Intelligent metasurface enabled cellular-free large-scale MIMO network resource efficiency optimization method

Abstract

The invention provides an intelligent metasurface enabled cellular-free large-scale MIMO (Multiple Input Multiple Output) network resource efficiency optimization method. The method comprises the following steps: establishing an intelligent metasurface enabled cellular-free large-scale MIMO network model under a Rician fading channel; based on the cellular-free large-scale MIMO network model, performing channel estimation by using a linear minimum mean square error method; based on the cellular-free large-scale MIMO network model, constructing a resource efficiency closed expression when a maximum ratio merging receiving scheme is used; constructing an optimization problem formula for maximizing the network resource efficiency according to the resource efficiency closed expression; and according to the optimization problem expression, designing an optimization method for cooperatively controlling a user power control coefficient and intelligent metasurface phase shift. The spectrum efficiency and the energy efficiency of the intelligent metasurface energized cellular-free large-scale MIMO network are improved, the compromise between the spectrum efficiency and the energy efficiency is effectively realized, the optimization of the network performance is realized, and an important theoretical support is provided for the development of a 6G Internet of Things network.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and more particularly to a method for optimizing resource efficiency in cellular-free massive MIMO networks powered by intelligent metasurfaces. Background Technology

[0002] With the rapid growth in demand for wireless communication, traditional cellular networks face significant challenges due to scarce spectrum resources and severe inter-cell interference. To address these challenges, cellular-free massive MIMO networks have emerged. By deploying a large number of distributed access points, they simultaneously serve multiple users on the same time-frequency resources, effectively improving signal quality and system capacity through macro diversity. Despite these advantages, cellular-free massive MIMO networks also have significant limitations. Firstly, the deployment of numerous access points leads to a substantial increase in hardware costs, energy consumption, and signal processing complexity. Secondly, in complex wireless propagation environments, signals are susceptible to multipath fading and shadowing effects, reducing communication quality and reliability.

[0003] Intelligent metasurfaces (RIS), as an emerging technology, offer a new approach to solving the aforementioned problems. RIS consists of numerous low-cost, passive reflective elements that can flexibly control the phase and amplitude of reflected signals through software programming, thereby altering the wireless propagation environment, enhancing signal strength, and reducing interference. Introducing RIS into non-cellular massive MIMO networks holds promise for further improving network performance while reducing system costs and energy consumption.

[0004] Currently, some research has explored non-cellular massive MIMO networks powered by smart metasurfaces (RIS), but many problems remain to be solved. For example, existing resource allocation algorithms fail to fully consider the characteristics of smart metasurfaces (RIS), resulting in low network resource utilization. The introduction of smart metasurfaces (RIS) makes channel models more complex, making it difficult to meet high-precision requirements using traditional channel estimation methods to overcome practical problems such as imperfect channel state information. Furthermore, current research on the collaborative optimization problem between smart metasurfaces (RIS) and users is relatively limited, and no effective solution has yet been developed. Against this backdrop, developing a method that can fully leverage the resource efficiency of non-cellular massive MIMO networks powered by smart metasurfaces (RIS) has significant practical implications and application value. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a method for optimizing resource efficiency in cellular-free massive MIMO networks powered by intelligent metasurfaces. This invention obtains imperfect channel state information using the Linear Least Mean Square Error (LMMSE) method and, combined with the Maximum Ratio Combining (MRC) receiver scheme, derives a closed-form expression for network resource efficiency. Furthermore, it designs an alternating optimization algorithm to jointly optimize the power control coefficient and the phase shift of the RIS (Resource Optimization Scale), effectively improving the spectral efficiency and energy efficiency of the RIS-enabled cellular-free massive MIMO network. This effectively achieves a trade-off between spectral efficiency and energy efficiency, thereby optimizing network performance and providing important theoretical support for the development of 6G IoT networks.

[0006] The technical means employed in this invention are as follows: A method for optimizing resource efficiency in cellular-free massive MIMO networks empowered by intelligent metasurfaces includes: S1. Establish a cellular-free large-scale MIMO network model empowered by intelligent metasurfaces (RIS) under Ricean fading channels; S2. Based on the aforementioned non-cellular massive MIMO network model, channel estimation is performed using the linear minimum mean square error (LMMSE) method. S3. Based on the aforementioned non-cellular massive MIMO network model, construct a closed-form expression for resource efficiency (RE) when using the maximum ratio combining (MRC) reception scheme; S4. Based on the closed-form expression of resource efficiency (RE), construct the optimization problem formula for maximizing network resource efficiency; S5. Based on the optimization problem, design optimization methods for the collaborative control user power control coefficients and the phase shift of the intelligent metasurface (RIS).

[0007] Further, step S1 specifically includes: S11. Construct a network model, which consists of... Base stations (BS) A smart metasurface (RIS) and Composed of several Internet of Things (IoT) users, each base station is configured with Each smart metasurface is equipped with a root antenna. Each user is equipped with a single antenna and one passive reflector element. All base stations and smart metasurfaces are interconnected with the central processing unit (CPU) via backhaul links. A data-sharing transmission strategy is used for collaborative processing. The network operates in time-division duplex mode, with a channel coherence interval length of [missing information]. The pilot length is The rest One symbol is used for data transmission; S12. Represent the base station, user, smart metasurface, and the set of reflection units for each smart metasurface as follows: , , and The total elements of the intelligent metasurface are ,in ; S13, Assuming all The user to the If the channels between intelligent metasurfaces satisfy the line-of-sight (LoS) communication scenario, then all The user to the Channel matrix of a smart metasurface Represented as:

[0008] in, Indicates the first The user to the Large-scale path loss coefficient of an intelligent metasurface. Indicates the first The user to the A deterministic line-of-sight channel vector for a smart metasurface; S14. Using Ricean decay to represent the first... The first intelligent metasurface to the first The channel of each base station, then the channel Represented as:

[0009] in, , , It is the path loss factor. It is the Rice factor, representing the line-of-sight component. Intensity and non-line-of-sight components The strength ratio; in addition, All elements in the array follow an independent and identically distributed complex Gaussian distribution. ; S15. Assuming the channel between the user and the base station also follows Ricean fading, then all The user to the Channel between base stations Represented as:

[0010] in, , , Indicates the first The user to the Path loss factor for each base station Indicates the first The user to the Rice factor of each base station; Represents the line-of-sight component. This represents the non-line-of-sight component, where each element of the non-line-of-sight component is an independent and identically distributed complex Gaussian distribution. ; S16. Based on the formulas in steps S13, S14, and S15, all The user to the Aggregation channels of individual base stations for:

[0011] in, Represents a diagonal matrix. Indicates the first Phase shift of a smart metasurface.

[0012] Further, step S2 specifically includes: S21. During each channel coherence time interval, all users transmit mutually orthogonal pilot sequences. ,in For users The pilot, and the pilot length The pilot sequence satisfies Therefore, the first Pilot signals received by each base station Represented as:

[0013] in, This indicates the uplink pilot transmit power for all users. Indicates the first The additive white Gaussian noise at each base station has elements that are independently and identically distributed. ; S22. Based on the pilot signal received in step S21, in order to estimate the aggregation channel... , the formula and Multiplying them together gives:

[0014] S23. Based on step S22, the aggregated channel is obtained by applying the LMMSE channel estimation method. The estimated channel information is as follows:

[0015] in, , , , , ; S24, from the first The user to the The channel estimation and channel estimation error of the aggregation channel of each base station are respectively and ,in and They are and The List, and It is statistically independent and has the following statistical properties:

[0016]

[0017] in, yes The column, and , , , .

[0018] Further, step S3 specifically includes: S31. During the uplink data transmission phase, all Individual users With the assistance of intelligent metasurfaces, simultaneously to all The first base station transmits data signals. Therefore, the first... The signals received by each base station are:

[0019] in, Indicates the uplink data transmission power. Indicates the first The power control coefficient for each user, and ,vector Representation matrix The List, Indicates the first The data signals transmitted by each user satisfy and , Represents the expected value and the noise vector. Indicates the first Additive white Gaussian noise at each base station, with variance of... ; S32, Based on step S31, from all The signal from the first base station is forwarded to the central processing unit, which eventually receives the signal from the second base station. The data for each user is as follows:

[0020] To perform uplink data detection, an MRC receiver scheme is used, wherein the detection matrix... , It is a matrix The List, , Indicates the desired signal. This represents the uncertainty in beamforming gain. This indicates interference from other users. Indicates noise interference; S33. Based on the formula in step S32, calculate the... The signal-to-noise ratio (SINR) for each user is:

[0021] S34. Based on the formula in step S33, in a cellular-free large-scale MIMO network empowered by a smart metasurface, all The expression for the total spectral efficiency (SE) of a user is:

[0022] in, This indicates the effective transmission ratio within each coherent time interval; S35. Based on the formula in step S33, the total energy efficiency (EE) expression for the cellular-free large-scale MIMO network empowered by intelligent metasurface is:

[0023] in, Indicates transmission bandwidth. Indicates the first The efficiency of the transmit power amplifier used by each user, and meets the requirements. , Indicates total static power consumption. ,constant Indicates the first The first base station The static circuit power consumption of the antenna. Indicates the first The static circuit power consumption of a user, a constant Indicates the first The first intelligent metasurface (RIS) on Control power consumption related to each element; S36. Because the units of total spectral efficiency (SE) and total energy efficiency (EE) are inconsistent, it is physically inappropriate to directly combine them. To address this issue, a resource efficiency (RE) metric with consistent units is introduced, defined as:

[0024] in, and It is a weighting factor, and satisfies The denominator in the first term This represents the maximum power consumption and is considered a constant. S37. Based on the formula in step S36, by adopting the MRC receiving scheme, the closed-form expression for the resource efficiency (RE) of the smart metasurface-enabled cellular-free massive MIMO network is:

[0025] in, , , , , yes The List, , Indicates XOR; S38. Based on the formula in step S37, the closed-form expression for the resource efficiency (RE) of a cellular-free massive MIMO network empowered by intelligent metasurfaces under the MRC receiving scheme is equivalently rewritten as follows:

[0026] in: , , , , , , , , , , .

[0027] Further, step S4 specifically includes: S41. The resource efficiency (RE) optimization problem of using the MRC scheme in a cellular-free massive MIMO network powered by intelligent metasurfaces (RIS) is formulated as follows:

[0028] S42. Based on the formula in step S41, due to constraints... allow Take any value, when the power control coefficient When fixed, the resource efficiency (RE) optimization problem is transformed into optimizing the phase of the intelligent metasurface (RIS) reflective unit. sub-problems, Simplify to an unconstrained optimization problem ,Right now:

[0029] S43. Based on the formula in step S41, when the intelligent metasurface (RIS) reflective unit phase shifts... When fixed, the resource efficiency (RE) optimization problem is transformed into optimizing the power control coefficient. sub-problems, Simplified to ::

[0030] S44. Based on the formula in step S43, due to constraints... It is convex, and the objective function It is non-convex, and by applying multidimensional quadratic transformation techniques, it can be... Transform it into a convex optimization problem ::

[0031] in, , and It is an auxiliary variable introduced through a multidimensional quadratic transformation, under a fixed... In this case, and The expression for determining the optimal value is:

[0032]

[0033] S45. By using the accelerated gradient ascent algorithm to solve the optimization problem in step S42, the local optimal phase shift is obtained. The local optimal power control coefficients are obtained by using the multidimensional quadratic transformation technique to solve the optimization problem of S44. The two algorithms are iterated alternately until they converge to achieve the maximum resource efficiency.

[0034] Further, step S5 specifically includes: S51, Parameter initialization settings , , and ; S52. Initialize the parameters of the accelerated gradient ascent algorithm. , , , , ; S53, in a fixed position In the case of updating ,as follows:

[0035] in: , ,

[0036]

[0037] in, Represents the Hadamard product, a function about The partial derivative is denoted as ; S54. Set the step size according to the backtracking search method. ,renew ; S55, Update ,in ; S56, will and Substitute into the formula The objective function, when this value converges, will As output, execute step S57; otherwise, set... ,Will As output, execute step S53; S57. Initialize the technical parameters of the multidimensional quadratic transformation. , and ; S58, in a fixed position In the case of using the formula renew and use the formula renew ; S59, By substituting parameters , and Solving convex problems renew ; S510, will and Introducing the convex optimization problem The objective function, when this value converges, will As output, execute step S511; otherwise, set... ,Will As output, proceed to step S58; S511, when At that time, and As output, terminate the iteration process; otherwise, set... ,Will and As output, execute step S52.

[0038] Compared with the prior art, the present invention has the following advantages: 1. The present invention provides a resource efficiency optimization method for cellular-free massive MIMO networks empowered by intelligent metasurfaces. Unlike existing technologies that only focus on a single performance indicator, the present invention considers the joint optimization of two performance indicators, spectral efficiency and energy efficiency. By maximizing their weighted sum, i.e. resource efficiency, the spectral efficiency and energy efficiency of cellular-free massive MIMO networks are improved simultaneously, effectively solving the imbalance problem between the two performance indicators. This has certain practical significance for actual network design.

[0039] 2. The present invention provides a resource efficiency optimization method for non-cellular massive MIMO networks empowered by intelligent metasurfaces. By designing an alternating optimization algorithm of accelerated gradient ascent and multidimensional quadratic transformation to coordinate the control of the power control coefficients of different users and the phase shift of intelligent metasurfaces (RIS), the effective resource utilization of non-cellular massive MIMO networks is improved. Simulation experiments also verify that the algorithm of the present invention has effectiveness and fast convergence.

[0040] Based on the above reasons, this invention can be widely applied in fields such as wireless communication. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart of the method of the present invention.

[0043] Figure 2 The model structure of a cellular-free large-scale MIMO network powered by a smart metasurface (RIS) to which the present invention is applicable.

[0044] Figure 3 A simulation diagram showing the relationship between network resource efficiency and uplink data transmission power provided in an embodiment of the present invention.

[0045] Figure 4 The simulation diagram shows the relationship between network resource efficiency and weight factor under different resource allocation algorithms, as provided in the embodiments of the present invention.

[0046] Figure 5 The simulation diagram shows the relationship between the number of iterations and resource efficiency of the alternating optimization algorithm provided in the embodiment of the present invention. Detailed Implementation

[0047] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0048] It should be noted that the terms "comprising" and "having" and any variations thereof in the specification, claims and accompanying drawings of this invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such processes, methods, products or devices.

[0049] like Figure 1 As shown, this invention provides a method for optimizing resource efficiency in cellular-free massive MIMO networks powered by intelligent metasurfaces (RIS), comprising: S1. Establish a cellular-free large-scale MIMO network model empowered by intelligent metasurfaces (RIS) under Ricean fading channels; S2. Based on the aforementioned non-cellular massive MIMO network model, channel estimation is performed using the linear minimum mean square error (LMMSE) method. S3. Based on the aforementioned non-cellular massive MIMO network model, construct a closed-form expression for resource efficiency (RE) when using the maximum ratio combining (MRC) reception scheme; S4. Based on the closed-form expression of resource efficiency (RE), construct the optimization problem formula for maximizing network resource efficiency; S5. Based on the optimization problem, design optimization methods for the user power control coefficients and RIS phase shift of the collaborative control system.

[0050] In a specific implementation, as a preferred embodiment of the present invention, step S1 specifically includes: S11. Construct a network model, which consists of... Base stations (BS) RIS and Composed of several Internet of Things (IoT) users, each base station is configured with Each smart metasurface is equipped with a root antenna. Each user is equipped with a single antenna and one passive reflector element. All base stations and smart metasurfaces are interconnected with the central processing unit (CPU) via backhaul links. A data-sharing transmission strategy is used for collaborative processing. The network operates in time-division duplex mode, with a channel coherence interval length of [missing information]. The pilot length is The rest One symbol is used for data transmission; S12. Represent the base station, user, smart metasurface, and the set of reflection units for each smart metasurface as follows: , , and The total elements of the intelligent metasurface are ,in ; S13, Assuming all The user to the If the channels between intelligent metasurfaces satisfy the line-of-sight (LoS) communication scenario, then all The user to the Channel matrix of a smart metasurface Represented as:

[0051] in, Indicates the first The user to the Large-scale path loss coefficient of an intelligent metasurface. Indicates the first The user to the A deterministic line-of-sight channel vector for a smart metasurface; S14. Using Ricean decay to represent the first... The first intelligent metasurface to the first The channel of each base station, then the channel Represented as:

[0052] in, , , It is the path loss factor. It is the Rice factor, representing the line-of-sight component. Intensity and non-line-of-sight components The strength ratio; in addition, All elements in the array follow an independent and identically distributed complex Gaussian distribution. ; S15. Assuming the channel between the user and the base station also follows Ricean fading, then all The user to the Channel between base stations Represented as:

[0053] in, , , Indicates the first The user to the Path loss factor for each base station Indicates the first The user to the Rice factor of each base station; Represents the line-of-sight component. This represents the non-line-of-sight component, where each element of the non-line-of-sight component is an independent and identically distributed complex Gaussian distribution. ; S16. Based on the formulas in steps S13, S14, and S15, all The user to the Aggregation channels of individual base stations for:

[0054] in, Represents a diagonal matrix. Indicates the first Phase shift of a smart metasurface.

[0055] In a specific implementation, as a preferred embodiment of the present invention, step S2 specifically includes: S21. During each channel coherence time interval, all users transmit mutually orthogonal pilot sequences. ,in For users The pilot, and the pilot length The pilot sequence satisfies Therefore, the first Pilot signals received by each base station Represented as:

[0056] in, This indicates the uplink pilot transmit power for all users. Indicates the first The additive white Gaussian noise (AWGN) at each base station has elements that are independently and identically distributed. ; S22. Based on the pilot signal received in step S21, in order to estimate the aggregation channel... , the formula and Multiplying them together gives:

[0057] S23. Based on step S22, the aggregated channel is obtained by applying the LMMSE channel estimation method. The estimated channel information is as follows:

[0058] in, , , , , ; S24, from the first The user to the The channel estimation and channel estimation error of the aggregation channel of each base station are respectively and ,in and They are and The List, and It is statistically independent and has the following statistical properties:

[0059]

[0060] in, yes The column, and , , , .

[0061] In a specific implementation, as a preferred embodiment of the present invention, step S3 specifically includes: S31. During the uplink data transmission phase, all Individual users With the assistance of intelligent metasurfaces, simultaneously to all The first base station transmits data signals. Therefore, the first... The signals received by each base station are:

[0062] in, Indicates the uplink data transmission power. Indicates the first The power control coefficient for each user, and ,vector Representation matrix The List, Indicates the first The data signals transmitted by each user satisfy and , Represents the expected value and the noise vector. Indicates the first Additive white Gaussian noise at each base station, with variance of... ; S32, Based on step S31, from all The signal from the first base station is forwarded to the central processing unit, which eventually receives the signal from the second base station. The data for each user is as follows:

[0063] To perform uplink data detection, an MRC receiver scheme is used, wherein the detection matrix... , It is a matrix The List, , Indicates the desired signal. This represents the uncertainty in beamforming gain. This indicates interference from other users. Indicates noise interference; S33. Based on the formula in step S32, calculate the... The signal-to-noise ratio (SINR) for each user is:

[0064] S34. Based on the formula in step S33, in a cellular-free large-scale MIMO network empowered by a smart metasurface, all The expression for the total spectral efficiency (SE) of a user is:

[0065] in, This indicates the effective transmission ratio within each coherent time interval; S35. Based on the formula in step S33, the total energy efficiency (EE) expression for the cellular-free large-scale MIMO network empowered by intelligent metasurface is:

[0066] in, Indicates transmission bandwidth. Indicates the first The efficiency of the transmit power amplifier used by each user, and meets the requirements. , Indicates total static power consumption. ,constant Indicates the first The first base station The static circuit power consumption of the antenna. Indicates the first The static circuit power consumption of a user, a constant Indicates the first The first intelligent metasurface (RIS) on Control power consumption related to each element; S36. Because the units of total spectral efficiency (SE) and total energy efficiency (EE) are inconsistent, it is physically inappropriate to directly combine them. To address this issue, a resource efficiency (RE) metric with consistent units is introduced, defined as:

[0067] in, and It is a weighting factor, and satisfies The denominator in the first term This represents the maximum power consumption and is considered a constant. S37. Based on the formula in step S36, by adopting the MRC receiving scheme, the closed-form expression for the resource efficiency (RE) of the smart metasurface-enabled cellular-free massive MIMO network is:

[0068] in, , , , , yes The List, , Indicates XOR; S38. Based on the formula in step S37, the closed-form expression for the resource efficiency (RE) of a cellular-free massive MIMO network empowered by intelligent metasurfaces under the MRC receiving scheme is equivalently rewritten as follows:

[0069] in: , , , , , , , , , , .

[0070] In a specific implementation, as a preferred embodiment of the present invention, step S4 specifically includes: S41. The resource efficiency (RE) optimization problem of using the MRC scheme in a cellular-free massive MIMO network powered by intelligent metasurfaces (RIS) is formulated as follows:

[0071] S42. Based on the formula in step S41, due to constraints... allow Take any value, when the power control coefficient When fixed, the resource efficiency (RE) optimization problem is transformed into optimizing the phase of the intelligent metasurface (RIS) reflective unit. sub-problems, Simplify to an unconstrained optimization problem ,Right now:

[0072] S43. Based on the formula in step S41, when the intelligent metasurface (RIS) reflective unit phase shifts... When fixed, the resource efficiency (RE) optimization problem is transformed into optimizing the power control coefficient. sub-problems, Simplified to :

[0073] S44. Based on the formula in step S43, due to constraints... It is convex, and the objective function It is non-convex, and by applying multidimensional quadratic transformation techniques, it can be... Transform it into a convex optimization problem :

[0074] in, , and It is an auxiliary variable introduced through a multidimensional quadratic transformation, under a fixed... In this case, and The expression for determining the optimal value is:

[0075]

[0076] S45. By using the accelerated gradient ascent algorithm to solve the optimization problem in step S42, the local optimal phase shift is obtained. The local optimal power control coefficients are obtained by using the multidimensional quadratic transformation technique to solve the optimization problem of S44. The two algorithms are iterated alternately until they converge to achieve the maximum resource efficiency.

[0077] In a specific implementation, as a preferred embodiment of the present invention, step S5 specifically includes: S51, Parameter initialization settings , , and ; S52. Initialize the parameters of the accelerated gradient ascent algorithm. , , , , ; S53, in a fixed position In the case of updating ,as follows:

[0078] in: , ,

[0079]

[0080] in, Represents the Hadamard product, a function about The partial derivative is denoted as ; S54. Set the step size according to the backtracking search method. ,renew ; S55, Update ,in ; S56, will and Substitute into the formula The objective function, when this value converges, will As output, execute step S57; otherwise, set... ,Will As output, execute step S53; S57. Initialize the technical parameters of the multidimensional quadratic transformation. , and ; S58, in a fixed position In the case of using the formula renew and use the formula renew ; S59, By substituting parameters , and Solving convex problems renew ; S510, will and Introducing the convex optimization problem The objective function, when this value converges, will As output, execute step S511; otherwise, set... ,Will As output, proceed to step S58; S511, when At that time, and As output, terminate the iteration process; otherwise, set... ,Will and As output, execute step S52.

[0081] Example To verify the effectiveness of the present invention, the following simulation experiments were conducted. The specific simulation parameters for the cellular-free massive MIMO network powered by intelligent metasurfaces (RIS) are as follows: For RIS-enabled cellular-free massive MIMO network simulation scenarios, such as Figure 2 As shown, settings Each base station (BS) is configured with one base station (BS). One antenna, respectively deployed at , and Location. Building surface. A smart metasurface (RIS) is deployed in , and Location, each intelligent metasurface (RIS) is composed of It consists of passive reflective elements. All Users are randomly distributed in the following groups: Centered on, with radius Within the circular area, the user's uniform height is The large-scale path loss factor is , and ,in , and They represent the first The user to the The first intelligent metasurface (RIS), the first The first intelligent metasurface (RIS) to the first The first base station (BS), the first The user to the The distance between base stations (BS). Other simulation parameters include: , , , , , , , , , For all In the random phase shift (RPS) reference scheme, the phase shift of the intelligent metasurface (RIS) is set as follows: In the full power control (FPT) baseline scheme, the power control coefficients for all users are... To balance the trade-off between total spectral efficiency (SE) and total energy efficiency (EE), a maximum weighting factor is defined as:

[0082] When all users transmit at full power and the weighting factor is set to 1, At that time, the resource efficiency (RE) metric was simplified to total spectral efficiency (SE), ignoring the total energy efficiency (EE) component.

[0083] Figure 3Simulation graphs showing the relationship between network resource efficiency and uplink data transmission power were plotted. It can be seen that the theoretical closed-form expression is highly consistent with the Monte Carlo simulation results, proving the accuracy of the theoretical expression. Resource efficiency (RE) increases with power, first reaching a peak and then decreasing, or monotonically increasing but quickly stabilizing at a certain constant. The results indicate that blindly increasing power not only fails to continuously improve resource efficiency but may also lead to efficiency loss due to ineffective energy consumption, thus highlighting the crucial role of precise control of uplink power in resource allocation.

[0084] Figure 4 Resource efficiency (RE) performance under different resource allocation algorithms was plotted. These include schemes based on random phase shift and full power, schemes based on full power and accelerated gradient ascent (Algorithm 1), schemes based on random phase shift and multidimensional quadratic transformation (Algorithm 2), schemes based on alternating optimization algorithms of accelerated gradient ascent and multidimensional quadratic transformation (Algorithm 3), and schemes based on genetic algorithms. Compared to the baseline scheme of random phase shift and full power, both the accelerated gradient ascent algorithm and the multidimensional quadratic transformation technique can significantly improve resource efficiency (RE) when applied alone. Furthermore, compared to the four baseline schemes (RPS+FPT, RPS+Algorithm 2, FPT+Algorithm 1, and genetic algorithm), Algorithm 3 consistently achieves the highest resource efficiency (RE). Moreover, with… As the number of ... Under these conditions, using Algorithm 1 alone becomes a practical and efficient strategy for maximizing resource efficiency (RE). In contrast, under low... Under these conditions, Algorithm 3 remains the optimal choice. These findings validate the effectiveness of Algorithms 1-3 in improving resource efficiency (RE) and provide valuable insights into selecting appropriate optimization strategies under different network parameters.

[0085] Figure 5 Simulation plots were created showing the relationship between the number of iterations and resource efficiency in the alternating optimization algorithm. The algorithm demonstrates its convergence within a very short iteration cycle; furthermore, resource efficiency increases with... As the value of the optimization process gradually increases, the resource efficiency (RE) value at the end of the optimization process is significantly improved compared with its initial iteration point. This significant performance gain demonstrates the effectiveness and robustness of the alternating optimization algorithm.

[0086] In summary, this invention verifies the correctness and mathematical rigor of the theoretical expression of resource efficiency (RE) through simulation. The alternating optimization algorithm has a significant gain effect on network resource efficiency (RE) performance. The optimal strategy under different parameter scenarios is the key to the performance leap. At the same time, the fast convergence, effectiveness and stability of the algorithm fully demonstrate that the method of this invention has practical application significance and theoretical guidance.

[0087] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing resource efficiency in cellular-free massive MIMO networks empowered by intelligent metasurfaces, characterized in that, include: S1. Establish a cellular-free large-scale MIMO network model empowered by intelligent metasurfaces under Ricean fading channels; S2. Based on the aforementioned non-cellular massive MIMO network model, channel estimation is performed using the linear minimum mean square error method; S3. Based on the aforementioned non-cellular massive MIMO network model, construct a closed-form expression for resource efficiency when using the maximum ratio combining reception scheme. S4. Based on the closed-form expression for resource efficiency, construct an optimization problem that maximizes network resource efficiency; S5. Based on the optimization problem, design optimization methods for the collaborative control user power control coefficient and the intelligent metasurface phase shift.

2. The method for optimizing resource efficiency in a cellular-free massive MIMO network empowered by intelligent metasurfaces according to claim 1, characterized in that, Step S1 specifically includes: S11. Construct a network model, which consists of... One base station, A smart metasurface and Composed of IoT users, each base station is configured with Each smart metasurface is equipped with a root antenna. Each user is equipped with a single antenna and one passive reflector element. All base stations and smart metasurfaces are interconnected with the central processing unit via backhaul links. A data-sharing transmission strategy is used for collaborative processing. The network operates in time-division duplex mode, with a channel coherence interval length of [missing information]. The pilot length is The rest One symbol is used for data transmission; S12. Represent the base station, user, smart metasurface, and the set of reflection units for each smart metasurface as follows: , , and The total elements of the intelligent metasurface are ,in ; S13, Assuming all The user to the If the channel between several smart metasurfaces satisfies the line-of-sight communication scenario, then all The user to the Channel matrix of a smart metasurface Represented as: in, Indicates the first The user to the Large-scale path loss coefficient of an intelligent metasurface. Indicates the first The user to the A deterministic line-of-sight channel vector for a smart metasurface; S14. Using Ricean decay to represent the first... The first intelligent metasurface to the first The channel of each base station, then the channel Represented as: in, , , It is the path loss factor. It is the Rice factor, representing the line-of-sight component. Intensity and non-line-of-sight components The strength ratio; in addition, All elements in the array follow an independent and identically distributed complex Gaussian distribution. ; S15. Assuming the channel between the user and the base station also follows Ricean fading, then all The user to the Channel between base stations Represented as: in, , , Indicates the first The user to the Path loss factor for each base station Indicates the first The user to the Rice factor of each base station; Represents the line-of-sight component. This represents the non-line-of-sight component, where each element of the non-line-of-sight component is an independent and identically distributed complex Gaussian distribution. ; S16. Based on the formulas in steps S13, S14, and S15, all The user to the Aggregation channels of individual base stations for: in, Represents a diagonal matrix. Indicates the first Phase shift of a smart metasurface.

3. The method for optimizing resource efficiency in a cellular-free massive MIMO network empowered by intelligent metasurfaces according to claim 1, characterized in that, Step S2 specifically includes: S21. During each channel coherence time interval, all users transmit mutually orthogonal pilot sequences. ,in For users The pilot, and the pilot length The pilot sequence satisfies Therefore, the first Pilot signals received by each base station Represented as: in, This indicates the uplink pilot transmit power for all users. Indicates the first The additive white Gaussian noise at each base station has elements that are independently and identically distributed. ; S22. Based on the pilot signal received in step S21, in order to estimate the aggregation channel... , the formula and Multiplying them together gives: S23. Based on step S22, the aggregated channel is obtained by applying the LMMSE channel estimation method. The estimated channel information is as follows: in, , , , , ; S24, from the first The user to the The channel estimation and channel estimation error of the aggregation channel of each base station are respectively and ,in and They are and The List, and It is statistically independent and has the following statistical properties: in, yes The column, and , , , .

4. The method for optimizing resource efficiency in a cellular-free massive MIMO network empowered by intelligent metasurfaces according to claim 1, characterized in that, Step S3 specifically includes: S31. During the uplink data transmission phase, all Individual users With the assistance of intelligent metasurfaces, simultaneously to all The first base station transmits data signals. Therefore, the first... The signals received by each base station are: in, Indicates the uplink data transmission power. Indicates the first The power control coefficient for each user, and ,vector Representation matrix The List, Indicates the first The data signals transmitted by each user satisfy and , Represents the expected value and the noise vector. Indicates the first Additive white Gaussian noise at each base station, with variance of... ; S32, Based on step S31, from all The signal from the first base station is forwarded to the central processing unit, which eventually receives the signal from the second base station. The data for each user is as follows: To perform uplink data detection, an MRC receiver scheme is used, wherein the detection matrix... , It is a matrix The List, , Indicates the desired signal. This represents the uncertainty in beamforming gain. This indicates interference from other users. Indicates noise interference; S33. Based on the formula in step S32, calculate the... The signal-to-noise ratio for each user is: S34. Based on the formula in step S33, in a cellular-free large-scale MIMO network empowered by a smart metasurface, all The expression for the total spectral efficiency of a user is: in, This indicates the effective transmission ratio within each coherent time interval; S35. Based on the formula in step S33, the overall energy efficiency expression of the cellular-free large-scale MIMO network empowered by intelligent metasurface is: in, Indicates transmission bandwidth. Indicates the first The efficiency of the transmit power amplifier used by each user, and meets the requirements. , Indicates total static power consumption. ,constant Indicates the first The first base station The static circuit power consumption of the antenna. Indicates the first The static circuit power consumption of a user, a constant Indicates the first The first intelligent metasurface Control power consumption related to each element; S36. Introduce a resource efficiency metric with consistent units, defined as: in, and It is a weighting factor, and satisfies The denominator in the first term This represents the maximum power consumption and is considered a constant. S37. Based on the formula in step S36, by adopting the MRC receiving scheme, the closed-form expression for the resource efficiency of the smart metasurface-enabled cellular-free massive MIMO network is: in, , , , , yes The List, , Indicates XOR; S38. Based on the formula in step S37, the closed-form expression for the resource efficiency of a cellular-free massive MIMO network empowered by intelligent metasurfaces under the MRC receiving scheme is equivalently rewritten as follows: in: ， ， ， ， ， ， ， ， ， ， 。 5. The method for optimizing resource efficiency in a cellular-free massive MIMO network empowered by intelligent metasurfaces according to claim 1, characterized in that, Step S4 specifically includes: S41. The resource efficiency optimization problem of using the MRC scheme in a cellular-free massive MIMO network empowered by intelligent metasurfaces is expressed as: S42. Based on the formula in step S41, due to constraints... allow Take any value, when the power control coefficient When the resource efficiency optimization problem is fixed, it is transformed into optimizing the phase of the intelligent metasurface reflection unit. sub-problems, Simplify to an unconstrained optimization problem ,Right now: S43. Based on the formula in step S41, when the intelligent metasurface reflection unit undergoes a phase shift... When fixed, the resource efficiency optimization problem is transformed into optimizing the power control coefficient. sub-problems, Simplified to : S44. Based on the formula in step S43, due to constraints... It is convex, and the objective function It is non-convex, and by applying multidimensional quadratic transformation techniques, it can be... Transform it into a convex optimization problem : in, , and It is an auxiliary variable introduced through a multidimensional quadratic transformation, under a fixed... In this case, and The expression for determining the optimal value is: S45. By using the accelerated gradient ascent algorithm to solve the optimization problem in step S42, the local optimal phase shift is obtained. The local optimal power control coefficients are obtained by using the multidimensional quadratic transformation technique to solve the optimization problem of S44. The two algorithms are iterated alternately until they converge to achieve the maximum resource efficiency.

6. The method for optimizing resource efficiency in a cellular-free massive MIMO network empowered by intelligent metasurfaces according to claim 1, characterized in that, Step S5 specifically includes: S51, Parameter initialization settings , , and ; S52. Initialize the parameters for the accelerated gradient ascent algorithm. , , , , ; S53, in a fixed position In the case of updating ,as follows: in: ， ， in, Represents the Hadamard product, a function about The partial derivative is denoted as ; S54. Set the step size according to the backtracking search method. ,renew ; S55, Update ,in ; S56, will and Substitute into the formula The objective function, when this value converges, will As output, execute step S57; otherwise, set... ,Will As output, execute step S53; S57. Initialize the technical parameters of the multidimensional quadratic transformation. , and ; S58, in a fixed position In the case of using the formula renew and use the formula renew ; S59, By substituting parameters , and Solving convex problems renew ; S510, will and Introducing the convex optimization problem The objective function, when this value converges, will As output, execute step S511; otherwise, set... ,Will As output, proceed to step S58; S511, when At that time, and As output, terminate the iteration process; otherwise, set... ,Will and As output, execute step S52.

Images