A method and device for optimizing the number of SIM layers in a MIMO system in a mine environment
By introducing multi-layer SIMs into the mine MIMO system and optimizing the number of SIM layers and parameters, the problems of signal attenuation and high power consumption in the mine communication environment were solved, and efficient and secure mine communication was achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- XIAN UNIV OF SCI & TECH
- Filing Date
- 2026-06-18
- Publication Date
- 2026-07-24
AI Technical Summary
The communication environment in mines is complex. Existing MIMO systems suffer from severe signal attenuation in mines, which cannot meet the concurrent communication needs of multiple devices and users. Furthermore, high-power radio frequency components pose safety hazards, and traditional optimization methods have failed to fully unleash the system's energy efficiency limits.
In a mine MIMO system, a multi-layer SIM is introduced. By optimizing the number of SIM layers, phase shift matrix, and power allocation matrix, a channel model adapted to the three-dimensional space of the mine is constructed. The optimal number of SIM layers and parameters are solved by a hierarchical decoupling and alternating iterative algorithm to meet the explosion-proof power consumption constraints.
It significantly improves channel modeling accuracy and system energy efficiency, achieving highly reliable and low-power communication in mining environments, while synergistically enhancing communication performance and security indicators.
Smart Images
Figure CN122458057A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of wireless communication technology in mines, and particularly relates to a method and apparatus for optimizing the number of SIM layers in a MIMO system in a mine environment. Background Technology
[0002] In the current context of global energy structure transformation and increasingly prominent demands for industrial safety, mines, as the core carriers of energy supply, directly impact national energy security and industrial economic stability through the safety and efficiency of their underground operations. Mine communication systems, serving as the "lifeline" connecting underground work sites and surface command centers, not only undertake basic tasks such as daily production scheduling and equipment status monitoring, but also act as the core of emergency communication in response to sudden accidents such as gas exceeding limits and roof collapses. Their reliability, low power consumption, and explosion-proof compatibility have become key bottlenecks restricting the intelligent upgrading of mines.
[0003] The underground communication environment in mines is highly unique and complex: on the one hand, tunnels have a typical "narrow, deep, and curved" spatial structure, with a width of only 2-3 meters and branch roadways with an angle between 30° and 90°. The narrow and confined space severely restricts the propagation path of electromagnetic waves; on the other hand, the densely distributed metal supports and transportation equipment in the roadways form strong electromagnetic shielding, and the flying dust further aggravates signal attenuation, causing the non-line-of-sight (NLOS) propagation attenuation of the mainstream 5GHz communication frequency band to exceed 30dB, which seriously damages signal integrity.
[0004] Traditional Multiple-Input Multiple-Output (MIMO) communication technology requires dense deployment of radio frequency (RF) front-end equipment to compensate for signal attenuation and achieve full tunnel coverage. This not only leads to an exponential increase in hardware purchase, installation, and maintenance costs, but also the large amount of heat generated by the high-power RF components during operation can easily cause safety hazards such as underground gas explosions, fundamentally conflicting with the inherently safe explosion-proof standards for mines. Meanwhile, existing single-layer reconfigurable smart metasurface (RIS) technology can only achieve phase modulation in a two-dimensional plane, making it difficult to adapt to the multipath propagation characteristics of the three-dimensional space of mines. This results in blind spots in beam coverage, low signal energy utilization efficiency, and an inability to meet the concurrent communication needs of multiple devices and users underground.
[0005] Driven by these dual technological demands, stacked smart metasurface (SIM) technology, with its unique structural design and performance advantages, provides a revolutionary solution to the challenges of mine communication.
[0006] Currently, RIS / SIM-enabled MIMO communication systems have well-established technical support. Conventional methods used in the industry to improve system energy efficiency only involve joint optimization of the metasurface phase shift matrix and transmit power allocation. This optimization method is limited to the control of only two types of parameters and does not explore the optimization potential of other adjustable parameters. The obtained optimal energy efficiency value is only a local optimum under bivariate constraints, and the upper limit of system energy efficiency is not fully released. Summary of the Invention
[0007] The purpose of this invention is to provide a method and apparatus for optimizing the number of SIM layers in a MIMO system in a mining environment, which increases the number of optimization parameters based on existing optimization parameters, thereby fully releasing the upper limit of system energy efficiency.
[0008] This invention adopts the following technical solution: a method for optimizing the number of SIM layers in a MIMO system under mining conditions, applied to a SIM-assisted MIMO communication system under mining conditions, comprising the following steps: A SINR calculation model for user k is generated based on the transmission matrix of the SIM and the number of ultra-intelligent surface layers in the SIM. ;in, Indicates the SINR of user k. L represents the penetration loss factor of the ultra-smart surface layer in the SIM, and L represents the number of ultra-smart surface layers in the SIM. express Lth power, This represents the transmission power of the base station sending signals to user k. To indicate The List, This represents the channel state matrix between the base station and user k. express The List, This represents the SIM's transmission matrix when the base station sends a signal to user k, where K represents the number of users. This represents a user index that is different from user k. Indicates that the base station sends a message to the user SIM transmission matrix when transmitting signals The List, This represents the user's additive white Gaussian noise power; Calculate the system energy efficiency based on the SINR calculation model; Taking the maximization of system energy efficiency as the optimization problem and the communication conditions of the MIMO communication system in the mining environment as constraints, the number of super-intelligent surface layers, phase shift matrix and power allocation matrix of the base station in the SIM are obtained by solving the problem.
[0009] Another technical solution of the present invention: a SIM layer optimization device in a MIMO system in a mining environment, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the above-described method.
[0010] The beneficial effects of this invention are as follows: This invention incorporates the number of SIM layers into the SINR calculation formula of the mine MIMO system, which can accurately adapt to the narrow and enclosed transmission environment of the mine, which is characterized by multi-metal scattering and limited spatial dimensions. After introducing the layer factor reconstruction formula, the coupling relationship of useful signal power, inter-layer interference and multipath interference under different SIM layers can be quantified, which greatly improves the accuracy of channel modeling and SINR measurement, thereby fully releasing the upper limit of system energy efficiency. Attached Figure Description
[0011] Figure 1 This is a schematic diagram of the MIMO communication system framework in a mining environment according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating the energy efficiency optimization process of a mine SIM-MIMO system in an embodiment of the present invention. Figure 3 These are energy efficiency comparison curves at different SIM atomic scales in the embodiments of the present invention; Figure 4 This is a comparison curve of energy efficiency under different SIM layer configurations in this embodiment of the invention; Figure 5 The curves showing the energy efficiency as a function of the number of users under different algorithms in this embodiment of the invention are shown. Figure 6 This is a curve showing the relationship between the number of SIM layers and energy efficiency under different numbers of users in an embodiment of the present invention. Detailed Implementation
[0012] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.
[0013] Addressing the shortcomings of communication in mine environments, such as insufficient channel adaptability, existing research is largely based on ideal environments or ground-level modeling, failing to fully consider the "narrow, deep, and curved" three-dimensional spatial structure of mine tunnels and the strong attenuation characteristics of 5GHz signals exceeding 30dB. It also lacks a precise adaptation model between SIM and the complex channels in mines. Furthermore, the optimization dimensions are singular; a multi-objective coupled optimization model for SIM layer count, phase shift control accuracy, power allocation, and energy efficiency has not been established, making it difficult to achieve coordinated optimization of performance, cost, and energy consumption.
[0014] Therefore, how to construct an energy efficiency optimization mechanism for a SIM-enabled MIMO system that adapts to the three-dimensional confined space of a mine and takes into account explosion-proof power consumption constraints has become a technical problem that urgently needs to be solved for low-power and high-reliability communication in mines.
[0015] Existing SIM-assisted communication systems typically only optimize traditional communication parameters such as transmit power, precoding matrix, and beam direction, without taking the number of SIM layers as a core optimization variable. They ignore the decisive impact of the number of layers on beam gain, penetration loss, and system power consumption. As a result, the optimized system is unable to achieve optimal energy efficiency under the constraints of strong attenuation and explosion protection in mines. Beam coverage and energy consumption control cannot be balanced, and communication performance and safety indicators cannot be improved in a coordinated manner.
[0016] In existing MIMO communication systems for mining environments, base stations are densely deployed with radio frequency (RF) equipment, leading to a significant increase in hardware costs and power consumption. Furthermore, underground equipment must meet explosion-proof standards, and the high-power RF components of traditional base stations may pose heat dissipation risks. Single-layer RIS (Radio Frequency Identification) systems can only achieve two-dimensional phase modulation, which is insufficient to handle the complex multipath effects of the three-dimensional space in mines. This invention addresses this by integrating multi-layer cascaded SIMs into the base station, constructing a wave domain intelligent base station to achieve a collaborative control mechanism for electromagnetic metasurface units. This mechanism enables subwavelength-precision phase modulation of incident electromagnetic waves in three-dimensional space, achieving dynamic beamforming of the electromagnetic wave front field.
[0017] Mine tunnels are confined three-dimensional spaces characterized by narrowness, depth, and curvature. Electromagnetic wave transmission suffers severe attenuation and multipath interference, necessitating multi-layer SIMs to enhance three-dimensional beam focusing and coverage. However, increasing the number of SIM layers leads to a cumulative effect of interlayer dielectric penetration loss, and beamforming gain diminishes marginally with increasing layer count. Furthermore, intrinsically safe explosion-proof standards for mines impose strict upper limits on total system power consumption. The number of layers directly determines the balance between beam gain, penetration loss, and total power consumption: insufficient layers cannot compensate for the strong attenuation in mines, resulting in inadequate communication coverage and speed; excessive layers lead to a sharp increase in penetration loss and power consumption, exceeding explosion-proof power consumption limits and reducing energy efficiency. Therefore, optimizing the number of SIM layers is crucial. In the solution process, a fixed phase shift matrix and power allocation matrix are used within a block coordinate descent framework, decoupling the original optimization problem into an integer programming subproblem with only the number of SIM layers as the variable.
[0018] Based on the total power consumption constraint of intrinsically safe explosion-proof systems in mines, the feasible region and maximum feasible number of SIM layers are calculated. With the phase shift matrix and power allocation matrix fixed, a nonlinear fractional energy efficiency objective function is constructed, with the number of SIM layers as the only variable, where the number of layers is a positive integer discrete variable. The Newton-Raphson iteration method is used to solve for the extreme points of the number of layers in the continuous domain, and the strict monotonicity of the objective function is utilized to achieve fast convergence. Within the feasible number of layers, the nearest integers to the optimal solution in the continuous domain are selected as candidate numbers of layers, and the system energy efficiency corresponding to each candidate integer is calculated. The integer number of layers that maximizes the energy efficiency is selected as the optimal number of SIM layers for the mine near-field MIMO system.
[0019] The energy efficiency optimization method for MIMO systems of the present invention includes the following steps: Step 1: Obtain the parameters of the mine communication scenario, SIM hardware parameters, MIMO system parameters, and intrinsically safe explosion-proof constraint parameters for the mine; Step 2: Construct a SIM-assisted MIMO system architecture and channel model adapted to the mining scenario, and establish an accurate channel model that includes inter-layer transmission loss, near-field propagation law and underground channel attenuation characteristics; Step 3: Establish a coupling model of SIM-assisted downlink transmission, rate and power consumption, and quantitatively analyze the coupling effects of SIM layer number, phase shift control accuracy, penetration loss factor and power allocation strategy on system energy efficiency; Step 4: With the goal of maximizing system energy efficiency, construct an optimization problem that includes user rate QoS constraints, total power consumption constraints for mine explosion protection, positive integer number of SIM layers, phase shift range constraints, and penetration loss factor constraints. Step 5: Solve the optimization problem using a hierarchical decoupling and alternating iterative algorithm to obtain the optimal number of SIM layers, the optimal phase shift matrix, and the optimal power allocation scheme, thus completing the system energy efficiency optimization; Specifically, when constructing a SIM-assisted MIMO system architecture and channel model adapted to the mining scenario, the mine roadway is modeled, with a length of 500m, a width of 2-3m, and a branch angle of 30°-90°; the transmitter is configured... root antenna and Layer-stacked smart metasurfaces, receiver configuration Each SIM layer contains M superatoms. Based on Rayleigh-Sommerfeld diffraction theory, a multilayer transmission coefficient matrix and near-field line-of-sight (LoS) channel model for the SIM are established; a single-layer penetration loss factor is introduced. , Layer cumulative loss is expressed as This characterizes the dielectric attenuation of electromagnetic waves as they pass through a multilayer SIM.
[0020] When establishing the SIM-assisted downlink transmission, rate, and power consumption coupling model, the formulas for user received signal, signal-to-interference-plus-noise ratio (SINR), and single-user achievable rate are derived; total power consumption includes transmit power, SIM modulation power consumption, RF chain power consumption, and baseband fixed power consumption; system energy efficiency is defined as the ratio of total users and rate to total power consumption.
[0021] Construct an energy efficiency maximization optimization problem under multiple constraints, with the objective function being the maximization of system energy efficiency and the optimization variable being the number of SIM layers. Phase shift matrix Power allocation matrix The constraints include: User rate QoS constraints: The rate for a single user is not lower than the rate threshold. Total power consumption constraints for mine explosion-proof systems: The total power consumption of the system shall not exceed the upper limit of the explosion-proof standard; SIM phase shift constraint: The phase shift value range is The modulus is 1; SIM layer number constraint: The number of layers is a positive integer not less than 1; Penetration loss factor constraint: .
[0022] The optimization problem is solved using a hierarchical decoupling and alternating iterative algorithm, including: Step 1: Solving for the optimal number of SIM layers: Given the initial phase shift With power Solving for the extreme points in the continuous domain using Newton's iterative method The optimal integer layer number is obtained by filtering neighboring integers. ; Step 2: Power Allocation Optimization: Fix the Optimal Number of Layers With phase shift The non-convex problem is transformed into a convex problem using the WMMSE method, and then solved using the interior point method. Step 3: Phase Shift Matrix Optimization: Fix the Optimal Number of Layers With power The problem is transformed into maximizing beamforming gain, and the gradient descent method is used for iterative solution. Step 4: Alternately execute steps one through three until the energy efficiency converges and output the optimal parameters.
[0023] More specifically, this invention discloses a method for optimizing the number of SIM layers in a MIMO system under mining conditions, applicable to, for example... Figure 1 The SIM-assisted MIMO communication system in a mining environment, as shown, includes the following steps: Will This represents the number of ultra-intelligent surface layers at the base station. This represents its set. Assume that each super-intelligent surface layer at the transmitter consists of the same number of units. Specifically, using... This represents the number of units on each super-intelligent surface layer. The corresponding set is represented as... Based on this, it means It is the phase shift caused by the m-th unit on the transmission-mode intelligent surface layer, where It is the transmission coefficient of the m-th unit on the super-intelligent surface layer.
[0024] With the The transmission coefficient matrix associated with each super-intelligent surface layer is represented as follows: The transmission coefficient matrix of the first layer of the super-intelligent surface layer is: In the case where all superintelligent surface layers have an isomorphic lattice arrangement, each superintelligent surface layer is modeled as a uniform planar array, where two distinct units exist on the same emitting superintelligent surface layer. and The spacing between them is: (1) in, This indicates the spacing between adjacent units on the same super-intelligent surface layer. Furthermore, and Corresponding to along shaft and The first axis Each unit (the coordinate system is established as a whole based on SIM) and Corresponding to along shaft and The first axis Units.
[0025] (2) in, This is the maximum number of cells included in each row of the super-intelligent surface layer, assuming the super-intelligent surface layer is a square surface, then the number of cells it contains. .
[0026] By assuming that the spacing between all surfaces is uniform and that all surfaces are parallel, the super-intelligent surface layer will be... The unit above To the Units on the layer The transmission distance is defined as: (3) in, It is the spacing between any two adjacent super-smart surface layers in the SIM, where This indicates the thickness of the SIM.
[0027] Furthermore, by assuming that the center of the transmitting antenna array is aligned with the center of all the super-intelligent surface layers, and that the antenna arrays are arranged with element spacing... In a uniform linear array, starting from the base station's... The transmitting antenna is connected to the first layer of the super-intelligent surface layer in the SIM. The distance between units is provided by the following formula: (4) This indicates the total number of transmitting antennas at the base station. This indicates the wavelength of the emitted wave.
[0028] Regarding the information provided by Rayleigh-Sommerfeld diffraction theory from superintelligent surface layers The unit above To the super intelligent surface layer The unit above The transmission coefficient is given by the following formula: (5) in, This indicates the cell region at the transmitter SIM. Is the direction of propagation and the first The angle between the normal directions of the super-intelligent surface layer and the This represents the corresponding transmission distance. Therefore, the transmission matrix of the SIM can be written as: (6) in, , It is an ultra-intelligent surface layer and super intelligent surface layer The transmission coefficient matrix between them .
[0029] The description of the near-field LoS channel, without loss of generality, assumes that the SIM is implemented in the YZ plane, while all users are located in the XY plane. In the case of the user's ULA, assuming they are parallel to the y-axis, the signal coordinates of the SIM's outer surface are given by the following equation: (7) in, and , It refers to the SIM cell spacing.
[0030] Therefore, users The coordinates of antenna n are given by the following formula: (8) in, This refers to the user antenna spacing.
[0031] SIM Each unit (referring to any unit in the SIM) and the user The The near-field LoS channel between the antennas is The EM signal experiences free-space path loss when propagating in the spherical wavefront.
[0032] (9) in, and Representing users respectively The The antenna and the first antenna in the SIM The distance between units and the free space path loss coefficient. It is worth noting that... It can be written as: (10) in, and They are users The distance and angle relative to the base station.
[0033] This invention proposes a downlink transmission model that leads to corresponding achievable rates for SIM-assisted near-field MIMO systems. During downlink transmission, wave-based beamforming occurs due to SIM, which contrasts with conventional digital precoding, where each symbol is assigned a separate beamforming vector. In this setup, the BS selects a set of K appropriate antennas from all MBS antennas, since each data stream must be transmitted directly from its corresponding antenna at the BS.
[0034] user The received signal is given by the following formula: (11) in, It has a distribution AWGN, It is a symbol vector, where, for , and .also, It is a diagonal matrix, where its th... The diagonal term represents the allocation to the first... The first user's The square root of the power of each data stream. The total transmit power constraint at BS is given by the following formula: (12) in, This represents the transmit power budget at BS.
[0035] A near-field MIMO communication system model for mines is constructed. A multi-layer SIM transmission model is established based on Rayleigh-Sommerfeld diffraction theory. Multi-layer dielectric penetration loss is introduced to establish a channel model and energy efficiency calculation model suitable for explosion-proof and low-power scenarios in mines. The penetration loss directly affects the total number of users and the rate of the energy efficiency numerator by changing the channel transmission gain.
[0036] In non-ideal conditions, when electromagnetic waves pass through a medium, a small portion of the medium molecules absorb the electromagnetic wave energy and convert it into heat energy. There is also a small probability of reflection at the medium interface, causing some electromagnetic waves to be bounced back or not penetrated. Additionally, scattering within the medium can disrupt the propagation direction of the electromagnetic waves, preventing effective penetration. Therefore, considering the loss characteristics of the SIM (Single Inductor Module), a single-layer penetration loss factor ξ∈(0,1] is assigned to each layer, and it is assumed that the loss of each layer is equal. Thus, the cumulative loss across multiple layers can be expressed by an exponential function. To express.
[0037] Therefore, the present invention provides a method for optimizing the number of SIM layers in a MIMO system under mining conditions, applicable to a SIM-assisted MIMO communication system under mining conditions, comprising the following steps: A SINR calculation model for user k is generated based on the transmission matrix of the SIM and the number of ultra-intelligent surface layers in the SIM. ;in, Indicates the SINR of user k. L represents the penetration loss factor of the ultra-smart surface layer in the SIM, and L represents the number of ultra-smart surface layers in the SIM. express Lth power, This represents the transmission power of the base station sending signals to user k. To indicate The List, This represents the channel state matrix between the base station and user k. express The List, This represents the SIM's transmission matrix when the base station sends a signal to user k, where K represents the number of users. This represents a user index that is different from user k. Indicates that the base station sends a message to the user SIM transmission matrix when transmitting signals The List, This represents the user's additive white Gaussian noise power; Calculate the system energy efficiency based on the SINR calculation model; Taking the maximization of system energy efficiency as the optimization problem and the communication conditions of the MIMO communication system in the mining environment as constraints, the number of super-intelligent surface layers, phase shift matrix and power allocation matrix of the base station in the SIM are obtained by solving the problem.
[0038] No. The achievable rate for each user is .
[0039] The total power consumption of the base station is The transmission power is SIM control power consumption is ( (Power consumption of a single unit), RF chain power consumption is ( (Number of BS antennas) This is the power consumption of a single transmitter chain; the fixed power consumption of baseband processing is... .
[0040] Then the system energy efficiency is: (13) Where R represents user and rate.
[0041] The optimization objective is to maximize system energy efficiency, using an integer number of SIM layers. Phase shift matrix Power allocation matrix To optimize the variables, an optimization problem is established with user QoS, mine explosion-proof power consumption, and hardware characteristics as constraints. The optimization problem can be expressed as: (14) in, Indicates system energy efficiency. This represents the transmission power of the base station to the d-th antenna of user k. , Indicates the first in SIM The phase of the m-th unit in a super-intelligent surface layer. This represents the base station's transmit power, and M represents the number of cells in the super-intelligent surface layer. Indicates the controlled power consumption of the unit. This indicates the RF chain power consumption of the base station. This indicates the fixed power consumption of the baseband processing at the base station; This represents the rate of user k. The minimum achievable rate threshold for a single user (ensuring real-time data transmission for intelligent operations in the mine). This indicates the total power of the base station. This indicates the maximum total power consumption limit set by the explosion-proof standard. This represents the transmission matrix of the SIM. This represents the phase shift matrix of the Lth super-intelligent surface layer. This represents the transmission coefficient matrix between the (L-1)th and Lth superintelligent surface layers. , Indicates the first The transmittance coefficient of the Mth unit in the super-intelligent surface layer. .
[0042] in addition, It is a user rate QoS constraint; This indicates the total power consumption constraint for intrinsically safe explosion-proof mines. This refers to the base station's transmit power. It is a SIM transmission matrix. It is a SIM phase shift discrete constraint; It is a positive integer layer number constraint for SIM, which conforms to the hardware physical structure; It is a constraint on the phase shift range of each cell in the SIM layer, and the phase range. This ensures the continuity of phase modulation; This indicates a single-layer penetration loss factor constraint.
[0043] Due to constraints c1-c7, the number of SIM layers Phase shift matrix With power distribution The problems are mutually coupled and involve integer constraints, constant modulus non-convex constraints, and fractional objective functions. Problem P1 is a non-convex mixed integer fractional programming problem, which is difficult to solve directly by mixing the two to obtain the optimal solution.
[0044] Due to the difficulty in directly solving the problems caused by strong variable coupling, mixed integers, fractional non-convexity, and constant modulus constraints, a block coordinate descent BCD framework is used. By fixing some variables and solving single variable blocks sequentially, the original problem is decoupled into three sub-problems. Each sub-problem is solved sequentially and iteratively updated to continuously improve energy efficiency while meeting the low-power explosion-proof constraints of the mine. When the energy efficiency converges, the optimal parameters and the maximum energy efficiency under the green explosion-proof scenario of the mine are output.
[0045] In other words, the optimization problem is divided into three sub-problems: solving for the number of SIM layers, power allocation optimization, and phase shift matrix optimization. These three sub-problems are then solved iteratively in turn. In each round of iteration, the number of SIM layers is solved first.
[0046] More specifically, such as Figure 2 As shown, to solve this optimization problem, the feasible range of the number of SIM layers is first determined by power consumption constraints. Then, the given problem is divided into three sub-problems to be solved step by step: the optimal number of SIM layers, power allocation optimization, and phase shift matrix optimization. Finally, the sub-problems are solved by alternating iterations to form an iterative algorithm based on hierarchical decoupling and alternating optimization.
[0047] Subproblem 1: Solving for the optimal number of SIM layers.
[0048] Based on the total power consumption constraint of intrinsically safe explosion-proof systems in mines, the feasible region and maximum feasible number of SIM layers are calculated. With the phase shift matrix and power allocation matrix fixed, a nonlinear fractional energy efficiency objective function is constructed, with the number of SIM layers as the only variable, where the number of layers is a positive integer discrete variable. The Newton-Raphson iteration method is used to solve for the extreme points of the number of layers in the continuous domain, and fast convergence is achieved by utilizing the strict monotonicity of the objective function. Within the feasible number of layers, the nearest integers to the optimal solution in the continuous domain are selected as candidate numbers of layers, and the system energy efficiency corresponding to each candidate integer is calculated. The integer number of layers that maximizes the energy efficiency is selected as the optimal number of SIM layers for the mine near-field MIMO system.
[0049] Finding the optimal number of SIM layers is a high-dimensional, strongly coupled, mixed-integer, non-convex optimization problem, making it impossible to optimize three variable blocks simultaneously. To decouple variables, reduce complexity, and make the problem solvable, this invention employs a Block Coordinate Descent (BCD) framework. Multiple optimization variables are divided into several variable blocks, and only one variable block is optimized at a time, while all other variable blocks are fixed. Through iterative updates, the optimization gradually converges to a stationary point.
[0050] Therefore, in optimizing the number of layers At that time, and Set as a fixed value (i.e., a fixed continuous variable) and This decouples the mixed integer problem into a single-variable integer programming problem, significantly reducing the difficulty of solving it. This is achieved by assuming a phase shift matrix. and power distribution Using the initial values (uniform power distribution + random phase shift), we study the subproblem of layer optimization for problem P1.
[0051] Specifically, the total speed and total power consumption are respectively and To decouple mixed integer variables, a fixed phase shift matrix is used. With power allocation matrix Only for the number of floors Optimization is then performed. At this point, a fixed power consumption independent of the number of layers is defined: The total power consumption degrades to only that of... Related: (15) Due to the near-field LosS spherical wave channel characteristics in mines, the beamforming gain brought by multi-layer SIM increases logarithmically with the number of modulation layers: (16) In the formula, , , and All of these are constants related to the number of units, the number of users, and the channel.
[0052] Due to the penetration loss caused by cascading multiple dielectric layers, electromagnetic waves experience a loss factor as they pass through each layer. , The total layer loss decreases exponentially: The number of users, noise power, transmit power, mine NLOS attenuation, etc. are compared with... Irrelevant terms are merged into the system comprehensive coefficient. Then SINR can be expressed as (17) In the high SINR operating range that meets QoS in mine communication, a standard approximation is adopted. The total rate is linearly proportional to SINR, therefore (18) The total power consumption of intrinsically safe explosion-proof mines must meet the requirements. ,make Substitution The maximum feasible number of layers is obtained by solving the problem. number of floors It is a positive integer, that is .make and Generate a subproblem for solving the number of SIM layers with L as the only variable: (19) in, This represents the system energy efficiency with L as the variable. Indicates and The overall system coefficient obtained by combining irrelevant terms; This indicates that electromagnetic waves generate a loss factor with each layer of the ultra-intelligent surface layer they pass through. This represents the maximum feasible number of layers obtained by solving based on power constraints. This represents a fixed power consumption independent of the number of SIM layers.
[0053] Since the objective function is a nonlinear fractional function is a positive integer, Since it is a discrete function, it cannot be solved by differentiation using continuous optimization methods; Newton's iteration method utilizes... The strict monotonicity enables fast convergence, and the discrete screening satisfies the hardware integer constraint.
[0054] Therefore, solving the subproblem of determining the number of SIM layers includes: For optimization problems involving nonlinear, differentiable fractional functions, Newton's iteration method is used to find the extreme points in the continuous domain. ; Filter neighboring integers within the feasible range , and As a candidate layer number; Calculate the system energy efficiency for each candidate layer number Choose the system with the highest energy efficiency. Corresponding candidate layer number The optimal layer number is shown in Table 1 below, which illustrates the specific calculation method for this subproblem.
[0055] Table 1
[0056] Sub-problem 2: Power allocation optimization.
[0057] Within the block coordinate descent framework, the optimal number of SIM layers has been obtained by solving subproblem P1.1. Set the number of SIM layers to the optimal number of layers. ,Will Set to fixed values. In this case, the channel matrix, beamforming gain, and penetration loss are all constants, and the system energy efficiency maximization problem degenerates into a problem concerning only the power allocation matrix. The optimization problem.
[0058] Since the energy efficiency target is still a fractional non-convex structure (rate is a concave function, power consumption is a linear function; concave / linear is a typical non-convex fraction), it cannot be solved directly using convex optimization methods. Therefore, this invention introduces a weighted minimum mean square error (WMMSE) equivalent transformation, which strictly transforms the non-convex sum-rate maximization problem into a convex optimization problem, achieving a globally optimal solution.
[0059] fixed , After that, the total rate Power distribution only Determine the total power consumption. China's transmission power and The relevant terms are all constants, and the remaining terms are fixed. Therefore, the energy efficiency maximization problem degenerates into: (20) in The achievable speed for the user. For the total reachable rate of the system, Indicates user No. The signal-to-interference-plus-noise ratio of each data stream; The total power consumption of the system is Fixed power consumption (including radio frequency power consumption, base station basic power consumption, etc.) that is independent of power allocation. The maximum allowable transmit power of the system. For users No. The transmit power of each antenna; Set minimum QoS rate constraints for users to ensure that the communication quality of each user meets the preset requirements.
[0060] Since the objective function is a fraction of total speed and total power consumption, and the total speed about The function is concave, and the total power consumption is... about Since the objective function is a linear function, the ratio of the concave function to the linear function constitutes a typical non-convex fractional objective, which cannot be solved directly using convex optimization theory. Therefore, we first transform the fractional optimization problem into a non-convex problem through an equivalent transformation, and then introduce the WMMSE transformation to achieve convexity.
[0061] For the fractional objective of maximizing energy efficiency, the Dinkelbach algorithm is used to transform it into a parameterized non-fractional optimization problem. The core idea of this method is to iteratively update the energy efficiency parameters. This transforms the fractional programming problem into a series of unconstrained linear objective optimization problems, that is, it transforms the optimization problem into a parameterized non-fractional optimization problem: (twenty one) in, Represents the total reachable rate of the system. Indicates by The resulting power allocation matrix requires optimization. This represents the energy efficiency parameters that are updated iteratively.
[0062] The initial value can be set to Each iteration updates the solution based on the optimal solution from the previous round. until the convergence condition is met. ( (This is for the preset convergence precision). When the method converges, at this point... This represents the optimal energy efficiency of the system, and the corresponding power allocation. This is the optimal power allocation corresponding to maximizing energy efficiency.
[0063] During the iteration process, for a fixed , Since the term is a constant, the above optimization problem can be further simplified to a sum-rate maximization problem: (twenty two) At this point, the problem is transformed into maximizing the sum rate under power and QoS constraints. This problem is still a non-convex optimization problem and needs to be further convexized through WMMSE equivalent transformation.
[0064] For users No. There are one antenna, and the transmitted symbol is... (satisfy The receiver uses a linear equalizer. Equalize the received signal to obtain the symbol estimate. ,in To receive the signal, the expression is: (twenty three) in For users No. The receive channel vector of each data stream, This is the transmit channel vector for the data stream from the base station to the SIM. The first term represents Gaussian white noise, and the second term represents interference between multiple users.
[0065] Define user No. The symbol estimation error for each data stream is The mean square error is defined as follows: (twenty four) Will Substituting into the above equation and expanding, we get: (25) in For users No. The total interference power experienced by each data stream is a constant (because...). and fixed).
[0066] To minimize mean square error Regarding its equilibrium coefficient Taking the derivative and setting it to zero, we can obtain the optimal MMSE equilibrium coefficient: (26) The optimal equilibrium coefficient Substituting back into the MSE expression, we obtain the minimum mean squared error (MMSE): (27) This formula establishes a strict correspondence between minimum mean square error and signal-to-interference-plus-noise ratio (SINR), laying the foundation for the subsequent equivalent conversion between rate and MSE.
[0067] Substituting the relationship between minimum mean square error and signal-to-interference-plus-noise ratio into the rate formula, we get: (28) Summing over all users and data streams, the total system rate can be expressed as: (29) This equation shows that maximizing the total rate of the system is equivalent to maximizing the negative logarithm of the minimum mean square error.
[0068] Introducing weight variables (Based on the weights of each antenna), construct the weighted mean square error (WMMSE) objective function: (30) The first term is the weighted mean square error sum, and the second term is the logarithmic penalty term for the weight variables, used to ensure equivalence.
[0069] According to optimization theory, for any weight variable objective function about The maximum value can be obtained by adjusting the value of the maximum value. By taking the derivative and setting it to zero, we obtain the result. For about Differentiate: (31) Setting the derivative to 0 yields the optimal weights. Substituting the optimal weights into the objective function, we get: (32) when That is, when the equalizer reaches its optimal state, the first term is a constant. At this point, the maximum value of the objective function is: (33) Therefore, it can be proven that maximizing the sum rate is strictly equivalent to minimizing the WMMSE objective function, i.e. .
[0070] Substituting the definition of mean squared error into the WMMSE objective function while retaining all constraints of the original problem, we obtain the final form of the second power allocation subproblem. In other words, the power allocation subproblem is obtained by convexifying the non-fractional optimization problem through the WMMSE equivalent transformation: (34) in, This represents the estimated value of the received signal. This represents the weight matrix introduced in relation to the data flow. This indicates the number of antennas for user k. This represents the weighting factor for the d-th antenna of user k. This represents the transmitted signal of the d-th antenna of user k. To represent the estimated received signal value of user k's d-th antenna, For the transmit power budget, Indicates user No. The signal-to-interference-plus-noise ratio of each antenna; The transformation is based on the following: the minimum value of the weighted mean square error function is equivalent to the maximum value of the original summation rate (satisfying the KKT conditions), and the transformed objective function... Regarding the convex function. It is about The concave function, but yes The nonlinear function leads to The maximization problem is non-convex and cannot be solved directly using convex optimization methods. For convex optimization problems with linear inequality constraints, the interior-point method is used to solve this power allocation subproblem, iteratively updating... , and until the power change meets the convergence accuracy. The optimal power allocation matrix is obtained. The specific solution method is shown in Table 2.
[0071] Table 2 Subproblem 3: Phase shift matrix optimization.
[0072] The optimal number of SIM layers has been obtained by solving subproblem 1. Subproblem 2: Finding the optimal power allocation With the optimal beamforming vector Under the premise of setting the number of SIM layers as the optimal number of layers. Set the power allocation matrix as the optimal power allocation matrix. Only the phase shift matrix Φ is optimized to maximize the system energy efficiency. The overall system energy efficiency is defined as: (35) Among them, total power consumption .because , It has been fixed. For phase shift matrix Irrelevant constants; meanwhile, users No. The achievable rate of a data stream Its SINR is proportional to the received signal power: (36) Therefore, maximizing system energy efficiency is equivalent to maximizing the sum of the received signal power of all users and all data streams.
[0073] To address the cascaded transmission characteristics of multi-layer SIMs in mines, an inter-layer transmission gain correction coefficient is introduced. (Based on the optimal number of layers) (The decision is made to keep constants), and the final objective function is: (37) Based on the hardware physics properties of the superatoms in intelligent metasurfaces (SIM), the phase shift must satisfy two types of core constraints: 1. Unity modulus constraint: Superatoms only modulate the phase, not the amplitude; therefore, the modulus of a phase-shifting element is always 1. , of which is the first l The phase shift value of the m-th cell in the layer.
[0074] 2. Phase range constraint: The phase value is between 0 and... between: ,in For the first Layer The phase angle of each unit satisfies .
[0075] Integrating the objective function and constraints, we obtain the phase shift matrix optimization subproblem: (38) in, Indicates by The phase shift matrix formed, Represents the optimal power allocation matrix The transmit power of the base station to the d-th antenna of user k. This represents the channel state between the base station and the d-th antenna of user k. This represents the transmission coefficient matrix between the d-th antenna of user k. Indicates the first in SIM The phase shift matrix of the layer. It has a distribution AWGN, It is a symbol vector, where, for , and .also, It is a diagonal matrix, where its th... The diagonal term represents the allocation to the first... The first user's The square root of the power of each data stream. The total transmit power constraint at BS is given by the following formula: Solve the phase shift matrix optimization subproblem to obtain the optimal phase shift matrix. For problem P1.3, constant modulus constraint... (Non-convex constraints) result in the feasible region being a non-convex set; at the same time yes Nonlinear functions (channel matrix) yes The objective function is a composite function, and the objective function is non-convex, with no closed-form optimal solution. For high-dimensional variable optimization problems, an iterative optimization algorithm based on gradient descent is used to solve them. The objective function is derived to address phase shifts. The gradient expression is simplified by combining near-field channel range / angle information, the phase shift is updated along the gradient ascent direction, and then normalized. To the desired range; simultaneously, increase the gradient weight for directions with larger channel singularities to enhance near-field beamforming gain, until the beamforming gain change meets the convergence accuracy. The specific solution process is shown in Table 3 below.
[0076] Table 3 In summary, the optimization problem is solved by decoupling integer constraints in a hierarchical manner and alternately optimizing continuous variables until convergence. Furthermore, the overall method of this invention is summarized in Table 4 below.
[0077] Table 4 To verify the effectiveness of the method of the present invention, a comparative experiment was conducted with the existing mainstream methods. The channel parameters, mine environment parameters, hardware constraints, power consumption limits and QoS requirements of all experiments were kept consistent, with only the optimization algorithm and configuration scheme being different.
[0078] like Figure 3 As shown, in a mine roadway with a total length of 500m and a width of 2.5m, 5GHz near-field spherical wave communication, a far-field boundary distance of 11.06m, an intrinsically safe explosion-proof total power consumption limit of 5W, a user count of K=16, a minimum single-user rate QoS constraint of 2Mbps, and a single-layer penetration loss factor... With a SIM array diameter of 0.92 and a SIM array aperture of 0.576m (i.e., the width and height of the SIM), the system energy efficiency comparison curves under different SIM cell sizes are shown.
[0079] As shown in the figure, with the increase of the number of single-layer units in the SIM, the system's adjustable degrees of freedom are improved, the beamforming gain is significantly enhanced, and the energy efficiency shows a clear upward trend. However, when the unit size is too large, the interlayer coupling loss and control power consumption increase slightly, and the energy efficiency growth rate gradually slows down. When the number of SIM layers L=4, it indicates that there is an optimal unit size that balances gain and loss, verifying the adaptability of the model established in this invention to the SIM hardware scale.
[0080] like Figure 4 As shown, under the conditions of a total mine roadway length of 500m, a width of 2.5m, 5GHz near-field spherical wave communication, a far-field boundary distance of 11.06m, an intrinsically safe explosion-proof total power consumption limit of 5W, a number of users K=16, an array aperture of 0.576m, a single-layer penetration loss factor ξ=0.92, a single-user minimum rate QoS constraint of 2Mbps, and 4-layer SIM, the curves of system energy efficiency as a function of iteration number under different unit number configurations are displayed to verify the convergence performance of the energy efficiency optimization method of this invention.
[0081] As shown in the figure, the system energy efficiency under different unit number configurations increases rapidly with the number of iterations, and basically converges to a stable value after three iterations. This indicates that the optimization method proposed in this invention has a fast convergence speed and stable solution results. At the same time, the more SIM units there are, the higher the steady-state energy efficiency after convergence. The configuration with M=36 (6×6) can achieve higher energy efficiency, which verifies that a larger scale of SIM units can provide stronger beam focusing gain and effectively improve the energy utilization efficiency of the system.
[0082] like Figure 5 As shown, in a mine tunnel with a total length of 500m and a width of 2.5m, 5GHz near-field spherical wave communication, a far-field boundary distance of 11.06m, an intrinsically safe explosion-proof total power consumption limit of 5W, a SIM single-layer atomic scale of 24×24, an array aperture of 0.576m, and a single-layer penetration loss factor... Under the conditions of a minimum single-user rate (QoS) constraint of 2 Mbps and the near-field optimal number of SIM layers, the performance curves of system energy efficiency as a function of the number of users under different optimization algorithms are presented. The figure compares four schemes: the hierarchical decoupling + alternating optimization joint algorithm of this invention, the block coordinate descent (BCD) algorithm, the maximum transmission ratio (MRT) algorithm, and the random phase shift method. Specifically, the joint algorithm of this invention performs three-dimensional joint optimization of the near-field SIM layer number, phase shift matrix, and power allocation; the BCD algorithm only optimizes the phase shift matrix and power allocation with a fixed number of layers; the MRT algorithm only optimizes the far-field equivalent beam direction with a fixed number of layers and phase shift; and the random phase shift method uses random phase shift, uniform power allocation, and a fixed number of layers.
[0083] The results show that, under the same number of users and mine constraints, the energy efficiency of the joint optimization algorithm proposed in this invention is always higher than that of the other three comparative methods. As the number of users increases and multi-user interference intensifies, the advantages of the algorithm in power allocation and phase shift collaborative optimization become more prominent, further verifying the superiority of the algorithm in multi-user concurrent scenarios in mines.
[0084] like Figure 6 As shown, in a mine tunnel with a total length of 500m and a width of 2.5m, 5GHz near-field spherical wave communication, a far-field boundary distance of 11.06m, an intrinsically safe explosion-proof total power consumption limit of 5W, a SIM single-layer atomic scale of 24×24, an array aperture of 0.576m, and a single-layer penetration loss factor... Under the conditions of a minimum single-user rate (QR) of 0.92 and a QoS constraint of 2 Mbps, the relationship between the number of SIM layers and energy efficiency is shown for different numbers of users. As can be seen from the figure, when the number of users is 10, 16, 20, and 32, the energy efficiency first increases and then decreases with the number of SIM layers. The more users there are, the higher the requirement for beam control accuracy, and the corresponding optimal number of SIM layers increases, causing the peak value of the curve to shift to the right. This result is completely consistent with the coupling laws of beam gain, inter-layer loss, and multi-user interference in near-field MIMO channels in mines, proving the correctness and applicability of the optimization model and iterative solution method of this invention.
[0085] The present invention also discloses a SIM layer optimization device in a MIMO system in a mining environment, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the above-described method.
[0086] Those skilled in the art will understand that the steps described in the embodiments disclosed in this invention can be implemented either by electronic hardware or by a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application scenario and design constraints of the technical solution. Those skilled in the art can implement the above functions in different ways to meet different application needs, and such implementations should not be considered as departing from the scope of protection of this invention.
[0087] In the embodiments provided by this invention, the described method may also be implemented in other ways. For example, the description of the above embodiments is only illustrative in nature. Multiple modules or components may be integrated or integrated into other systems, and certain features or related steps may be omitted.
[0088] In summary, the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail based on the foregoing embodiments, those skilled in the art can still improve the technical solutions described in the embodiments or make equivalent substitutions for some technical features. As long as these improvements or substitutions do not cause the corresponding technical solutions to depart from the core essence and protection scope of the technical solutions of the embodiments of the present invention, they should all be included within the protection scope of the present invention.
[0089] In the above embodiments, the descriptions of each embodiment have different focuses. If there are any parts in a certain embodiment that are not described in detail or recorded, you can refer to the relevant descriptions of other embodiments.
Claims
1. A method for optimizing the number of SIM layers in a MIMO system under mining conditions, characterized in that, A MIMO communication system for SIM-assisted mining environments includes the following steps: A SINR calculation model for user k is generated based on the transmission matrix of the SIM and the number of ultra-intelligent surface layers in the SIM. ;in, Indicates the SINR of user k. L represents the penetration loss factor of the ultra-smart surface layer in the SIM, and L represents the number of ultra-smart surface layers in the SIM. express Lth power, This represents the transmission power of the base station sending signals to user k. To indicate The List, This represents the channel state matrix between the base station and user k. express The List, This represents the SIM's transmission matrix when the base station sends a signal to user k, where K represents the number of users. This represents a user index that is different from user k. Indicates that the base station sends a message to the user SIM transmission matrix when transmitting signals The List, This represents the user's additive white Gaussian noise power; Calculate the system energy efficiency based on the SINR calculation model. By taking maximizing system energy efficiency as the optimization problem and the communication conditions of the MIMO communication system in a mining environment as constraints, the number of super-intelligent surface layers, phase shift matrix, and power allocation matrix of the base station in the SIM are obtained.
2. The method for optimizing the number of SIM layers in a MIMO system under mining conditions as described in claim 1, characterized in that, The optimization problem and constraints are as follows: ， in, Indicates system energy efficiency. This represents the transmission power of the base station to the d-th antenna of user k. , Indicates the first in SIM The phase of the m-th unit in a super-intelligent surface layer. This represents the base station's transmit power, and M represents the number of cells in the super-intelligent surface layer. Indicates the controlled power consumption of the unit. This indicates the RF chain power consumption of the base station. This indicates the fixed power consumption of the baseband processing at the base station; This represents the rate of user k. The minimum achievable rate threshold for a single user. This indicates the total power of the base station. This indicates the maximum total power consumption limit set by the explosion-proof standard. This represents the transmission matrix of the SIM. Indicates the first Phase shift matrix of the super-intelligent surface layer This represents the transmission coefficient matrix between the (L-1)th and Lth superintelligent surface layers. , Indicates the first The transmittance coefficient of the Mth unit in the super-intelligent surface layer. .
3. The method for optimizing the number of SIM layers in a MIMO system under mining conditions as described in claim 2, characterized in that, The optimization problem is divided into three sub-problems: solving for the number of SIM layers, power allocation optimization, and phase shift matrix optimization. The three sub-problems are solved iteratively in turn. In each round of iteration, the number of SIM layers is solved first.
4. The method for optimizing the number of SIM layers in a MIMO system under mining conditions as described in claim 3, characterized in that, Solving for the number of SIM layers includes: Will and Set to a fixed value; make , Generate a subproblem for solving the number of SIM layers with L as the only variable: ， in, This represents the system energy efficiency with L as the variable. Indicates and The overall system coefficient obtained by combining irrelevant terms; , , and All of these are constants related to the number of units, the number of users, and the channel. This indicates that electromagnetic waves generate a loss factor with each layer of the ultra-intelligent surface layer they pass through. This represents the maximum feasible number of layers obtained by solving based on power constraints. This represents a fixed power consumption independent of the number of SIM layers.
5. The method for optimizing the number of SIM layers in a MIMO system under mining conditions as described in claim 4, characterized in that, Solving the subproblem of determining the number of SIM layers includes: For optimization problems involving nonlinear, differentiable fractional functions, Newton's iteration method is used to find the extreme points in the continuous domain. ; Filter neighboring integers within the feasible range , and As a candidate layer number; Calculate the system energy efficiency for each candidate layer number Choose the system with the highest energy efficiency. Corresponding candidate layer number As the optimal number of layers.
6. The method for optimizing the number of SIM layers in a MIMO system under mining conditions as described in claim 5, characterized in that, Power allocation optimization includes: Set the number of SIM layers to the optimal number. ,Will Set to a fixed value; The optimization problem is transformed into a parameterized non-integer optimization problem. ;in, Represents the total reachable rate of the system. Indicates by The resulting power allocation matrix requires optimization. This represents the energy efficiency parameters that are updated iteratively. The non-integer optimization problem is convexified by the WMMSE equivalent transformation to obtain the power allocator problem: ， in, This represents the estimated value of the received signal. This represents the weight matrix introduced in relation to the data flow. This indicates the number of antennas for user k. This represents the weighting factor for the d-th antenna of user k. This represents the transmitted signal of the d-th antenna of user k. To represent the estimated received signal value of user k's d-th antenna, For the transmit power budget, Indicates user No. The signal-to-interference-plus-noise ratio of each antenna; The power allocation subproblem is solved using the interior-point method to obtain the optimal power allocation matrix. .
7. The method for optimizing the number of SIM layers in a MIMO system under mining conditions as described in claim 6, characterized in that, Phase shift matrix optimization includes: Set the number of SIM layers to the optimal number. Set the power allocation matrix as the optimal power allocation matrix. This leads to the phase shift matrix optimization subproblem: ， in, Indicates by The phase shift matrix formed, Represents the optimal power allocation matrix The transmit power of the base station to the d-th antenna of user k. This represents the channel state between the base station and the d-th antenna of user k. This represents the transmission coefficient matrix between the d-th antenna of user k. Indicates the first in SIM l The phase shift matrix of the layer; Solve the phase shift matrix optimization subproblem to obtain the optimal phase shift matrix.
8. A SIM layer optimization device for a MIMO system in a mining environment, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1-7.