Massive MIMO (Multiple Input Multiple Output) communication and sensing integrated multi-target green linear precoding method and system

Abstract

The invention discloses a Massive MIMO (Multiple Input Multiple Output) communication and sensing integrated multi-target green linear precoding method and system, relates to the technical field of wireless communication, and aims to solve the problem that two conflicting targets, namely communication energy efficiency and sensing energy efficiency, cannot be optimized at the same time in the conventional method. The method is characterized by comprising the following steps: deriving a perception energy efficiency expression under multiple perception targets based on the Cramer-Rao bound, and establishing a communication perception integrated communication energy efficiency and perception energy efficiency system model; the method comprises the following steps: establishing a pre-coding method for maximizing communication energy efficiency and sensing energy efficiency of joint power distribution when multiple targets and multiple users coexist, establishing a multi-target linear pre-coding optimization model of joint power distribution, and taking a pre-coding method with a closed-form solution as a selectable pre-coding scheme; the optimization target is to maximize the communication energy efficiency and the sensing energy efficiency of the ISAC system; and a multi-target optimization problem is solved through a multi-target quantum energy valley mechanism, maximization of communication energy efficiency and sensing energy efficiency is realized, and an optimal ISAC multi-target green linear precoding scheme is obtained.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and more specifically, to a Massive MIMO integrated sensing multi-objective green linear precoding method and system. Background Technology

[0002] In recent years, Integrated Sensing and Communication (ISAC) technology has become a key component in the deployment of next-generation wireless communication networks. ISAC base stations enable spectrum sharing between communication devices and sensing targets; however, with the increasing variety and number of devices, this undoubtedly poses challenges to system energy efficiency. Simultaneously, next-generation wireless networks advocate for green and carbon-reducing development, requiring improved energy efficiency and reduced power consumption in ISAC systems to achieve green development. Massive Multiple Input Multiple Output (MIMO) improves spectral efficiency through spatial diversity, meeting the multi-device access requirements of integrated systems.

[0003] Deploying Massive MIMO-based integrated communication and sensing systems requires addressing downlink inter-device interference while improving spatial gain. Integrated joint beamforming and precoding techniques reduce inter-device signal interference by designing the base station transmitter, thereby improving both communication and sensing performance. The communication performance of an ISAC system is typically measured by the signal-to-interference-plus-noise ratio (SINR) of the communication user, with the Crammé-Rao bound (CRB) serving as a lower bound for system parameter estimation. Existing integrated communication and sensing joint beamforming and precoding designs can be categorized into communication-centric, radar-centric, and joint designs. Joint designs can guarantee good parameter estimation performance while maintaining communication rates, but they also increase system complexity and energy consumption. Furthermore, the increased transmit power of integrated base stations to meet the service demands of multiple devices inevitably increases energy costs, contradicting the emerging requirements of carbon neutrality and environmental sustainability in next-generation wireless networks. Existing energy-saving communication designs lack consideration for sensing functions and cannot be directly applied to ISAC. Therefore, energy-saving performance indicators for integrated communication and sensing designs need to be considered.

[0004] A search of existing technical literature revealed that Jiaqi Zou et al. [1] This paper aims to maximize the energy efficiency of communication and sensing in ISAC systems separately, and expresses the trade-off between communication and sensing energy consumption with a Pareto front-end as the optimization objective. However, it only considers MIMO with a single sensing target and lacks a massive MIMO sensing system with multiple far-field targets, failing to optimize both communication and sensing performance simultaneously. (Minghe Zhu et al.)[2] For the ISAC system architecture of UAVs equipped with passive intelligent reflective surfaces, the goal is to maximize the energy efficiency of the communication system while ensuring the quality of user communication services and the target perception signal-to-noise ratio. However, this approach only considers the perception signal-to-noise ratio as a constraint and does not take into account the system's perception energy efficiency, thus failing to improve the system's perception performance.

[0005] Existing research indicates that optimizing the energy efficiency of communication and sensing in ISAC systems is a Pareto front-end optimization problem. However, current integrated communication and sensing precoding methods cannot simultaneously optimize the conflicting objectives of communication and sensing energy efficiency, necessitating the design of novel multi-objective algorithms. Summary of the Invention

[0006] The technical problem to be solved by this invention is:

[0007] Existing integrated communication and sensing precoding methods cannot simultaneously optimize the conflicting objectives of communication energy efficiency and sensing energy efficiency.

[0008] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows:

[0009] This invention designs a precoding method for maximizing communication and sensing energy efficiency in joint power allocation for multi-objective, multi-user coexistence scenarios. It provides an expression for sensing energy efficiency in multi-sensory target scenarios, selects a precoding scheme with a closed-form solution, and designs a multi-objective quantum energy valley mechanism to solve this multi-objective optimization problem. This method overcomes the application limitations of existing Massive MIMO integrated sensing precoding methods, maximizes communication and sensing energy efficiency, and provides ideas and methods for the green and energy-saving design and engineering implementation of future integrated sensing systems.

[0010] This invention provides a Massive MIMO integrated sensing multi-objective green linear precoding method, comprising the following steps:

[0011] Step 1: Based on the Cramer-Rao bound, derive the expression for sensing energy efficiency under multi-sensing targets, and establish an integrated communication and sensing energy efficiency system model;

[0012] Step 2: Establish a multi-objective linear precoding optimization model for joint power allocation, and select precoding methods with closed-form solutions as the precoding schemes. The optimization objective is to maximize the communication energy efficiency and sensing energy efficiency of the ISAC system.

[0013] Step 3: Initialize the particle population of the multi-objective quantum energy valley mechanism, put the non-dominated solution into the storage repository, construct a hypercube grid according to the objective function value corresponding to the external storage repository, and use roulette wheel to select the non-dominated solution with the highest leadership density as the optimal particle;

[0014] Step 4: Calculate the enrichment energy level of the particle from the objective function value, determine the decay type of the particle based on the neutron enrichment level and the stability level of the particle, and generate new particles.

[0015] Step 5: Map the quantum position of the newly generated particle to a position and calculate the objective function value. Determine the updated particle according to the selection mechanism and update the non-dominated solutions in the repository.

[0016] Step 6: Repeat steps 4 and 5 until the maximum number of iterations is reached; output the global optimal quantum position of the multi-objective quantum energy valley mechanism, and after mapping transformation, obtain the optimal ISAC multi-objective green linear precoding scheme.

[0017] Furthermore, step one includes the following steps:

[0018] Considering the scenario of a monostatic radar integrating dual functions, a large-scale MIMO sensing-integrated base station includes... Root transmitting antenna and Root receiver antenna, The array has an element spacing of [missing information]. A uniform linear array, within the communication and sensing range of the base station. A far-field sensing target and With multiple single-antenna users existing simultaneously, the sensing and communication signals transmitted by the base station are within the same time-frequency resource block, and the channel state information is perfectly known; the base station downlink joint beamforming precoding matrix is... , Indicates the first Precoded vectors for each user , Indicates the first Beamforming vector of a sensing target , Indicates conjugate transpose;

[0019] No. The signal received by each communication user is , , Indicates the sampling point. Indicates the maximum number of sampling points. For base station to user The channel vector, Indicates the first During the second sampling, the base station transmitted to the... Communication signals of individual users The Gaussian white noise has zero mean and a variance of . ,user The signal-to-noise ratio of the received signal is expressed as , If we take the absolute value, then the communication energy efficiency of the integrated sensing system is expressed as: , This indicates the total transmit power of the base station;

[0020] The base station uses the echo signal to perform direction-of-arrival estimation on the sensed target, and the target response matrix. , ,in This represents the transpose of a vector. Indicates the first A far-field sensing target amplitude, Indicates the first The guidance vector of a perceived target. Indicates the first An angle of arrival, Indicates the spacing between array elements. The signal wavelength is used to determine the angle of the sensing target in the integrated sensing base station. The estimated Craméro boundary is composed of The calculation yielded, where Represents the target response matrix For vectors Taking the first derivative, the sensing energy efficiency of the synesthetic system is expressed as: ,in, Indicates matrix trace. This indicates that the matrix is ​​inverted. This indicates the maximum number of sampling points.

[0021] Furthermore, step two includes the following steps:

[0022] user When using maximum ratio transmission and regular zero-forcing precoding, the normalized precoding vectors are as follows: , , Represents the magnitude of a vector. Represents the channel matrix. As a regularization factor, for An identity matrix of dimension; for the perceived target Considering the use of maximum ratio transmission precoding and null space zero-forcing precoding, the normalized precoding vectors are respectively , , for An identity matrix of dimension; the downlink joint beamforming precoding matrix of an integrated base station is represented by a normalized precoding vector as follows: ,in, and These represent the base station's communication with the user. and perceived target Allocated transmission power, and These represent the base station's communication with the user. and perceived target The normalized precoding vector, , , This indicates the base station power allocation method, for the total transmit power of the base station. ,in, Indicates the transmit power threshold. This indicates the maximum transmit power of the base station, and the total power of the transmitter is... ,at this time, Indicates energy conversion efficiency. This represents antenna power consumption; therefore, communication energy efficiency is expressed as... Perceived energy efficiency is expressed as ;

[0023] Establish a multi-objective linear precoding optimization model for joint power allocation, with the optimization objective being: Constraint 1 is satisfied as follows: Constraint 2 is Constraint 3 is .

[0024] Furthermore, step three includes the following steps:

[0025] First, set the number of particles to... The maximum number of iterations is The dimension of the particle search space is , ;No. Sub-particles The quantum position is , , ,particle No. The quantum position of the dimension is initialized as ,in, for Uniform random numbers between particles The Dimensional position is The mapping relationship is ,in, express Random numbers that follow a uniform distribution between , when At that time, particles The Wei represents communication users Precoding selection, , ;when At that time, particles The Dimension represents the perceived target Precoding selection, , ;when At that time, the first The first particle The dimension represents the downlink power allocation of the integrated base station, and the power allocation vector is... satisfy ,in ;when At that time, the first The first particle Wei represents the base station transmission threshold. , ;

[0026] The first Sub-particles Substitute the position into the two objective functions , Calculate the corresponding objective function value; if for the th , particles, simultaneously satisfying: , If at least one strict inequality holds, then we say that the inequality is true in the th case. Generational particles Dominant Particle If initially all other particles cannot dominate the particle... So, the particles The location is placed in an external repository that stores non-dominated solutions, and the repository size is [size missing]. , Indicates the first repository size ;

[0027] For objective function 1 Perform mesh generation, and set the mesh boundary spread factor to... The number of grids is For objective function 2 Similarly, the grid edges are divided to form a grid containing... A hypercube network of 1,000 grid cells is constructed. The specific location of each non-dominated solution within the established hypercube network is determined based on the objective function value, yielding the number of non-dominated solutions in each grid cell. Then, a roulette wheel selection process is used to select the non-dominated solution with the highest leadership density as the optimal particle. The leadership grid density is calculated using the following formula: , Indicates the first The leader grid density of the grid cell, Indicates the first The number of non-dominated solutions in a grid cell, of which , , The leader density factor is represented by the quantum position of the optimal particle. .

[0028] Furthermore, step four includes the following steps:

[0029] calculate The second iteration Neutron enrichment levels of individual particles , No. The maximum enrichment energy level in the next iteration is denoted as The minimum value is denoted as , No. The nearest neighbor particle in the next iteration is represented as The central particle is Nearest neighbor particles are The average of the non-dominated solutions from the external repositories is obtained. Indicates rounding up. for Calculate the uniformly random number between the two values. The Euclidean distance of a particle from other particles The one with the smallest distance The average of the particles is obtained by taking the particle size. The central particle, for Random integers between [a certain range];

[0030] Calculate the first based on enrichment energy level The enrichment boundary of the next iteration , No. The second iteration The stability of a particle is expressed as ,in, for A uniform random number between, the first The iteration factor for the next iteration is Compare the neutron enrichment level with the boundary value; if Then the particle undergoes a decay process; if , indicating particle occurrence and decay, in which Represents particles stability, express A set of random numbers that follow a uniform distribution and are not repeated. middle as well as Each element forms a vector and ,in At this time, the particles occur Two particles are generated during decay: and ,in, , , and Representing the nearest neighbor particles and the central particle respectively Wei; if , indicating that it has occurred decay, particle release The ray forms two particles, and after decomposition, the first particle... The quantum rotation angles are updated to: , ,in, and They are respectively The particles are updated using a simulated quantum rotation gate, where random numbers following a uniform distribution are used. occur The quantum positions at the time of decay are respectively , ;

[0031] if This indicates that the particle When the enrichment level is less than the boundary value, the optimal particle either receives or emits electrons. The quantum rotation angle is updated to Update quantum position using simulated quantum rotation gate .

[0032] Furthermore, step five includes the following steps:

[0033] For the g+1 generation particles ,if The quantum positions of the two new particles produced by the decay process of the particle will then be determined. and Mapped to position and Calculate the dominance relationship, if Dominate ,and Simultaneously dominates the particles before the iteration Then the (g+1)th generation particle The quantum position is ;like Dominate ,and Simultaneously dominates the particles before the iteration Then the (g+1)th generation particle The quantum position is Otherwise, do not update, that is ;

[0034] if Then compare the new particle with the g-th generation particle. The dominance relationship will determine the quantum position of the new particle. Mapped to position And calculate all objective function values, if Dominate Then the (g+1)th generation particle quantum position ;otherwise, ;

[0035] Calculate the non-dominated solutions of the (g+1)th generation population and transfer them to an external repository; if the number of non-dominated solutions in the repository exceeds its storage capacity after updating, then... Then calculate the deleted grid density corresponding to all non-dominated solutions in the external repository: , Indicates the first Grid density for cell deletion, Indicates the first The number of non-dominated solutions in a grid cell This involves deleting the density factor, then using a roulette wheel to select the grid with the highest density, and randomly deleting non-dominated solutions from the grid with the highest density until the storage limit is reached, thus becoming the [number]th [cell]. The repository is used to calculate the leader grid density and obtain the updated optimal particle. The quantum position of the optimal particle is .

[0036] This invention provides a Massive MIMO integrated sensing multi-objective green linear precoding system. The system has a program module corresponding to the steps of any of the above-described technical solutions, and executes the steps in the Massive MIMO integrated sensing multi-objective green linear precoding method during runtime.

[0037] This invention provides a computer-readable storage medium storing a computer program configured to, when invoked by a processor, implement the steps of the Massive MIMO integrated sensing multi-objective green linear precoding method described in any of the above technical solutions.

[0038] Compared with the prior art, the beneficial effects of the present invention are:

[0039] (1) To address the problem that existing precoding methods for Massive MIMO that integrate communication and sensing cannot simultaneously improve communication and sensing energy efficiency, a more effective green linear precoding method based on the multi-objective quantum energy valley mechanism for Massive MIMO is designed. A direct coding scheme with a closed-form solution is designed for multi-objective multi-user coexistence scenarios, and the multi-objective quantum energy valley algorithm is used to solve the multi-objective hybrid optimization problem.

[0040] (2) The Massive MIMO sensing-integrated precoding method designed in this invention designs a multi-objective linear precoding optimization model for joint power allocation, gives the sensing energy efficiency expression for multiple sensing objectives based on the Cramer-Rao bound, establishes a Pareto optimization model to maximize sensing and communication energy efficiency, adopts a linear precoding method with closed-form solution to jointly allocate power and improve system energy efficiency, and the designed multi-objective quantum energy valley mechanism can solve the optimization objectives with high precision, realizing the joint optimization of the two conflicting objectives of communication energy efficiency and sensing energy efficiency.

[0041] (3) A multi-objective quantum energy valley algorithm was designed by combining the quantum energy valley mechanism with the grid mechanism, leader selection, and archiving mechanism. Simulation experiments proved the effectiveness of the ISAC green linear precoding method based on the multi-objective quantum energy valley mechanism. The precoding method with closed-form solution achieves high-precision estimation while improving communication and sensing energy efficiency, which can reduce the difficulty of hardware implementation and provide ideas and methods for the green and energy-saving design and engineering implementation of future integrated sensing systems. Attached Figure Description

[0042] Figure 1 This is a flowchart of the Massive MIMO synesthetic integrated multi-objective green linear precoding method in an embodiment of the present invention;

[0043] Figure 2 The image shows a comparison of the Pareto optimal frontier obtained by applying the multi-objective quantum energy valley algorithm, the multi-objective gray wolf algorithm, and the multi-objective particle swarm algorithm to optimize Massive MIMO integrated communication and sensing energy efficiency in this embodiment of the invention.

[0044] Figure 3 This is a comparison diagram of the Pareto optimal frontier obtained by optimizing the precoding scheme using the multi-objective quantum energy valley algorithm in this embodiment of the invention, and the optimal precoding scheme obtained by optimizing two objective functions using the energy valley algorithm respectively, corresponding to the values ​​of the two objective functions. Detailed Implementation

[0045] To enable those skilled in the art to better understand the present invention, exemplary embodiments or examples of the present invention will be described below in conjunction with the accompanying drawings. Obviously, the described embodiments or examples are merely some, not all, of the embodiments or examples of the present invention. All other embodiments or examples obtained by those skilled in the art based on the embodiments or examples of the present invention without inventive effort should fall within the scope of protection of the present invention.

[0046] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0047] Combination Figure 1 As shown, this invention provides a Massive MIMO integrated sensing multi-objective green linear precoding method, comprising the following steps:

[0048] Step 1: Based on the Cramer-Rao bound, derive the expression for sensing energy efficiency under multi-sensing targets, and establish an integrated communication and sensing energy efficiency system model. The specific steps are as follows:

[0049] Considering the scenario of a monostatic radar integrating dual functions, a large-scale MIMO sensing-integrated base station includes... Root transmitting antenna and Root receiver antenna, The array has an element spacing of [missing information]. A uniform linear array. Within the communication and sensing range of the base station... A far-field sensing target and A single-antenna user exists simultaneously. , The sensing and communication signals transmitted by the base station are located within the same time-frequency resource block, and the channel state information is perfectly known. The base station downlink joint beamforming precoding matrix is ​​as follows: , Indicates the first Precoded vectors for each user , Indicates the first Beamforming vector of a sensing target .

[0050] No. The signal received by each communication user is , , Indicates the sampling point. Indicates the maximum number of sampling points. For base station to user The channel vector, Indicates the first During the second sampling, the base station transmitted to the... Communication signals of individual users The Gaussian white noise has zero mean and a variance of . ,user The signal-to-noise ratio of the received signal can be expressed as: , If we take the absolute value, then the communication energy efficiency of the integrated sensing system can be expressed as: , This indicates the total transmission power of the base station.

[0051] The base station uses the echo signal to perform direction-of-arrival estimation on the sensed target, and the target response matrix. , ,in This represents the transpose of a vector. Indicates the first A far-field sensing target amplitude, Indicates the first The guidance vector of a perceived target. Indicates the first An angle of arrival, Indicates the spacing between array elements. The signal wavelength is used to determine the angle of the sensing target in the integrated sensing base station. The estimated Craméro boundary can be determined by The calculation yielded, where This indicates the conjugate transpose. Represents the target response matrix For vectors Taking the first derivative, the sensing energy efficiency of the synesthetic system can be expressed as: ,in, Indicates matrix trace. This indicates that the matrix is ​​inverted. This indicates the maximum number of sampling points.

[0052] Step 2: Establish a multi-objective linear precoding optimization model for joint power allocation. Precoding methods with closed-form solutions are selected as alternative precoding schemes. The optimization objective is to maximize the communication and sensing energy efficiency of the ISAC system, which is a Pareto optimization problem. The specific steps are as follows:

[0053] Communication users Maximum ratio transmission precoding and regular zero-forcing precoding can be selected as linear precoding methods. When using maximum ratio transmission and regular zero-forcing precoding, the normalized precoding vectors are respectively , , Represents the magnitude of a vector. Represents the channel matrix. As a regularization factor, for An identity matrix of dimension; for the perceived target Considering the use of maximum ratio transmission precoding and null space zero-forcing precoding, the normalized precoding vectors are respectively , , for An identity matrix of dimension 1. The downlink joint beamforming precoding matrix of an integrated base station can be represented using normalized precoding vectors as follows: ,in, and These represent the base station's communication with the user. and perceived target Allocated transmission power, and These represent the base station's communication with the user. and perceived target The normalized precoding vector, , , This indicates the base station power allocation method, for the total transmit power of the base station. ,in, Indicates the transmit power threshold. This indicates the maximum transmit power of the base station, and the total power of the transmitter is... ,at this time, Indicates energy conversion efficiency. This represents antenna power consumption; therefore, communication energy efficiency can be expressed as... Perceived energy efficiency can be expressed as .

[0054] A multi-objective linear precoding optimization model for joint power allocation is established. The optimization scheme for system energy efficiency has two objectives: the first is to maximize the communication energy efficiency of the ISAC system, i.e. The second is to maximize perceived energy efficiency, that is... The optimization objective is then... Constraints must be met That is, the user's average transmission rate is greater than or equal to the minimum transmission rate. Constraint 2 is That is, transmit power threshold To meet the transmit power limit of the integrated base station, constraint 3 is... That is, the Fisher information matrix is ​​a positive semidefinite matrix.

[0055] Step 3: Initialize the particle population of the multi-objective quantum energy valley mechanism, release the non-dominated particles into the storage repository, construct a hypercube grid based on the objective function value corresponding to the external storage repository, and use roulette wheel selection to choose the optimal particle. The specific steps are as follows:

[0056] First, set the number of particles to... The maximum number of iterations is The dimension of the particle search space is , . No. Sub-particles The quantum position is , , .particle No. The quantum position of the dimension is initialized as ,in, for Uniformly random numbers between particles. The Dimensional position is The mapping relationship is ,in, express Random numbers that follow a uniform distribution between [a certain number of elements]. When At that time, particles The Wei represents communication users Precoding selection, , ;when At that time, particles The Dimension represents the perceived target Precoding selection, , ;when At that time, the first The first particle The dimension represents the downlink power allocation of the integrated base station, and the power allocation vector is... satisfy ,in ;when At that time, the first The first particle Wei represents the base station transmission threshold. , .

[0057] The first Sub-particles Substitute the position into the two objective functions , Calculate the corresponding objective function value. If for the ... , particles, simultaneously satisfying: , If at least one strict inequality holds, then we say that the inequality is true in the th case. Generational particles Dominant Particle If initially all other particles cannot dominate the particle... So, the particles The location is placed in an external repository that stores non-dominated solutions, and the repository size is [size missing]. That is, the maximum amount of data stored in the repository A non-dominated solution, Indicates the first repository size .

[0058] When constructing the hypercube mesh, an adaptively expanding mesh boundary is first created based on the objective function values ​​of each objective function in the external repository, and then the mesh is generated. For objective function 1, i.e. The boundary spread factor of the mesh is set to , Multiplying by the difference between the maximum and minimum values ​​represents the boundary expansion value. The upper bound of the grid is the maximum objective function value plus the boundary expansion value, and the lower bound is the minimum objective function value minus the boundary expansion value. The number of grid cells is... That is, dividing the grid side length into Parts; for objective function 2, i.e. Similarly, the grid edges are divided to form a grid containing... A hypercube network of individual grid cells is constructed. The specific location of each non-dominated solution within the established hypercube grid is determined based on the objective function value. For the optimal particle, the leader grid density needs to be calculated. This is done by first obtaining the number of non-dominated solutions in each grid cell, and then using a roulette wheel selection method to choose the non-dominated solution with the highest leader density as the optimal particle. The leader grid density calculation formula is as follows: , Indicates the first The leader grid density of the grid cell, Indicates the first The number of non-dominated solutions in a grid cell, of which , , The leader density factor is represented by the quantum position of the optimal particle. .

[0059] Step 4: Averaging the objective function values ​​yields the enrichment energy level of the particles. Based on the neutron enrichment level and stability level of the particles, the decay type of the particles is determined, and new particles are generated. The specific steps are as follows:

[0060] According to the The objective function value of the next iteration particle , Take the modulo value and use it as the first... The neutron-enriched energy levels of each particle, i.e. , No. The maximum enrichment energy level in the next iteration is denoted as The minimum value is denoted as . No. The nearest neighbor particle in the next iteration is represented as The central particle is Nearest neighbor particles are formed by The average of the non-dominated solutions from the external repositories is obtained. Indicates rounding up. for A uniformly random number between [a certain number of elements]. Calculate the [number of elements]. The Euclidean distance of a particle from other particles The one with the smallest distance The average of the particles is obtained by taking the particle size. The central particle, for A random integer between [a certain range].

[0061] Calculate the first based on enrichment energy level The enrichment boundary of the next iteration , No. The second iteration The stability of a particle is expressed as ,in, for A uniform random number between, the first The iteration factor for the next iteration is Compare the neutron enrichment level with the boundary value; if Then the particle undergoes a decay process; if , indicating particle occurrence and decay, in which Represents particles stability, express A set of random numbers that follow a uniform distribution and are not repeated. middle as well as Each element forms a vector and ,in At this time, the particles occur During decay, two particles are generated, namely... and ,in, , , and Representing the nearest neighbor particles and the central particle respectively Wei; if , indicating that it has occurred decay, particle release The ray forms two particles, and after decomposition, the first particle... The quantum rotation angles are updated to: , ,in, and They are respectively The particles are updated using a simulated quantum rotation gate, where random numbers following a uniform distribution are used. occur The quantum positions at the time of decay are respectively , .

[0062] if This indicates that the particle When the enrichment level is less than the boundary value, the optimal particle either receives or emits electrons. The quantum rotation angle is updated to Update quantum position using simulated quantum rotation gate .

[0063] Step 5: Map the quantum position of the newly generated particle to a position and calculate the objective function value. Determine the updated particle according to the selection mechanism and update the non-dominated solution in the repository. The specific steps are as follows:

[0064] For the g+1 generation particles ,if The quantum positions of the two new particles produced by the decay process of the particle will then be determined. and Mapped to position and Calculate the dominance relationship, if Dominate ,and Simultaneously dominates the particles before the iteration Then the (g+1)th generation particle The quantum position is ;like Dominate ,and Simultaneously dominates the particles before the iteration Then the (g+1)th generation particle The quantum position is Otherwise, do not update, that is .

[0065] if At this point, it is only necessary to compare the new particle with the g-th generation particle. The dominance relationship will determine the quantum position of the new particle. Mapped to position And calculate all objective function values, if Dominate Then the (g+1)th generation particle quantum position ;otherwise, .

[0066] Calculate the non-dominated solutions for the (g+1)th generation population and add them to an external repository. If the number of updated non-dominated solutions in the repository exceeds its storage capacity... Then calculate the deleted mesh density corresponding to all non-dominated solutions in the external repository. The formula for calculating the deleted mesh density is: , Indicates the first Grid density for cell deletion, Indicates the first The number of non-dominated solutions in a grid cell, of which , , This involves deleting the density factor, then using a roulette wheel to select the grid with the highest density, and randomly deleting non-dominated solutions from the grid with the highest density until the storage limit is reached, thus becoming the [number]th [cell]. The repository is used to calculate the leader grid density and obtain the updated optimal particle. The quantum position of the optimal particle is .

[0067] Step 6: Determine if the maximum number of iterations has been reached. If it is not achieved, then Return to step four; if the target is reached, terminate the loop iteration and output the global optimal quantum position of the multi-objective quantum energy valley mechanism. After mapping transformation, the global optimal position is the optimal ISAC multi-objective green linear precoding scheme.

[0068] The Massive MIMO integrated sensing multi-objective green linear precoding method (algorithm) proposed in this invention is the underlying technical core of this invention, and various products can be derived based on the algorithm.

[0069] Based on the method proposed in this invention, a Massive MIMO integrated sensing multi-objective green linear precoding system is developed using a programming language. This system has program modules corresponding to the steps of the above-mentioned technical solution, and executes the steps in the Massive MIMO integrated sensing multi-objective green linear precoding method during runtime.

[0070] The developed system (software) computer program is stored on a computer-readable storage medium. This computer program is configured to implement the steps of the Massive MIMO integrated sensing multi-objective green linear precoding method described above when invoked by a processor. In other words, the invention is materialized on a carrier, becoming a computer program product.

[0071] Various implementations of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, application-specific integrated circuits (ASICs), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include: implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0072] The computational programs (also referred to as programs, software, software applications, or code) of this invention include machine instructions of a programmable processor and can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, device, and / or apparatus (e.g., disk, optical disk, memory, programmable logic device PLD) for providing machine instructions and / or data to a programmable processor, including machine-readable media that receive machine instructions as machine-readable signals. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0073] The beneficial effects of the present invention will be described below with reference to specific embodiments and comparative examples.

[0074] Example 1

[0075] In this embodiment, the ISAC multi-objective green linear precoding method based on the multi-objective quantum energy valley mechanism designed in this invention is denoted as MOQEVO; the existing model includes: the ISAC multi-objective green linear precoding method based on the multi-objective particle swarm optimization algorithm, denoted as MOPSO. [3] The ISAC multi-objective green linear precoding method based on the multi-objective gray wolf optimization algorithm is denoted as MOGWO. [4] The ISAC green linear precoding method based on the energy valley optimization algorithm is denoted as EVO. [5] .

[0076] In the simulation experiment, consider a square simulated area with a side length of 500 m, with the base station located at the center of the area, at coordinates [coordinates missing]. The location center coordinates of the communication user Randomly generated within a 75 m square area, the target's center coordinates are sensed. Randomly generated within a 75 m square area, the maximum output power of the base station is... Number of communication users Perceive target number Number of transmitting and receiving antennas in an integrated base station Minimum transmission rate for communication users The element spacing of both the transmitting and receiving arrays is half a wavelength, enabling target sensing. amplitude Base station energy conversion efficiency Antenna power consumption The path loss of the communication channel was modeled using the 3GPP Urban Microcell model, with a carrier frequency of 1.9 GHz, a bandwidth of 20 MHz, and a noise variance of [missing information]. for Maximum number of sampling points The channel is a correlated Rayleigh fading channel. , , , Given the spatial correlation matrix, the channel state information is perfectly known. The regularization factor for RZF precoding is... .

[0077] The parameters for the multi-objective quantum energy valley mechanism are set as follows: number of particles and size of the storage reservoir. Maximum number of iterations Search space dimension In a hypercube mesh, the boundary spread factor , number of grids Leadership density factor Delete density factor .

[0078] Figure 2 A comparison of Pareto optimal frontiers obtained by applying the multi-objective quantum energy valley algorithm, the multi-objective gray wolf algorithm, and the multi-objective particle swarm optimization algorithm to optimize Massive MIMO integrated sensing communication and sensing energy efficiency. From Figure 2 As can be seen, the Pareto optimal solution set obtained by the multi-objective particle swarm optimization algorithm yields the lowest communication and sensing energy efficiency, while the Pareto optimal solution set obtained by the multi-objective quantum energy valley optimization algorithm yields the highest communication and sensing energy efficiency. Furthermore, for any Pareto optimal solution obtained by the multi-objective particle swarm optimization algorithm and the multi-objective gray wolf optimization algorithm, at least one Pareto optimal solution can be found in the Pareto optimal solution set obtained by the multi-objective quantum energy valley optimization algorithm. The designed quantum energy valley optimization algorithm can obtain a better-performing and more diversified joint power allocation precoding scheme, thereby improving the energy efficiency of the ISAC system.

[0079] Figure 3 This graph compares the Pareto optimal front obtained by applying the multi-objective quantum energy valley algorithm to optimize the precoding scheme with the optimal precoding schemes obtained by applying the energy valley algorithm to optimize communication energy efficiency and sensing energy efficiency, respectively, corresponding to the two objective function values. Figure 3 As can be seen, the two optimal solutions obtained by the energy valley algorithm are both dominated by some Pareto optimal solutions obtained by the multi-objective quantum energy valley algorithm, which demonstrates the superiority of the designed multi-objective quantum energy valley algorithm.

[0080] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

[0081] The documents cited in this invention include:

[0082] [1] "Energy-Efficient Beamforming Design for Integrated Sensingand Communications Systems" published by Jiaqi Zou et al. in "IEEE Transactions on Communications" (2024, 72(6):3766-3782)

[0083] [2] "Energy efficiency optimization for UAV-mounted RIS-aided integratedsensing and communication" published by Minghe Zhu et al. in "Chinese Journal of Aeronautics" (2025, 103920)

[0084] [3]Coello C et al. Handling Multiple Objectives with Particle SwarmOptimization. "IEEE Transactions on Evolutionary Computation", 2004, 8(3): 256-279.

[0085] [4] Seyedali Mirjalili et al. Multi-objective grey wolf optimizer : A novel algorithm for multi-criterion optimization. 《Expert Systems with Applications》, 2016, 47: 106-119.

[0086] [5] M. Azizi et al. Energy Valley Optimizer: A Novel Metaheuristic Algorithm for Global and Engineering Optimization. 《Scientific Reports》, 2023, 1(13):226.

Claims

1. A Massive MIMO synesthetic integrated multi-objective green linear precoding method, characterized in that, Includes the following steps: Step 1: Based on the Cramer-Rao bound, derive the expression for sensing energy efficiency under multi-sensing targets, and establish an integrated communication and sensing energy efficiency system model; Step 2: Establish a multi-objective linear precoding optimization model for joint power allocation, and select precoding methods with closed-form solutions as the precoding schemes. The optimization objective is to maximize the communication energy efficiency and sensing energy efficiency of the ISAC system. Step 3: Initialize the particle population of the multi-objective quantum energy valley mechanism, put the non-dominated solution into the storage repository, construct a hypercube grid according to the objective function value corresponding to the external storage repository, and use roulette wheel to select the non-dominated solution with the highest leadership density as the optimal particle; Step 4: Calculate the enrichment energy level of the particle from the objective function value, determine the decay type of the particle based on the neutron enrichment level and the stability level of the particle, and generate new particles. Step 5: Map the quantum position of the newly generated particle to a position and calculate the objective function value. Determine the updated particle according to the selection mechanism and update the non-dominated solutions in the repository. Step 6: Repeat steps 4 and 5 until the maximum number of iterations is reached; output the global optimal quantum position of the multi-objective quantum energy valley mechanism, and after mapping transformation, obtain the optimal ISAC multi-objective green linear precoding scheme.

2. The method according to claim 1, characterized in that, Step one includes the following steps: Considering the scenario of a monostatic radar integrating dual functions, a large-scale MIMO sensing-integrated base station includes... Root transmitting antenna and Root receiver antenna, The array has an element spacing of [missing information]. A uniform linear array, within the communication and sensing range of the base station. A far-field sensing target and With multiple single-antenna users existing simultaneously, the sensing and communication signals transmitted by the base station are within the same time-frequency resource block, and the channel state information is perfectly known; the base station downlink joint beamforming precoding matrix is... , Indicates the first Precoded vectors for each user , Indicates the first Beamforming vector of a sensing target , Indicates conjugate transpose; No. The signal received by each communication user is , , Indicates the sampling point. Indicates the maximum number of sampling points. For base station to user The channel vector, Indicates the first During the second sampling, the base station transmitted to the... Communication signals of individual users The Gaussian white noise has zero mean and a variance of . ,user The signal-to-noise ratio of the received signal is expressed as , If we take the absolute value, then the communication energy efficiency of the integrated sensing system is expressed as: , This indicates the total transmit power of the base station; The base station uses the echo signal to perform direction-of-arrival estimation on the sensed target, and the target response matrix. , ,in This represents the transpose of a vector. Indicates the first A far-field sensing target amplitude, Indicates the first The guidance vector of a perceived target. Indicates the first An angle of arrival, Indicates the spacing between array elements. The signal wavelength is used to determine the angle of the sensing target in the integrated sensing base station. The estimated Craméro boundary is composed of The calculation yielded, where Represents the target response matrix For vectors Taking the first derivative, the sensing energy efficiency of the synesthetic system is expressed as: ,in, Indicates matrix trace. This indicates that the matrix is ​​inverted. This indicates the maximum number of sampling points.

3. The method according to claim 2, characterized in that, Step two includes the following steps: user When using maximum ratio transmission and regular zero-forcing precoding, the normalized precoding vectors are as follows: , , Represents the magnitude of a vector. Represents the channel matrix. As a regularization factor, for An identity matrix of dimension; for the perceived target Considering the use of maximum ratio transmission precoding and null space zero-forcing precoding, the normalized precoding vectors are respectively , , for An identity matrix of dimension; the downlink joint beamforming precoding matrix of an integrated base station is represented by a normalized precoding vector as follows: ,in, and These represent the base station's communication with the user. and perceived target Allocated transmission power, and These represent the base station's communication with the user. and perceived target The normalized precoding vector, , , This indicates the base station power allocation method, for the total transmit power of the base station. ,in, Indicates the transmit power threshold. This indicates the maximum transmit power of the base station, and the total power of the transmitter is... ,at this time, Indicates energy conversion efficiency. This represents antenna power consumption; therefore, communication energy efficiency is expressed as... Perceived energy efficiency is expressed as ; Establish a multi-objective linear precoding optimization model for joint power allocation, with the optimization objective being: Constraint 1 is satisfied as follows: Constraint 2 is Constraint 3 is .

4. The method according to claim 3, characterized in that, Step three includes the following steps: First, set the number of particles to... The maximum number of iterations is The dimension of the particle search space is , ;No. Sub-particles The quantum position is , , ,particle No. The quantum position of the dimension is initialized as ,in, for Uniform random numbers between particles The Dimensional position is The mapping relationship is ,in, express Random numbers that follow a uniform distribution between , when At that time, particles The Wei represents communication users Precoding selection, , ;when At that time, particles The Dimension represents the perceived target Precoding selection, , ;when At that time, the first The first particle The dimension represents the downlink power allocation of the integrated base station, and the power allocation vector is... satisfy ,in ;when At that time, the first The first particle Wei represents the base station transmission threshold. , ; The first Sub-particles Substitute the position into the two objective functions , Calculate the corresponding objective function value; if for the th , particles, simultaneously satisfying: , If at least one strict inequality holds, then we say that the inequality is true in the th case. Generational particles Dominant Particle If initially all other particles cannot dominate the particle... So, the particles The location is placed in an external repository that stores non-dominated solutions, and the repository size is [size missing]. , Indicates the first repository size ; For objective function 1 Perform mesh generation, and set the mesh boundary spread factor to... The number of grids is For objective function 2 Similarly, the grid edges are divided to form a grid containing... A hypercube network of 1,000 grid cells is constructed. The specific location of each non-dominated solution within the established hypercube network is determined based on the objective function value, yielding the number of non-dominated solutions in each grid cell. Then, a roulette wheel selection process is used to select the non-dominated solution with the highest leadership density as the optimal particle. The leadership grid density is calculated using the following formula: , Indicates the first The leader grid density of the grid cell, Indicates the first The number of non-dominated solutions in a grid cell, of which , , The leader density factor is represented by the quantum position of the optimal particle. .

5. The method according to claim 4, characterized in that, Step four includes the following steps: calculate The second iteration Neutron enrichment levels of individual particles , No. The maximum enrichment energy level in the next iteration is denoted as The minimum value is denoted as , No. The nearest neighbor particle in the next iteration is represented as The central particle is Nearest neighbor particles are The average of the non-dominated solutions from the external repositories is obtained. Indicates rounding up. for Calculate the uniformly random number between the two values. The Euclidean distance of a particle from other particles The one with the smallest distance The average of the particles is obtained by taking the particle size. The central particle, for Random integers between [a certain range]; Calculate the first based on enrichment energy level The enrichment boundary of the next iteration , No. The second iteration The stability of a particle is expressed as ,in, for A uniform random number between, the first The iteration factor for the next iteration is Compare the neutron enrichment level with the boundary value; if Then the particle undergoes a decay process; if , indicating particle occurrence and decay, in which Represents particles stability, express A set of random numbers that follow a uniform distribution and are not repeated. middle as well as Each element forms a vector and ,in At this time, the particles occur Two particles are generated during decay: and ,in, , , and Representing the nearest neighbor particles and the central particle respectively Wei; if , indicating that it has occurred decay, particle release The ray forms two particles, and after decomposition, the first particle... The quantum rotation angles are updated to: , ,in, and They are respectively The particles are updated using a simulated quantum rotation gate, where random numbers following a uniform distribution are used. occur The quantum positions at the time of decay are respectively , ; if This indicates that the particle When the enrichment level is less than the boundary value, the optimal particle either receives or emits electrons. The quantum rotation angle is updated to Update quantum position using simulated quantum rotation gate .

6. The method according to claim 5, characterized in that, Step five includes the following steps: For the g+1 generation particles ,if The quantum positions of the two new particles produced by the decay process of the particle will then be determined. and Mapped to position and Calculate the dominance relationship, if Dominate ,and Simultaneously dominates the particles before the iteration Then the (g+1)th generation particle The quantum position is ;like Dominate ,and Simultaneously dominates the particles before the iteration Then the (g+1)th generation particle The quantum position is ; Otherwise, do not update, that is ; if Then compare the new particle with the g-th generation particle. The dominance relationship will determine the quantum position of the new particle. Mapped to position And calculate all objective function values, if Dominate Then the (g+1)th generation particle quantum position ; otherwise, ; Calculate the non-dominated solutions of the (g+1)th generation population and transfer them to an external repository; if the number of non-dominated solutions in the repository exceeds its storage capacity after updating, then... Then calculate the deleted grid density corresponding to all non-dominated solutions in the external repository: , Indicates the first Grid density for cell deletion, Indicates the first The number of non-dominated solutions in a grid cell This involves deleting the density factor, then using a roulette wheel to select the grid with the highest density, and randomly deleting non-dominated solutions from the grid with the highest density until the storage limit is reached, thus becoming the [number]th [cell]. The repository is used to calculate the leader grid density and obtain the updated optimal particle. The quantum position of the optimal particle is .

7. A Massive MIMO integrated multi-objective green linear precoding system, characterized in that, The system has a program module corresponding to the steps of the method described in any one of claims 1 to 6, and executes the steps in the Massive MIMO integrated sensing multi-objective green linear precoding method described above when running.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program configured to, when invoked by a processor, implement the steps of the Massive MIMO synesthetic multi-objective green linear precoding method according to any one of claims 1 to 6.

Images