the sub-channel. If the user is associated with the sub-channel, then a i =1; if the user is not associated with the sub-channel, then a i =0. Furthermore, the operational steps for determining the semantic similarity maximization problem will be described in detail later, and therefore will not be repeated here.

[0066] However, since Equation 6a is non-convex and depends on the text generation model used to reconstruct the text, it cannot be solved using traditional optimization algorithms. To address this issue, this application employs a reinforcement learning algorithm with an attention network to compute the semantic similarity between the original text and the reconstructed text.

[0067] Therefore, in order to solve the optimization problem in Equation 6a, the drone, as the intelligent agent, first creates an initial policy π. θ And using the initial strategy π θ Sampling of semantic communication actions to generate action sets Then, the drone defines the expected reward corresponding to the set of actions. And maximize expected reward (Right now, This generates an optimized initial policy. Then, the optimized initial policy π can be updated using the standard gradient ascent method. θThe parameter θ(S308) is calculated using the following formula:

[0068]

[0069] Where α is the learning rate. It is the gradient of the parameter θ.

[0070] Furthermore, the UAV uses the updated policy parameters to determine the relationship between the importance distribution of semantic information and semantic similarity (S308). The semantic information corresponds to the original text. This will be described in detail later, and therefore will not be repeated here.

[0071] Finally, based on the determined relationship, the UAV determines the optimal solution to the resource allocation optimization problem and the optimal solution to the semantic information transmission strategy optimization problem, thereby maximizing the semantic similarity between the original text and the recovered text (S310).

[0072] As described in the background section, in existing communication methods, drones are often used as intermediaries to connect users and base stations for semantic data transmission. That is, the semantic data transmission process is: user (terminal device) – drone – base station. Specifically, the terminal device encodes the original text data to generate corresponding semantic data and transmits this semantic data to the drone. The drone, acting as a relay device, transmits the received semantic data to the base station. The base station then decodes the semantic data and generates the recovered text. In this process, the semantic similarity between the recovered text generated by the base station and the original text is crucial to measuring the quality of this semantic communication process; therefore, it is necessary to maximize semantic similarity. Semantic similarity includes semantic accuracy and semantic completeness.

[0073] However, existing optimization algorithms cannot solve the problem of maximizing semantic similarity, thus failing to improve the quality of the semantic communication process.

[0074] Since this application identifies the semantic similarity maximization problem based on the original text and the recovered text, it can be further decomposed into a resource allocation optimization problem and a semantic information transmission strategy optimization problem. The resource allocation optimization problem is related to the uplink channel capacity allocated by the UAV to the terminal equipment and the uplink channel capacity allocated by the UAV to the base station; the semantic information transmission strategy optimization problem is related to the selection of semantic information.

[0075] Furthermore, since this application creates an initial strategy, updates the parameters of the optimized initial strategy using the standard gradient ascent method, and uses the updated parameters to determine the relationship between the importance distribution of semantic information and semantic similarity, it can determine the optimal solution to both the resource allocation optimization problem and the semantic information transmission strategy optimization problem based on the determined relationship. This allows for maximizing the semantic similarity between the original text and the recovered text.

[0076] This achieves the technical effect of maximizing the semantic similarity between the original text and the recovered text, thereby improving the quality of the semantic communication process. It also solves the technical problem in existing technologies where traditional optimization algorithms cannot solve the semantic similarity maximization problem, thus failing to improve the quality of the semantic communication process.

[0077] Optionally, the operation of determining the original text and generating corresponding semantic information based on the original text includes: extracting data from the original text using a terminal device and modeling the extracted data using a knowledge graph to obtain semantic information composed of multiple semantic triples. Further, optionally, the operation of recovering the semantic information and generating recovered text includes: recovering the semantic information using a text generation model pre-set by the base station and generating recovered text.

[0078] Specifically, firstly, the processor constructs a semantic communication model for transmitting semantic information based on terminal devices, drones, and base stations. This semantic communication model includes a system model, a perception model, a queuing model, and a semantic model.

[0079] System model creation: The system model includes N users, M drones, cellular base stations, and edge servers, such as... Figure 2 As shown in the figure. The system model is a multi-time-slot system, with each time slot represented by t.

[0080] Creation of the perception model: The perception duration of the i-th user in the t-th time slot. Represented as:

[0081]

[0082] in, Qi(t) represents the perception duration of user i in the t-th time slot, and Qi(t) represents the amount of data that was not processed. i This indicates the rate at which sensor data is generated.

[0083] The energy consumed by user i in data perception, i.e. It can be expressed by the following formula:

[0084]

[0085] in, Qi(t) represents the energy consumed by user i in data perception, and e represents the amount of data that was not processed. i This represents the perceptual cost per bit.

[0086] Queuing model creation: The queuing model considers a multi-task model. The amount of data Qi(t) that cannot be processed in time will be cached at user i and then migrated to drone j. The cache queue state at the user end is represented as follows:

[0087]

[0088] Where Q1(t) represents the data that the first user terminal did not have time to process, ..., Q N (t) represents the data that the Nth user terminal did not have time to process.

[0089] Creating a semantic model: using w i,n It represents words, symbols, or punctuation marks in text data, hence the name w. i,n The tokens are used to transmit text data from user i to base station k. The data consists of a series of tokens, as shown below:

[0090]

[0091] Where V represents words, and N represents vocabulary. i It is the original text L i The number of symbols in the text.

[0092] Then, the processor in the terminal device processes the original text L i Data extraction is performed, and the extracted data is modeled using a knowledge graph to obtain semantic information G. i The amount of semantic information extracted will be far less than that of the original text L. i The amount of data is due to the semantic information G. i The dual-token relationship between entity pairs can reduce the amount of raw text L i Redundant context in.

[0093] Further, a transmission model is created. First, the coordinates of user i, drone j, and base station k are represented in a three-dimensional coordinate system. The coordinates of user i are q. i =[q i,x ,q i,y The coordinates of the drone are... Furthermore, the flight altitude H of drone j remains constant. The coordinates of base station k are q0 = [x0, y0, H0]. Here, H0 represents the altitude of base station k relative to user i. The channel gain between the i-th user and the j-th drone is:

[0094]

[0095] The channel gain between the j-th drone and base station k is:

[0096]

[0097] Where α = 2, and the distance between the i-th user and the j-th drone is d. i,j (t), the distance d between the j-th drone and the base station k m,0 (t). β0 is the channel gain per unit distance.

[0098] As mentioned above, UAV j will allocate Q uplink orthogonal resource blocks to user i using orthogonal frequency division multiple access (OFDMA). Assuming each user i can only occupy one uplink orthogonal resource block, and each uplink orthogonal resource block can only be allocated to one user, then user i will transmit semantic information G. i The uplink channel capacity is:

[0099]

[0100] Where W represents the bandwidth of each uplink orthogonal resource block. i,j (t) indicates that user i will send semantic information G i The transmission power transmitted to UAV j. q Represents the allocation of resource block P to the user i,j The interference of (t). N0 is the noise power spectral density. And assume that the delay for each user i to transmit data to drone j is limited to T.

[0101] Semantic information G transmitted by drone j to base station k i The uplink channel capacity is:

[0102]

[0103] Among them, P j,0 (t) indicates that the drone j will transmit semantic information G. i The transmission power transmitted to base station k. j,0 (t) represents the channel gain between the drone j and the base station k.

[0104] Finally, base station k uses a pre-set text generation model to recover semantic information and generate recovered text.

[0105] Optionally, the operation of determining the semantic similarity maximization problem based on the original text and the restored text includes: determining the semantic accuracy between the original text and the restored text; determining the semantic integrity between the original text and the restored text; and determining the semantic similarity maximization problem based on semantic accuracy and semantic integrity. Further optionally, the operation of determining the semantic similarity maximization problem based on semantic accuracy and semantic integrity includes: determining the semantic similarity between the original text and the restored text based on semantic accuracy and semantic integrity; determining the constraints corresponding to the semantic similarity; and maximizing the semantic similarity while satisfying the constraints, thus determining the semantic similarity maximization problem.

[0106] Specifically, the semantic accuracy between the original text and the recovered text is:

[0107]

[0108] Where, σ(L′) i (a i (t),G′ i ),w′ i,m ) indicates the recovery of text L′ i (α i (t),G′ i In the restored sentence text w′ i,m The number of times it appears; σ(L) i ,w′ i,m ) indicates in the original text L i In the middle, the restored sentence text w′ i,m Number of times it appears.

[0109] The semantic integrity between the original text and the recovered text is as follows:

[0110]

[0111] Where, σ(L′) i (a i (t),G′ i ),w′ i,m ) indicates the recovery of text L′ i (α i (t),G′ i In the restored sentence text w′ i,m The number of times σ(L) appears i ,w′ i,m ) indicates in the original text L i In the middle, the restored sentence text w′ i,m Number of times it appears.

[0112] Based on Formulas 2 and 3 above, the text L′ is restored. i(α i (t),G′ i Semantic similarity can be represented as:

[0113]

[0114] Here, φ∈(0,1) is a parameter used to adjust the contribution of semantic accuracy and semantic completeness to semantic similarity. That is, φ is used to balance the two. Increasing the value of φ will increase the impact of semantic accuracy on semantic similarity. In other words, as φ increases, semantic accuracy becomes more important for semantic similarity. θ i This is an additional penalty for short texts, which can be expressed by the following formula:

[0115]

[0116] When the recovered text is greater than or equal to the original text, it indicates that the recovered text is relatively complete, and θ i =1; When the recovered text is smaller than the original text, it indicates that the recovered text is incomplete.

[0117] Furthermore, constraints corresponding to semantic similarity are determined. Specifically, referring to the above description, this embodiment assumes that each user i can only occupy one uplink orthogonal resource block, and each uplink orthogonal resource block can only be allocated to one user i for semantic information transmission, as shown in formulas 6b, 6c, and 6d above. The latency of semantic information transmission should be less than a given transmission latency threshold T, as shown in formula 6e above. The movement of the UAV within adjacent time slots does not exceed the maximum movement distance, as shown in formula 6f above. The distance between two UAVs is greater than a given distance threshold, as shown in formula 6g above.

[0118] Finally, under the above constraints, the semantic similarity is maximized, thus defining the semantic similarity maximization problem.

[0119] Optionally, the operation of creating an initial policy and updating the parameters of the optimized initial policy using the standard gradient ascent method includes: creating an initial policy; sampling actions related to semantic communication according to the initial policy and generating an action set; defining the expected reward of the action set and maximizing the expected reward to determine the optimized initial policy, wherein the expected reward corresponds to the semantic similarity; and updating the parameters of the optimized initial policy using the standard gradient ascent method.

[0120] Specifically, in order to solve the problem of maximizing semantic similarity, the drone first creates an initial strategy π. θ , where θ is the parameter of the initial strategy. Semantic communication actions are sampled to obtain an action set D. The action set... To evaluate the strategy that maximizes semantic similarity, the expected reward of the action in D is defined as:

[0121]

[0122] Where s represents the action state, a d Represents the set of actions, π θ Let R represent the optimized initial policy, where the reward for choosing action a in action state s is R. Furthermore, this formula is equivalent to the semantic similarity maximization problem.

[0123] Optimized initial policy π θ The goal is to maximize the semantic similarity between the original text and the recovered text. That is,

[0124] Then, the optimized initial policy π can be updated using the standard gradient ascent method. θ The parameter θ in the formula is:

[0125]

[0126] in, Indicates the learning rate. This represents the gradient of the parameter θ.

[0127] By iterating through the above update steps, the optimized initial policy π is obtained. θ The parameter θ can be used to find the relationship between the importance distribution of all semantic information and semantic similarity. This allows us to obtain the optimal solutions to the resource allocation optimization problem and the semantic information transmission strategy optimization problem, thereby maximizing the semantic similarity between the original text and the recovered text.

[0128] Optionally, the operation of determining the relationship between the importance distribution of semantic information and semantic similarity using the updated parameters includes: pre-determining the importance distribution of semantic information; and determining the relationship between the importance distribution of semantic information and semantic similarity based on the importance distribution of semantic information. Further, optionally, the operation of pre-determining the importance distribution of semantic information includes: constructing an importance vector, where the importance vector represents the importance distribution of semantic triples in the semantic information; labeling sentence texts in the original text and labeling semantic triples; determining a first importance to indicate the relevance between semantic triples and labeled sentence texts; determining a second importance to indicate the relevance between semantic triples and the original text; determining an importance vector based on the first and second importance; and determining the importance distribution of semantic information based on the importance vector.

[0129] Specifically, first, an importance vector is created. Wherein the importance vector f i (Gi ) is used to represent the importance distribution of semantic triples in semantic information.

[0130] Then, the base station processed the original text. The sentence text w i,n Mark and semantic information semantic triples in Perform tagging. Then, define the tags. The vector is Used to mark w i,n The vector is in, Represents semantic triples The sentence text itself. D x This represents the dimension of each tag vector.

[0131] Semantic triples With the original text L i The middle mark w i,n The correlation between them can be expressed as:

[0132]

[0133] in, It is a semantic triple The number of markers in the middle. Using the trained parameters, the attention network can calculate (W... tri x i,b g ) T (W tok x i,n And among them, (W) tri x i,b g ) T (W tok x i,n ) represents a semantic triple. Middle Mark With the original text L i The middle mark w i,n The correlation between them.

[0134] Define semantic triples With the original text L i The importance of the correlation between them can be expressed by the following formula:

[0135]

[0136] Semantic information G i The importance distribution can be represented by the following formula:

[0137]

[0138] in,

[0139] Therefore, by using a policy gradient-based reinforcement learning algorithm for importance assessment, the base station can analyze the relationship between importance distribution and semantic similarity measurement, thereby optimizing semantic information selection (i.e., semantic information delivery strategy optimization) and resource block allocation (i.e., resource allocation optimization).

[0140] Furthermore, it is worth noting that the policy gradient-based reinforcement learning algorithm consists of six parts: 1) agent, 2) environment, 3) action, 4) state, 5) policy, and 6) reward. The specific process is as follows:

[0141] Input: The raw text L for each user i i and transmission delay threshold T.

[0142] Initialization: Randomly generate the initial policy parameters θ and the interference I for each channel. q Task learning rate δ, number of iterations E

[0143] 1: Calculate the importance distribution f(G) according to the above importance distribution formula. i )

[0144] 2: Fori = 1 → Edo

[0145] 3: Utilizing the initial strategy π θ The trajectory of the mobile phone drone j is α=[α1,...,α U ]

[0146] 4: Update the parameter θ of the initial policy πθ using the standard gradient ascent method (iterative step 4)

[0147] 5: endfor

[0148] Thus, according to the first aspect of this embodiment, the semantic similarity between the original text and the recovered text can be maximized, thereby improving the quality of the semantic communication process.

[0149] In addition, refer to Figure 1 As shown, according to a second aspect of this embodiment, a storage medium is provided. The storage medium includes a stored program, wherein, when the program is executed, a processor performs any of the methods described above.

[0150] Therefore, according to this embodiment, the semantic similarity between the original text and the recovered text can be maximized, thereby improving the quality of the semantic communication process.

[0151] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0152] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0153] Example 2

[0154] Figure 4 A resource optimization apparatus 400 utilizing unmanned aerial vehicle-assisted semantic communication according to a first aspect of this embodiment is shown, which corresponds to the method described according to the first aspect of Embodiment 1. Reference Figure 4 As shown, the device 400 includes: a semantic information generation module 410, used to determine the original text and generate corresponding semantic information based on the original text; a restored text generation module 420, used to restore the semantic information and generate restored text; a maximization problem determination module 430, used to determine a semantic similarity maximization problem based on the original text and the restored text, wherein semantic similarity is used to indicate the semantic accuracy and semantic integrity between the original text and the restored text, and the semantic similarity maximization problem includes a resource allocation optimization problem and a semantic information transmission strategy optimization problem; an update module 440, used to create an initial strategy and update the parameters of the optimized initial strategy using the standard gradient ascent method; a relationship determination module 450, used to determine the relationship between the importance distribution of semantic information and semantic similarity using the updated parameters; and a solution module 460, used to determine the optimal solution of the resource allocation optimization problem and the optimal solution of the semantic information transmission strategy optimization problem based on the determined relationship, thereby maximizing the semantic similarity between the original text and the restored text.

[0155] Optionally, the semantic information generation module 410 includes: a semantic information generation submodule, which is used to extract data from the original text using a terminal device and to model the extracted data using a knowledge graph to obtain semantic information composed of multiple semantic triples.

[0156] Optionally, the recovery text generation module 420 includes: a recovery text generation module 420, used to recover semantic information and generate recovery text by utilizing a text generation model pre-set by the base station.

[0157] Optionally, the maximization problem determination module 430 includes: a semantic accuracy determination module for determining the semantic accuracy between the original text and the restored text; a semantic integrity determination module for determining the semantic integrity between the original text and the restored text; and a maximization problem determination submodule for determining the semantic similarity maximization problem based on semantic accuracy and semantic integrity.

[0158] Optionally, the maximization problem determination submodule includes: a semantic similarity determination module, used to determine the semantic similarity between the original text and the recovered text based on semantic accuracy and semantic integrity; a constraint determination module, used to determine the constraint conditions corresponding to the semantic similarity; and a first determination module, used to maximize the semantic similarity under the condition of satisfying the constraint conditions, and to determine the semantic similarity maximization problem.

[0159] Optionally, the update module 440 includes: a creation module for creating an initial policy; a generation module for sampling actions related to semantic communication according to the initial policy and generating an action set; an expected reward definition module for defining the expected reward of the action set and maximizing the expected reward to determine the optimized initial policy, wherein the expected reward corresponds to the semantic similarity; and an update submodule for updating the parameters of the optimized initial policy using the standard gradient ascent method.

[0160] Optionally, the relationship determination module 450 includes: a second determination module for pre-determining the importance distribution of semantic information; and a second determination module for determining the relationship between the importance distribution of semantic information and semantic similarity based on the importance distribution of semantic information.

[0161] Optionally, the second determining module includes: a construction module for constructing an importance vector, wherein the importance vector represents the importance distribution of semantic triples in the semantic information; a labeling module for labeling sentence texts in the original text and labeling semantic triples; a third determining module for determining a first importance for indicating the relevance between semantic triples and labeled sentence texts; a fourth determining module for determining a second importance for indicating the relevance between semantic triples and the original text; a fifth determining module for determining an importance vector based on the first and second importance; and a sixth determining module for determining the importance distribution of semantic information based on the importance vector.

[0162] Therefore, according to this embodiment, the semantic similarity between the original text and the recovered text can be maximized, thereby improving the quality of the semantic communication process.

[0163] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0164] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0165] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0166] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0167] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0168] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0169] 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 principle 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 resource optimization method utilizing unmanned aerial vehicles (UAVs) to assist semantic communication, characterized in that, include: Determine the original text and generate corresponding semantic information based on the original text; The semantic information is recovered, and the recovered text is generated; Based on the original text and the recovered text, a semantic similarity maximization problem is determined, wherein the semantic similarity is used to indicate the semantic accuracy and semantic completeness between the original text and the recovered text. The semantic similarity maximization problem includes a resource allocation optimization problem and a semantic information transmission strategy optimization problem, wherein the semantic similarity maximization problem is expressed by the following formula: The constraints corresponding to the semantic similarity maximization problem are expressed by the following formula: s.t.a i,q (t)∈{0,1},i∈U,q∈Q,t∈T; ||q(t+1)-q(t)|| 2 ≤(ν max Δ T ) 2 ; Create an initial policy and update the parameters of the optimized initial policy using the standard gradient ascent method; The updated parameters are used to determine the relationship between the importance distribution of the semantic information and the semantic similarity. as well as Based on the established relationship, the optimal solutions to the resource allocation optimization problem and the semantic information transmission strategy optimization problem are determined, thereby maximizing the semantic similarity between the original text and the recovered text.

2. The method according to claim 1, characterized in that, The operation of determining the original text and generating corresponding semantic information based on the original text includes: The original text is extracted using a terminal device, and the extracted data is modeled using a knowledge graph to obtain semantic information composed of multiple semantic triples.

3. The method according to claim 1, characterized in that, The operation of recovering the semantic information and generating the recovered text includes: The semantic information is recovered using a text generation model pre-set in the base station, and the recovered text is generated.

4. The method according to claim 1, characterized in that, Based on the original text and the recovered text, the operation for maximizing semantic similarity is determined, including: Determine the semantic accuracy between the original text and the recovered text; Determine the semantic integrity between the original text and the recovered text; and Based on the semantic accuracy and semantic completeness, the problem of maximizing semantic similarity is determined.

5. The method according to claim 3, characterized in that, Based on the semantic accuracy and semantic completeness, the operations for determining the semantic similarity maximization problem include: Based on the semantic accuracy and semantic integrity, the semantic similarity between the original text and the recovered text is determined; Determine the constraints corresponding to the semantic similarity; and Under the constraints, maximize the semantic similarity and determine the semantic similarity maximization problem.

6. The method according to claim 1, characterized in that, The operations of creating an initial policy and updating the parameters of the optimized initial policy using the standard gradient ascent method include: Create an initial strategy; Actions related to semantic communication are sampled according to the initial strategy, and a set of actions is generated; Define the expected reward of the action set, and maximize the expected reward to determine the optimized initial policy, wherein the expected reward corresponds to the semantic similarity; and Update the parameters of the optimized initial policy using the standard gradient ascent method.

7. The method according to claim 6, characterized in that, The operation of determining the relationship between the importance distribution of semantic information and the semantic similarity using the updated parameters includes: Predetermine the importance distribution of the semantic information; and Based on the importance distribution of the semantic information, determine the relationship between the importance distribution of the semantic information and the semantic similarity.

8. The method according to claim 7, characterized in that, The operation of pre-determining the importance distribution of the semantic information includes: Construct an importance vector, wherein the importance vector is used to represent the importance distribution of semantic triples in the semantic information; The sentence text in the original text is marked, and the semantic triples are marked; Determine the first importance of the semantic triples used to indicate the relevance between the semantic triples and the tagged sentence text; Determine the second importance used to indicate the relevance between the semantic triples and the original text; Based on the first importance and the second importance, the importance vector is determined; Based on the importance vector, the importance distribution of the semantic information is determined.

9. A storage medium, characterized in that, The storage medium includes a stored program, wherein, when the program is executed, the method described in any one of claims 1 to 8 is performed by a processor.

10. A resource optimization device utilizing unmanned aerial vehicle (UAV)-assisted semantic communication, characterized in that, include: The semantic information generation module is used to determine the original text and generate corresponding semantic information based on the original text; The restored text generation module is used to restore the semantic information and generate restored text; The maximization problem determination module is used to determine a semantic similarity maximization problem based on the original text and the recovered text, wherein the semantic similarity is used to indicate the semantic accuracy and semantic completeness between the original text and the recovered text. The semantic similarity maximization problem includes a resource allocation optimization problem and a semantic information transmission strategy optimization problem, wherein the semantic similarity maximization problem is expressed by the following formula: The constraints corresponding to the semantic similarity maximization problem are expressed by the following formula: s.t.a i,q (t)∈{0,1},i∈U,q∈Q,t∈T; ||q(t+1)-q(t)|| 2 ≤(ν max Δ T ) 2 ; The update module is used to create an initial policy and update the parameters of the optimized initial policy using the standard gradient ascent method. A relationship determination module is used to determine the relationship between the importance distribution of the semantic information and the semantic similarity using the updated parameters; as well as The solution module is used to determine the optimal solution of the resource allocation optimization problem and the optimal solution of the semantic information transmission strategy optimization problem based on the determined relationship, so as to maximize the semantic similarity between the original text and the recovered text.

Images