High energy efficiency semantic communication method, system and device based on layered metasurface

Abstract

The application relates to a high-energy-efficiency semantic communication method, system and device based on a laminated metasurface, a minimum semantic rate target, a minimum semantic similarity threshold and maximum transmission power are set, a semantic energy efficiency maximization is taken as an objective function, under a non-ideal angle error condition of a wireless channel, a minimum semantic rate target, a minimum semantic similarity threshold, maximum transmission power and amplitude normalization of each reflection unit of the laminated metasurface are taken as constraint conditions, an optimization model to be solved is constructed with a digital beamformer and a metasurface phase shift matrix as optimization variables, a non-ideal channel model based on an angle error is used, and wireless channel non-ideal angle error parameters are discretized, finally, a semantic maximization minimization method is used to solve the optimal digital beamformer, and a cyclic coordinate descent algorithm is used to solve the optimal metasurface phase shift matrix, semantic communication optimization of a communication system is achieved, and semantic energy efficiency maximization of the laminated metasurface in communication application is realized.

Description

Technical Field

[0001] The present invention relates to the field of semantic communication technology, and in particular to a high-energy-efficiency semantic communication method, system and device based on stacked metasurfaces. Background Art

[0002] Multiple-input, multiple-output (MIMO) communication systems have been widely adopted in fifth-generation (5G) wireless networks to improve spectral efficiency and support massive connectivity. However, MIMO systems are driven by the use of hundreds of antennas in base stations (BSs), which results in prohibitive hardware costs and power consumption due to the large number of radio frequency (RF) chains required for each antenna. While the development of hybrid analog-digital MIMO communication systems has significantly reduced the number of required RF chains, they still require a large number of phase shifters, resulting in low energy efficiency (EE).

[0003] The rise of two emerging innovative technologies—reconfigurable smart metasurfaces (RIS) and semantic communication (SemCom)—can enable energy-efficient holographic MIMO communications. Specifically, smart metasurfaces are artificial surfaces composed of a large number of low-cost meta-atoms. They can impose phase shifts on electromagnetic waves, reconfiguring the wireless propagation environment in an energy-efficient manner, thereby facilitating the use of large-scale arrays. Furthermore, semantic communication eliminates irrelevant information and transmits compressed critical information, thereby reducing system power consumption and improving energy efficiency (EE). Most existing research on smart metasurfaces relies on single-layer architectures, which have been proven to extend service coverage, enhance physical layer security, and improve achievable rate and energy efficiency. However, previous studies have shown that single-layer smart metasurfaces have limited beamforming design freedom and are unable to suppress multi-user interference. To overcome these shortcomings, the industry has proposed stacked metasurfaces (SMs), which vertically stack multiple metasurfaces to form a multi-layered metasurface structure. However, achieving the maximum semantic energy efficiency of stacked metasurfaces in communication applications remains a key technical challenge. Summary of the Invention

[0004] In response to the problems existing in the above-mentioned traditional technologies, a high-energy-efficiency semantic communication method based on stacked metasurfaces, a communication system and a computer device are provided.

[0005] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0006] On the one hand, a high-energy-efficiency semantic communication method based on a stacked metasurface is provided, which is applied to a communication system. The communication system includes a base station configured with a digital antenna, a stacked metasurface, and multiple user-end devices configured with a single antenna;

[0007] The high-energy-efficiency semantic communication method comprises the following steps:

[0008] Setting a minimum semantic rate target, a minimum semantic similarity threshold and a maximum transmission power of the communication system;

[0009] Taking the maximum semantic energy efficiency of the communication system as an objective function, under the condition of non-ideal angle error of the wireless channel, taking the minimum semantic rate target, the minimum semantic similarity threshold, the maximum transmission power and the amplitude normalization of each reflection unit of the layered metasurface as constraint conditions, an optimization model to be solved is constructed, with the digital beamformer and the metasurface phase shift matrix as optimization variables;

[0010] A non-ideal channel model based on angle error is constructed, and the non-ideal angle error of the wireless channel is parameterized;

[0011] The optimal digital beamformer is solved by using the semantic maximization minimization method, and the optimal metasurface phase shift matrix is solved by using the cyclic coordinate descent algorithm;

[0012] The optimal digital beamformer and the optimal metasurface phase shift matrix are used to optimize the semantic communication of the communication system.

[0013] On the other hand, a communication system is also provided, comprising a base station configured with a digital antenna, a layered metasurface and a plurality of user terminal devices configured with a single antenna, the communication system uses the optimal digital beamformer and the optimal metasurface phase shift matrix to optimize semantic communication; wherein the acquisition of the optimal digital beamformer and the optimal metasurface phase shift matrix comprises the following steps:

[0014] Setting a minimum semantic rate target, a minimum semantic similarity threshold and a maximum transmission power of the communication system;

[0015] Taking the maximum semantic energy efficiency of the communication system as an objective function, under the condition of non-ideal angle error of the wireless channel, taking the minimum semantic rate target, the minimum semantic similarity threshold, the maximum transmission power and the amplitude normalization of each reflection unit of the layered metasurface as constraint conditions, an optimization model to be solved is constructed, with the digital beamformer and the metasurface phase shift matrix as optimization variables;

[0016] A non-ideal channel model based on angle error is constructed, and the non-ideal angle error of the wireless channel is parameterized;

[0017] The optimal digital beamformer is solved by using the semantic maximization minimization method, and the optimal metasurface phase shift matrix is solved by using the cyclic coordinate descent algorithm;

[0018] The optimal digital beamformer and the optimal metasurface phase shift matrix are used to optimize the semantic communication of the communication system.

[0019] On the other hand, a computer device is provided, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned high-energy-efficiency semantic communication method based on stacked metasurfaces when executing the computer program.

[0020] One of the above technical solutions has the following advantages and beneficial effects:

[0021] The above-mentioned high-energy-efficiency semantic communication method, system and device based on stacked metasurfaces set the minimum semantic rate target, the minimum semantic similarity threshold and the maximum transmission power, and take the maximization of semantic energy efficiency as the objective function. Under the condition of non-ideal angular error of the wireless channel, the minimum semantic rate target, the minimum semantic similarity threshold, the maximum transmission power and the amplitude normalization of each reflection unit of the stacked metasurface are used as constraints. An optimization model to be solved with a digital beamformer and a metasurface phase shift matrix as optimization variables is constructed. The non-ideal channel model based on the angular error is used and the non-ideal angular error parameters of the wireless channel are discretized. Finally, the semantic maximization minimization method is used to solve the optimal digital beamformer and the cyclic coordinate descent algorithm is used to solve the optimal metasurface phase shift matrix for semantic communication optimization of the communication system, thereby realizing the maximization of the semantic energy efficiency of the stacked metasurface in communication applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the conventional technology, the following briefly introduces the drawings required for use in the embodiments or the conventional technology descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0023] Figure 1 1 is a flow chart of a method for high-energy-efficiency semantic communication based on stacked metasurfaces in one embodiment;

[0024] Figure 2 FIG. 1 is a schematic diagram of the architecture of a communication system in one embodiment. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and Examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the description of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0026] It should be noted that the reference to "embodiment" in this document means that the specific features, structures or characteristics described in conjunction with the embodiment may be included in at least one embodiment of the present invention. The presentation of this phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It will be understood by those skilled in the art that the embodiments described herein may be combined with other embodiments. The term "and / or" used in the specification of the present invention and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.

[0027] The following describes the embodiments of the present invention in detail with reference to the accompanying drawings.

[0028] In one embodiment, a high-energy-efficiency semantic communication method based on a stacked metasurface is provided, which is applied to a communication system comprising a base station configured with a digital antenna, a stacked metasurface, and a plurality of user-end devices configured with a single antenna. Figure 1 As shown, the above-mentioned high-energy-efficiency semantic communication method based on stacked metasurfaces may include the following processing steps S10 to S18:

[0029] S10, setting a minimum semantic rate target, a minimum semantic similarity threshold, and a maximum transmit power for the communication system;

[0030] S12, taking the maximization of the semantic energy efficiency of the communication system as the objective function, under the condition of non-ideal angle error of the wireless channel, with the minimum semantic rate target, minimum semantic similarity threshold, maximum transmission power and amplitude normalization of each reflective unit of the stacked metasurface as constraints, constructs an optimization model to be solved with the digital beamformer and metasurface phase shift matrix as optimization variables;

[0031] S14, constructing a non-ideal channel model based on angle error and discretizing the non-ideal angle error of the wireless channel into parameters;

[0032] S16, using the semantic maximization minimization method to solve the optimal digital beamformer, and using the cyclic coordinate descent algorithm to solve the optimal metasurface phase shift matrix;

[0033] S18, semantic communication optimization of the communication system using the optimal digital beamformer and the optimal metasurface phase shift matrix.

[0034] It can be understood that the high energy efficiency semantic communication method based on stacked metasurfaces provided in this embodiment can be applied to Figure 2 In the communication system application environment shown in the figure, the communication system can be equipped with A base station (which can be recorded as BS) with digital antennas, a Alayers of intelligent super surface and each layer of intelligent super surface has The stacked metasurface (which can be recorded as SM) of uniform plane units and K User terminals with a single antenna (referred to as User1 to User K ).

[0035] You can define collections 、 and The channel between the digital antenna of the base station and the stacked metasurface of the first layer is , No. a Layer-by-layer metasurface and a The channel between the +1-layer stacked metasurfaces is , base station and the k The channel between the user end devices is Using the semantic transfer framework based on deep learning, the original sentence Through neural networks Mapped into semantic symbols, i.e. .remember For the k User semantic symbols are transmitted through a digital beamformer before transmission. Processing. a The metasurface phase shift matrix of the stacked metasurface is Therefore, the semantic beamformer transmitted to the user end device can be expressed as Therefore, k The signal received at the user terminal device is ,in, is additive noise, is the average noise power. The equivalent analog beamformer generated by SM is defined as ,Right now:

[0036]

[0037] DeepSC Receiver Received degraded semantic symbols After that, extract and restore the original sentence In order to evaluate the quality of semantic communication, the semantic rate (suts / s) is widely used, namely:

[0038]

[0039] in, For the original sentence The expected amount of semantic information in , is the bandwidth, is the average number of semantic symbols per word, For the original sentence The expected number of words, Sentences for recovery With the original sentence The semantic similarity between . For the k The signal-to-noise ratio of each user terminal device:

[0040]

[0041] However, It mainly depends on the neural network used , its black box nature makes it impossible to accurately characterize it. A generalized logic function can handle this problem, which converts the semantic rate Maps to:

[0042]

[0043] in, , , and is a non-negative parameter. And the total power consumption of the communication system considered is:

[0044]

[0045] in, For efficiency, is the power required by the RF chain, is the power consumed by each meta-atom, is the power consumption of the baseband processor. Therefore, the semantic energy efficiency is expressed as:

[0046]

[0047] The above-mentioned high-energy-efficiency semantic communication method based on stacked metasurfaces sets the minimum semantic rate target, the minimum semantic similarity threshold and the maximum transmission power, and takes the maximization of semantic energy efficiency as the objective function. Under the condition of non-ideal angular error of the wireless channel, the minimum semantic rate target, the minimum semantic similarity threshold, the maximum transmission power and the amplitude normalization of each reflection unit of the stacked metasurface are used as constraints. An optimization model to be solved with the digital beamformer and the metasurface phase shift matrix as optimization variables is constructed. The non-ideal channel model based on the angular error and the non-ideal angular error parameters of the wireless channel are discretized. Finally, the semantic maximization minimization method is used to solve the optimal digital beamformer and the cyclic coordinate descent algorithm is used to solve the optimal metasurface phase shift matrix for semantic communication optimization of the communication system, thereby realizing the maximization of the semantic energy efficiency of the stacked metasurface in communication applications.

[0048] In particular, the high-energy-efficiency semantic communication method based on the stacked metasurface can include the following processing steps:

[0049] Step 1: Set the minimum semantic rate target of the communication system , the minimum semantic similarity threshold , and the maximum transmission power .

[0050] Step 2: Take the semantic energy efficiency maximization as the objective function, under the non-ideal angle error of the wireless channel, take the minimum semantic rate target , the minimum semantic similarity threshold , the maximum transmission power , and the amplitude normalization of each reflection unit of the smart metasurface as the constraint conditions, take the user-side device k expected digital beamformer and the metasurface phase shift matrix as the optimization variables, and construct an optimization model to be solved:

[0051]

[0052] wherein C1 to C4 represent each constraint condition respectively.

[0053] Step 3: Construct a non-ideal channel model based on the angle error and parameterize the non-ideal angle error of the wireless channel.

[0054] In particular, let the channel belong to a given uncertainty set, i.e. , and be the upper and lower limits of the azimuth angle respectively, and be the upper and lower limits of the elevation angle respectively. The interval angle within the non-ideal angle error of the wireless channel is uniformly selected as:

[0055]

[0056] wherein and are the sample numbers of and respectively, , . Thus is under the worst-case CSI (Channel State Information). The signal-to-noise ratio can be re-expressed as:

[0057]

[0058] Step 4: Use the semantic maximization and minimization method to solve the optimal digital beamformer (which can be recorded as ), and the optimal metasurface phase shift matrix (which can be recorded as ).

[0059] In one embodiment, in the process of solving the optimal digital beamformer using the semantic maximization-minimization method, a semantic proxy function is used to transform the untractable quasi-convexity in the digital optimization subproblem into a tractable quasi-convexity.

[0060] Specifically, first use the given metasurface phase shift matrix Designing a Digital Beamformer , the corresponding digital optimization sub-problem can be expressed as:

[0061]

[0062]

[0063] in, , the goal is to convert the unprocessable quasi-convex is transformed into a tractable quasi-convex, thereby solving the corresponding numerical optimization subproblem. In the following proposition, Semantic proxy function:

[0064] Proposition 1 (Semantic Proxy Function): Given a stationary point , quasi-convex It can be approximated as ,Right now:

[0065]

[0066] in, And the intermediate parameters:

[0067]

[0068] This Proposition 1 can be proved by sufficient experimental examples and curve fitting. By using Proposition 1, As A simple function, namely:

[0069]

[0070] However, the denominator With the form of sum and ratio, the corresponding digital optimization subproblem is still difficult to solve. Therefore, a secondary transformation is used for processing, namely:

[0071]

[0072] in, is an auxiliary variable introduced, and Therefore, the objective function can be further rewritten as Finally, the fractional semantic energy efficiency form is processed using the Dinkelbach method (an efficient algorithm for solving fractional programming problems). To elaborate, an additional variable , the fractional Sem-EE is converted into a non-fractional function, namely:

[0073]

[0074] Obviously, this is a given and The non-homogeneous separable QCQP (Quadratic ConstrainedQuadratic Programming) problem can be solved efficiently using CVX (a commonly used Matlab convex optimization solver). and Can be updated to:

[0075]

[0076] Among them, the superscript Indicates the number of iterations. In this iterative framework, the digital beamformer can be found .

[0077] In the pair After optimization, the focus is on designing the SM metasurface phase shift matrix , and its corresponding simulation optimization subproblem is given by:

[0078]

[0079] Obviously, due to the multiple constraints of quality of service (QoS) and the unit module constraints of meta-atoms, solving the above simulation optimization sub-problems is challenging. Therefore, we first add multiple QoS constraints to the non-negative Lagrange multiplier The simulation optimization subproblem is then rewritten as:

[0080]

[0081] Among them, the intermediate parameters are as follows:

[0082]

[0083] Subsequently, we turned to the CCD (conjugate gradient method) algorithm to process the unit module constraints of the meta-atom, and thus obtained the optimal closed-form solution Specifically, you can first Expands to:

[0084]

[0085] in, It should be noted that due to the Hermitian matrix The unit module constraints of and meta-atoms make the second equation valid. Therefore, Can be decomposed into N These scalar subproblems can be updated using the following CCD framework:

[0086]

[0087] Here Then, solve the above equation A scalar quantum problem. Ignoring the constant term, It can be expressed as:

[0088]

[0089] Obviously, the optimal closed-form solution is:

[0090]

[0091] However, the above CCD algorithm is based on a fixed non-negative Lagrange multiplier Therefore, we turn to find the optimal , which can be obtained from the following complementary relaxation conditions:

[0092]

[0093] in, . K-dichotomy can be used to search for the optimal For the sake of simplicity, we can refer to the existing K-dichotomy search for similar understanding. Finally, by and Perform iterative optimization to obtain the optimal , to obtain the optimal metasurface phase shift matrix .

[0094] In one embodiment, a communication system is provided, comprising a base station configured with a digital antenna, a stacked metasurface, and multiple user-end devices configured with single antennas. The communication system utilizes an optimal digital beamformer and an optimal metasurface phase shift matrix to optimize semantic communication. Acquiring the optimal digital beamformer and the optimal metasurface phase shift matrix includes the following steps:

[0095] setting a minimum semantic rate target, a minimum semantic similarity threshold, and a maximum transmit power for the communication system;

[0096] constructing, under a non-ideal angle error condition of a wireless channel, an optimization model to be solved with the digital beamformer and the metasurface phase shift matrix as optimization variables, taking the minimum semantic rate target, the minimum semantic similarity threshold, the maximum transmit power, and amplitude normalization of each reflection unit of the layered metasurface as constraint conditions, and taking semantic energy efficiency maximization of the communication system as an objective function;

[0097] constructing a non-ideal channel model based on angle error and parameter discretization of the non-ideal angle error of the wireless channel;

[0098] solving the optimal digital beamformer by using a semantic maximization minimization method, and solving the optimal metasurface phase shift matrix by using a cyclic coordinate descent algorithm;

[0099] performing semantic communication optimization on the communication system by using the optimal digital beamformer and the optimal metasurface phase shift matrix.

[0100] The above communication system sets a minimum semantic rate target, a minimum semantic similarity threshold, and a maximum transmit power, constructs, under a non-ideal angle error condition of a wireless channel, an optimization model to be solved with the digital beamformer and the metasurface phase shift matrix as optimization variables, taking the minimum semantic rate target, the minimum semantic similarity threshold, the maximum transmit power, and amplitude normalization of each reflection unit of the layered metasurface as constraint conditions, and taking semantic energy efficiency maximization of the communication system as an objective function, constructs a non-ideal channel model based on angle error and parameter discretization of the non-ideal angle error of the wireless channel, and finally solves the optimal digital beamformer by using a semantic maximization minimization method and solves the optimal metasurface phase shift matrix by using a cyclic coordinate descent algorithm, for semantic communication optimization of the communication system, thereby achieving semantic energy efficiency maximization of the layered metasurface in communication applications.

[0101] In one embodiment, in the process of solving the optimal digital beamformer by using a semantic maximization minimization method, a semantic proxy function is used to convert an unprocessable quasi-convex in a digital optimization sub-problem into a processable quasi-convex.

[0102] It can be understood that specific limitations on the above communication system can be referred to the corresponding limitations of the high-energy-efficiency semantic communication method based on the layered metasurface in the above text, which will not be repeated here.

[0103] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the following processing steps when executing the computer program: setting a minimum semantic rate target, a minimum semantic similarity threshold, and a maximum transmission power of the communication system; taking the maximization of the semantic energy efficiency of the communication system as the objective function, under the condition of non-ideal angular error of the wireless channel, taking the minimum semantic rate target, the minimum semantic similarity threshold, the maximum transmission power, and the amplitude normalization of each reflection unit of the stacked metasurface as constraints, constructing an optimization model to be solved with a digital beamformer and a metasurface phase shift matrix as optimization variables; constructing a non-ideal channel model based on angular error and discretizing the non-ideal angular error of the wireless channel; using the semantic maximization minimization method to solve the optimal digital beamformer, and using the cyclic coordinate descent algorithm to solve the optimal metasurface phase shift matrix; using the optimal digital beamformer and the optimal metasurface phase shift matrix to perform semantic communication optimization on the communication system.

[0104] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the above-described method embodiments. Any reference to memory, storage, database, or other media used in the various embodiments provided herein may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus DRAM (RDRAM), and DDR DRAM.

[0105] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0106] The above embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that a person skilled in the art would be able to make numerous modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of the present invention. Therefore, the scope of the present invention shall be determined by the appended claims.

Claims

1. A high-energy-efficiency semantic communication method based on stacked metasurfaces, characterized in that: Applied to a communication system comprising a base station configured with a digital antenna, a stacked metasurface, and a plurality of user-end devices configured with a single antenna; The energy-efficient semantic communication method comprises the following steps: Setting the minimum semantic rate target, minimum semantic similarity threshold, and maximum transmit power for the communication system; Taking maximizing the semantic energy efficiency of the communication system as the objective function, under the condition of non-ideal angle error of the wireless channel, and with the minimum semantic rate target, the minimum semantic similarity threshold, the maximum transmit power, and the amplitude normalization of each reflective unit of the stacked metasurface as constraints, an optimization model to be solved is constructed with a digital beamformer and a metasurface phase shift matrix as optimization variables; Construct a non-ideal channel model based on angle error and discretize the non-ideal angle error of the wireless channel into parameters; The optimal digital beamformer is solved using the semantic maximization-minimization method, and the optimal metasurface phase shift matrix is ​​solved using the cyclic coordinate descent algorithm. The communication system is optimized for semantic communication using an optimal digital beamformer and an optimal metasurface phase shift matrix.

2. The high-energy-efficiency semantic communication method based on stacked metasurfaces according to claim 1 is characterized in that: In the process of solving the optimal digital beamformer using the semantic maximization-minimization method, the semantic proxy function is used to transform the untractable quasi-convexity in the digital optimization subproblem into a tractable quasi-convexity.

3. A communication system, characterized in that: The system includes a base station configured with a digital antenna, a stacked metasurface, and multiple user-end devices configured with single antennas. The communication system utilizes an optimal digital beamformer and an optimal metasurface phase shift matrix to optimize semantic communication. Acquiring the optimal digital beamformer and the optimal metasurface phase shift matrix includes the following steps: Setting the minimum semantic rate target, minimum semantic similarity threshold, and maximum transmit power for the communication system; Taking maximizing the semantic energy efficiency of the communication system as the objective function, under the condition of non-ideal angle error of the wireless channel, and with the minimum semantic rate target, the minimum semantic similarity threshold, the maximum transmit power, and the amplitude normalization of each reflective unit of the stacked metasurface as constraints, an optimization model to be solved is constructed with a digital beamformer and a metasurface phase shift matrix as optimization variables; Construct a non-ideal channel model based on angle error and discretize the non-ideal angle error of the wireless channel into parameters; The optimal digital beamformer is solved using the semantic maximization-minimization method, and the optimal metasurface phase shift matrix is ​​solved using the cyclic coordinate descent algorithm. The communication system is optimized for semantic communication using an optimal digital beamformer and an optimal metasurface phase shift matrix.

4. The communication system according to claim 3, wherein: In the process of solving the optimal digital beamformer using the semantic maximization-minimization method, the semantic proxy function is used to transform the untractable quasi-convexity in the digital optimization subproblem into a tractable quasi-convexity.

5. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the energy-efficient semantic communication method according to claim 1 or 2 are implemented.

Images