A multi-base station cooperative wireless resource allocation method based on uplink rate segmentation

Abstract

The application discloses a kind of multi-base station cooperation wireless resource allocation method based on uplink rate segmentation, it is related to electronic communication technical field, uplink multicarrier RSMA by the message of each user is segmented to multiple subcarriers, and using appropriate decoding order, to effectively utilize all available degrees of freedom;In order to deal with the non-convexity of transmission rate maximization problem and reduce centralized computing burden, a two-stage method is developed: first, a low-complexity base station selection method based on matching game is proposed, which is used to segment the user set;Next, a cooperative distributed scheme is proposed, which divides the resource allocation and beamforming tasks to the corresponding base stations;Using continuous convex approximation and dual decomposition technique, a semi-closed form solution for power and subcarrier allocation is derived, and fractional programming and alternating direction method of multipliers are used to optimize the receive beamforming, and the simulation results verify the effectiveness of the application in terms of transmission rate gain and computational complexity.

Description

Technical Field

[0001] This invention relates to the field of electronic communication technology, and in particular to a multi-base station cooperative radio resource allocation method based on uplink rate segmentation. Background Technology

[0002] In recent years, given the dramatic increase in demand for high-speed data transmission and the widespread adoption of dense wireless networks, the need for technologies that can achieve excellent resource utilization efficiency has become increasingly urgent. As a cutting-edge multiple access technology in modern multi-antenna communication networks, Rate-Splitting Multiple Access (RSMA) has become a highly promising and efficient solution due to its ability to dynamically partially decode multi-user interference and treat the remaining interference as noise.

[0003] Specifically, RSMA divides a user's information into multiple sub-information and transmits them using overlay coding, enabling multiple users to access the same orthogonal resource block and allowing arbitrary forms of joint decoding combinations. It cleverly integrates and encompasses existing methods, representing an advanced form of spatial partitioned multiple access (treating interference purely as noise), non-orthogonal multiple access (completely decoding interference), and orthogonal multiple access (allocating orthogonal radio resources to completely avoid interference). With its outstanding capabilities, RSMA is well-positioned to address many emerging challenges in the 6G field, thus attracting widespread attention from industry and academia.

[0004] However, implementing RSMA in wireless networks involves several key issues, especially interference management and decoding sequence design during information transmission. Furthermore, current research on uplink RSMA is primarily based on the pre-defined conditions of single-cell scenarios, focusing on mitigating interference within the cell.

[0005] On the other hand, due to the severe shortage of spectrum resources, the instantaneous capacity of information transmission faces enormous challenges. One effective strategy to address this problem is to achieve spatial frequency reuse in multi-cell networks by deploying dense base stations, rather than allocating independent spectrum for each cell. When transmitting synchronously on the same frequency band, the transmission power not only affects the target receiver but also causes co-channel interference to non-target receivers. Therefore, to overcome the interference control problems caused by frequency reuse, multi-cell networks must implement efficient resource management and interference coordination to ensure reliable end-to-end communication.

[0006] Although existing work has made many improvements to dense cell resource management methods, most studies are still focused on single-antenna, single-carrier and traditional multiple access scenarios, which cannot fully leverage the advantages of dense spectrum reuse. Therefore, how to maximize the uplink transmission service rate in multi-cell cellular systems has become an urgent problem to be solved. Summary of the Invention

[0007] To address the above technical problems, this invention provides a multi-base station cooperative radio resource allocation method based on uplink rate segmentation, comprising the following steps:

[0008] S1. Establish the uplink RSMA transmission model;

[0009] S2. Establish the optimization problem and constraints of a multi-cell uplink communication system;

[0010] S3. Perform a global partitioning of the user set and determine which base station each user uploads to;

[0011] S4. Solve the wireless resource allocation scheme in a distributed manner;

[0012] S5. Construct a multi-cell distributed cooperation framework based on uplink RSMA.

[0013] The technical solution further defined in this invention is:

[0014] Furthermore, in step S1, the uplink RSMA transmission model includes I multi-antenna base stations and K cellular users, equipped with Nr receiving antennas. Each single-antenna user communicates with a specific base station using its receiving antenna, and the entire spectrum is divided into N orthogonal subcarriers. In uplink RSMA, each user terminal splits its message into M parts and transmits them simultaneously to the access base station. Subsequently, the base station uses serial interference cancellation technology to decode the sub-messages of all users according to a predetermined decoding order. The sets of users, carriers, sub-messages, base stations, and antennas are defined sequentially as follows: and

[0015] Let the m-th sub-information of user k be (k, m), and the n-th subcarrier of base station i be (i, n). Let... The signal to be transmitted on (i, n) represents the sub-information (k, m).

[0016]

[0017] in, Let represent the (k, m)th sub-information on the (i, n)th subcarrier, and satisfy . express The transmission power;

[0018] Define the channel information of user k on the (i, n)th subcarrier as follows: The received signal of base station i on carrier n is represented as:

[0019]

[0020] in, Indicates receive beamforming, z i，n Indicates power as Gaussian white noise.

[0021] As described above, in a multi-base station cooperative radio resource allocation method based on uplink rate segmentation, in step S1, base station i uses continuous interference cancellation to decode all sub-information in the received signal. The entire multi-cell system contains a total of L = K × M sub-information items, where the number of sub-information items on the (i, n)th subcarrier is L. i，n satisfy

[0022] definition For all possible decoding orders of all sub-information, and optimize the decoding order variables. Where ψ I，N This represents the decoding order of the sub-information on the (i, n)th carrier, i.e. And the decoding order of the m-th sub-information of user k on the (i, n)-th carrier is as follows The base station decodes and removes code with higher priority according to the priority decoding order. The sub-message is treated as interference, while the remaining sub-messages are treated as interference; therefore, the sub-messages... The signal-to-interference-plus-noise ratio is expressed as:

[0023]

[0024] Among them, h j，i， n represents the channel information of user j on the (i, n)th carrier. This represents the transmit power of the (j, l)th sub-information on the (i, n)th subcarrier; This indicates that the decoding order on subcarrier (i, n) is later than... The set of remaining sub-messages, Let represent the intra-cell interference power suffered by sub-information (k, m) on the (i, n)th subcarrier, and

[0025]

[0026] Therefore, based on formula (3), the spectral efficiency of user k can be written as:

[0027]

[0028] in, Indicates sub-message The signal-to-interference-to-noise ratio.

[0029] As described above, in a multi-base station cooperative radio resource allocation method based on uplink rate segmentation, step S2 sets that each user can only select one base station to access, and the transmit power... The following formula must be satisfied:

[0030]

[0031] Where sgn(·) represents the symbol function; each sub-information can only occupy a single subcarrier, and this constraint is expressed as:

[0032]

[0033] Assume the maximum transmit power of each base station is P. max The non-negativity constraint for power allocation is expressed as:

[0034]

[0035] The receive beamforming at the base station is represented by a unit constant modulus constraint:

[0036]

[0037] Minimum transmission rate per user r k Must meet:

[0038]

[0039] in, Denotes the minimum transmission rate threshold for the k-th user; define d k，m Let be a predefined nonnegative value used to determine the splitting ratio of the m-th sub-message, and with the following constraints:

[0040]

[0041] The constraint on the selection of decoding order is expressed as:

[0042]

[0043] By jointly optimizing base station selection, decoding sequence, carrier allocation, and power control to maximize system performance and rate, the objective function is expressed as follows:

[0044] As described above, in a multi-base station cooperative radio resource allocation method based on uplink rate segmentation, step S3 involves defining an auxiliary variable y = [y...]. 1，1 y 1，2, ..., y K，I ]∈{0,1} KI , where y k，i =1 indicates that the k-th user has accessed the i-th base station. The constraint shown in equation (6) is rewritten as:

[0045]

[0046] A matching game is performed, using a priority list of base stations and user terminals as input data, and outputting matching relationships; in the initial stage, the associated objects of all user terminals are set to the nearest base station;

[0047] make This represents the average channel norm of the k-th user terminal across all subcarriers at the i-th base station. In each interaction of the game, the central unit comprehensively considers... The selection of base stations for users is based on the number of users in the current cell; the selection of base stations for user terminals to be accessed is designed based on the following criteria:

[0048]

[0049] in, Let represent the set of users within base station i, and obtain the user cooperation vector y according to the above formula.

[0050] As described above, in a multi-base station cooperative radio resource allocation method based on uplink rate segmentation, in step S4, given a base station y, each base station is responsible for its own resource allocation design in a distributed manner; firstly, continuous slack variables are introduced. The constraints shown in equation (7) are rewritten as follows:

[0051]

[0052] When receiving beamforming W i Given the decoding order, the joint optimization problem of power control and carrier allocation for the i-th base station is expressed as:

[0053]

[0054] st(8), (10), (12), (16)

[0055] The objective function for the above problem can be further expressed as:

[0056]

[0057] in, This indicates that the decoding order on carrier (i, n) is not less than The sub-information set, i.e.

[0058] Allocate higher data rates to higher-priority sub-messages; this is achieved by utilizing priority indexes and introducing slack variables. The problem shown in equation (17) can be equivalently rewritten as:

[0059]

[0060] Solving the above problem lies in optimizing the coupling between variables and Non-concavity; for linear products Using the log-exponential reconstruction method to and Rewritten as and Then Further rewritten as:

[0061]

[0062] because and about and It is a joint convex function, and the continuous convex approximation method is used iteratively to approximate it. By approximation, we have:

[0063]

[0064] and Represented as:

[0065]

[0066] in, express about and The constant gradient;

[0067] By combining equations (20) to (22) to replace The problem shown in equation (19) is approximated as a convex lower bound maximization problem, whose optimal solution is unique. The problem is solved by iteratively updating the lower bound until it no longer tightens, using the Lagrange dual decomposition theory.

[0068] As described above, in a multi-base station cooperative radio resource allocation method based on uplink rate segmentation, step S4 uses Lagrange dual decomposition theory to solve the problem shown in equation (19), and the corresponding Lagrange function is expressed as:

[0069]

[0070] Where, μ i , λi , ρ i v i Let Lagrange multipliers be represented; the objective function of the dual problem is expressed as:

[0071]

[0072] Among them, v i satisfy The dual problem can be simplified to:

[0073]

[0074] in, Based on the Caro-Kun-Tucker conditions, the dual problem is solved using an iterative algorithm based on gradient descent.

[0075] As described above, in a multi-base station cooperative radio resource allocation method based on uplink rate segmentation, step S4 involves solving the dual problem using an iterative algorithm based on gradient descent, according to the Caro-Kuhn-Tucker condition. First, the Lagrangian function is calculated. The gradient is:

[0076]

[0077] in,

[0078] information The iterative process of the transmit power is expressed as:

[0079]

[0080] Where (e) and ξ represent the iteration index and step size, respectively;

[0081] Then, the carrier allocation is represented as:

[0082]

[0083] The (k, m)th piece of information was assigned the maximum The corresponding carrier (i, n), that is:

[0084]

[0085] The minimization of the dual problem for each base station, as shown in equation (25), is solved using the subgradient method; the Lagrange dual variables are updated according to the following formula:

[0086]

[0087] Where (u) is the iteration index, Indicates positive step size; projection operator In the above formula, it is defined as a feasible lower bound.

[0088] As described above, in a multi-base station cooperative radio resource allocation method based on uplink rate segmentation, step S4 involves designing a receive beamforming method, first defining... and Will Rewritten as:

[0089]

[0090] This leads to the following optimization problem:

[0091]

[0092]

[0093] By utilizing the Lagrange duality transformation, the objective function of the above equation can be rewritten as:

[0094]

[0095] Introducing auxiliary variables definition

[0096] By reintroducing the auxiliary variable and employing a quadratic transformation, the above equation can be rewritten as:

[0097]

[0098] in, express The conjugate of, and The optimal solution is obtained by taking the first-order partial derivative of the above equation, that is:

[0099]

[0100] Substituting equation (38) into the problem shown in equation (35), regarding W i The problem is represented in the following form:

[0101]

[0102] in, The problem above can be solved using the alternating direction multiplier method.

[0103] As described above, in a multi-base station cooperative radio resource allocation method based on uplink rate segmentation, step S4 uses the alternating direction multiplier method to solve the problem shown in equation (39). First, w im Introducing auxiliary variable q i，nRewriting the problem shown in equation (39) in the form of an augmented Lagrangian function, we have:

[0104]

[0105] in, Indicates the possible values ​​of beamforming variables; and These represent the real and imaginary parts of a Lagrange multiplier, respectively. express Indicator functions;

[0106] The problem is decomposed into multiple local problems using the alternating direction multiplier method, and the update of each variable is represented as follows:

[0107]

[0108] in, t represents the iteration index, and κ is the step size; since The convexity will The update solution is as follows:

[0109]

[0110] Then, by removing irrelevant beam vectors, equation (42) is derived as follows:

[0111]

[0112] in, Indicates (·) in Projection on;

[0113] Finally, based on the joint optimization of channel gain and rate segmentation ratio of the user terminal, the optimal decoding order is mathematically represented by a preset decoding order optimization criterion. Specifically, the decoding order optimization criterion is as follows: for all sub-messages on the (i, n)th subcarrier, their approximately optimal decoding order follows... Sort in descending order.

[0114] The beneficial effects of this invention are:

[0115] In this invention, by dividing the message stream of each user terminal into multiple sub-messages and assigning differentiated rates, joint optimization of user association, power allocation, subcarrier scheduling, decoding order, and receive beamforming in a multi-cell, multi-carrier system is achieved, thereby maximizing system performance and rate. This not only constructs a novel uplink multi-dimensional resource allocation theoretical framework, but also verifies the superiority of RSMA technology in complex interference environments through multi-domain collaborative optimization of space-frequency-power-decoding, thereby effectively improving the throughput performance of cellular systems and demonstrating significant advantages in reducing computational complexity. Attached Figure Description

[0116] Figure 1 This is a schematic diagram of the overall process of the present invention;

[0117] Figure 2 This is a system model diagram of the present invention;

[0118] Figure 3 This is a schematic diagram showing the changes in the number of carriers versus the system and rate in the simulation experiment results of an embodiment of the present invention;

[0119] Figure 4 This is a schematic diagram illustrating the relationship between the number of different users and the system and speed in the simulation experiment results of an embodiment of the present invention. Detailed Implementation

[0120] This embodiment mainly designs a collaborative distributed multi-cell multicarrier RSMA (CDMM-RSMA) resource allocation method. In this method, by dividing the message stream of each user terminal into multiple sub-messages and allocating differentiated rates, the joint optimization of user association, power allocation, subcarrier scheduling, decoding order, and receive beamforming in a multi-cell multicarrier system is achieved, thereby maximizing the system and rate. This embodiment not only constructs a novel uplink multi-dimensional resource allocation theoretical framework, but also verifies the superiority of RSMA technology in complex interference environments through multi-domain collaborative optimization of space-frequency-power-decoding, etc.

[0121] This embodiment provides a multi-base station cooperative radio resource allocation method based on uplink rate segmentation, such as... Figure 1 As shown, it includes the following steps:

[0122] S1. Establish an uplink RSMA transmission model, such as Figure 2 As shown, the system includes I multi-antenna base stations and K cellular users. Each single-antenna user communicates with a specific base station equipped with Nr receiving antennas, and the entire spectrum is divided into N orthogonal subcarriers. In uplink RSMA, each user terminal splits its message into M parts and transmits them simultaneously to the access base station. Subsequently, the base station uses serial interference cancellation technology to decode the sub-messages of all users according to a predetermined decoding order. The sets of users, carriers, sub-messages, base stations, and antennas are defined as follows: as well as

[0123] Let the m-th sub-information of user k be (k, m), and the n-th subcarrier of base station i be (i, n). Let... The signal to be transmitted on (i, n) represents the sub-information (k, m).

[0124]

[0125] in, Let represent the (k, m)th sub-information on the (i, n)th subcarrier, and satisfy . express The transmit power; the channel information of user k on the (i, n)th subcarrier is defined as... The received signal of base station i on carrier n is represented as:

[0126]

[0127] in, Indicates receive beamforming, z i，n Indicates power as Gaussian white noise.

[0128] Next, base station i uses continuous interference cancellation to decode all sub-information in the received signal. The entire multi-cell system contains a total of L = K × M sub-information, where L is the number of sub-information on the (i, n)th subcarrier. i，n satisfy definition For all possible decoding orders of all sub-information, and optimize the decoding order variables. Where ψ I，N This represents the decoding order of the sub-information on the (i, n)th carrier, i.e. And the decoding order of the m-th sub-information of user k on the (i, n)-th carrier is as follows

[0129] The base station decodes and removes code with higher priority according to the priority decoding order. The sub-message is treated as interference, while the remaining sub-messages are treated as interference; therefore, the sub-messages... The signal-to-interference-plus-noise ratio is expressed as:

[0130]

[0131] Among them, h j，i，n This represents the channel information of user j on the (i, n)th carrier. This represents the transmit power of the (j, l)th sub-information on the (i, n)th subcarrier; This indicates that the decoding order on subcarrier (i, n) is later than... The set of remaining sub-messages, Let represent the intra-cell interference power suffered by sub-information (k, m) on the (i, n)th subcarrier, and

[0132]

[0133] Therefore, based on formula (48), the spectral efficiency of user k can be written as:

[0134]

[0135] in, Indicates sub-message The signal-to-interference-to-noise ratio.

[0136] S2. Optimization problem and constraints of a multi-cell uplink communication system; considering that each user can only select one base station for access. The following formula must be satisfied:

[0137]

[0138] Where sgn(·) represents the sign function.

[0139] To reduce the decoding complexity of the receiver, each sub-information can only occupy a single subcarrier. This constraint is expressed as:

[0140]

[0141] Assume the maximum transmit power of each base station is P. max The non-negativity constraint for power allocation is expressed as:

[0142]

[0143] The receive beamforming at the base station is represented by a unit constant modulus constraint:

[0144]

[0145] Minimum transmission rate per user r k Must meet:

[0146]

[0147] in, Denotes the minimum transmission rate threshold for the k-th user; define d k，m Let be a predefined nonnegative value used to determine the splitting ratio of the m-th sub-message, and with the following constraints:

[0148]

[0149] The constraint on the selection of decoding order is expressed as:

[0150]

[0151] By jointly optimizing base station selection, decoding sequence, carrier allocation, and power control to maximize system performance and rate, the objective function is expressed as follows:

[0152] S3. Provide a global method for partitioning the user set and determine which base station each user uploads to; first, define an auxiliary variable y = [y 1，1 y 1，2 , ..., y K，I ]∈{0,1} KI , where y k，i =1 indicates that the k-th user has accessed the i-th base station. The constraint shown in equation (51) is rewritten as:

[0153]

[0154] Considering signaling overhead and computational complexity, this embodiment designs a distributed, low-complexity user association method. Specifically, it proposes a matching game scheme called "instantaneous acceptance," which takes priority lists of base stations and user terminals as input data and outputs matching relationships. In the initial stage, the association objects of all user terminals are set to the nearest base station.

[0155] make This represents the average channel norm of the k-th user terminal across all subcarriers at the i-th base station. In each interaction of the game, the central unit comprehensively considers... The base station is selected for the user based on the number of users in the current cell; this is because, under unit transmit power conditions, high channel gain means that the user terminal can obtain a higher average signal-to-interference-plus-noise ratio and lower interference from other cells.

[0156] Furthermore, a smaller number of active user terminals in each cell helps alleviate competition for wireless spectrum resources, especially for user terminals located at the cell edge. Therefore, the selection of base stations for user terminals to be accessed is designed based on the following criteria:

[0157]

[0158] in, Let represent the set of users within base station i, and obtain the user cooperation vector y according to the above formula.

[0159] S4. Design a distributed solution for wireless resource allocation; given a base station y, each base station is responsible for its own resource allocation design in a distributed manner; firstly, introduce continuous slack variables. The constraints shown in equation (52) are rewritten as follows:

[0160]

[0161] When receiving beamforming W i Given the decoding order, the joint optimization problem of power control and carrier allocation for the i-th base station is expressed as:

[0162]

[0163] st(53), (55), (57), (61)

[0164] The objective function for the above problem can be further expressed as:

[0165]

[0166] in, This indicates that the decoding order on carrier (i, n) is not less than The sub-information set, i.e.

[0167] Considering that the proportion coefficients of each sub-message within the same user terminal are predetermined, higher data rates can be allocated to sub-messages with higher priority; this can be achieved by utilizing priority indexes and introducing slack variables. The problem shown in equation (62) can be equivalently rewritten as:

[0168]

[0169] Solving the above problem mainly involves optimizing the coupling between variables and The non-concaveness of .

[0170] For linear products Using the log-exponential reconstruction method to and Rewritten as and Then Further rewritten as:

[0171]

[0172] because and about and It is a joint convex function, and the continuous convex approximation method is used iteratively to approximate it. By approximation, we have:

[0173]

[0174] Head Represented as:

[0175]

[0176] in, express about and The constant gradient.

[0177] Therefore, by combining equations (65) to (67) to replace The problem shown in equation (64) can be approximated as a convex lower bound maximization problem. Its optimal solution is unique and can be updated iteratively until the lower bound no longer tightens. It is worth noting that although this problem can be solved by the standard convex optimization solver, the polynomial complexity brought by the interior point method is still relatively high. Therefore, we consider using the Lagrange dual decomposition theory to solve the problem shown in equation (64).

[0178] The corresponding Lagrange function is expressed as:

[0179]

[0180] Where, μ i , λ i , ρ i v i Let Lagrange multipliers be represented; the objective function of the dual problem is expressed as:

[0181]

[0182] To avoid the duality problem tending to infinity, v i satisfy Therefore, the dual problem can be simplified to:

[0183]

[0184] in, Based on the Caro-Kun-Tucker conditions, an iterative algorithm based on gradient descent is designed to solve the dual problem.

[0185] First, calculate the Lagrange function. The gradient is:

[0186]

[0187] in, Therefore, information The iterative process of the transmit power is expressed as:

[0188]

[0189] Where (e) and ξ represent the iteration index and step size, respectively.

[0190] Then, the carrier allocation is represented as:

[0191]

[0192] Furthermore, the (k, m)th piece of information was assigned the maximum The corresponding carrier (i, n), that is:

[0193] The minimization of the dual problem for each base station, as shown in equation (70), is solved using the subgradient method.

[0194] The Lagrange dual variable is updated according to the following formula:

[0195]

[0196] Where (u) is the iteration index, Indicates positive step size; projection operator In the above formula, it is defined as a feasible lower bound.

[0197] Next, we design the receiving beamforming method, first defining... and Will Rewritten as:

[0198]

[0199] This leads to the following optimization problem:

[0200]

[0201] By utilizing the Lagrange duality transformation, the objective function of the above equation can be rewritten as:

[0202]

[0203] Introducing auxiliary variables definition

[0204] By reintroducing the auxiliary variable and employing a quadratic transformation, the above equation can be rewritten as:

[0205]

[0206] in, express The conjugate of, and The optimal solution is obtained by taking the first-order partial derivative of the above equation, that is:

[0207]

[0208] Substituting equation (83) into the problem shown in equation (80), regarding W... iThe problem is represented in the following form:

[0209]

[0210] in, The above problem is a quadratic programming problem, and the constant modulus constraint increases the difficulty of solving the problem. We consider using the alternating direction multiplier method to solve the above problem.

[0211] First, for w i，n Introducing auxiliary variable q i，n Rewriting the problem shown in equation (84) in the form of an augmented Lagrange function, we have:

[0212]

[0213] in, Indicates the possible values ​​of beamforming variables; and These represent the real and imaginary parts of a Lagrange multiplier, respectively. express Indicator functions.

[0214] The alternating direction multiplier method decomposes a complex problem into multiple smaller local problems, with the update of each variable represented as follows:

[0215]

[0216] in, t represents the iteration index, and κ is the step size.

[0217] because The convexity will The update solution is as follows:

[0218]

[0219] Then, by removing irrelevant beam vectors, equation (87) is derived as follows:

[0220]

[0221] in, Indicates (·) in The projection on the surface.

[0222] Finally, a decoding order optimization criterion for multiple sub-messages on the same subcarrier is proposed. Based on the joint optimization of the channel gain and rate segmentation ratio of the user terminal, the mathematical representation of the optimal decoding order is achieved through the following criterion: For all sub-messages on the (i, n)th subcarrier, the approximate optimal decoding order is as follows: Sort in descending order.

[0223] S5. Construct a multi-cell distributed cooperation framework based on uplink RSMA and propose a multi-base station distributed cooperation scheme based on uplink rate segmentation. Its core calculation process can be executed locally on each base station side. Specifically, the scheme includes two key parts: a) parallel user association based on "instantaneous acceptance" matching game; b) joint optimization of resource allocation and beamforming alternately implemented in distributed base stations (using the decoding order criterion described above). The update strategies of all base stations will achieve distributed cooperation through continuous interaction until the convergence condition or the maximum number of iterations is reached.

[0224] In this example, the algorithm performance is verified through simulation experiments. It is assumed that there are K users served by M base stations, and the path loss varies with distance as 120.9 + 37.6lgd, where d is the distance from the user to the base station in kilometers. The shadow fading is represented by a log-normal random variable with a mean of zero and a variance of 8dB. The small-scale fading coefficient follows a Rayleigh distribution with unit variance. Each base station is located at the center of a cell with a radius of 300 meters, and the users are uniformly distributed within the cell. The experimental parameters are shown in Table 1 below.

[0225] Table 1

[0226]

[0227] Experiment (A): As the transmit power increases, compare the changes in system and rate between the method of this embodiment and the baseline method.

[0228] Experiment (B): Comparing the computational complexity of the baseline method under different parameters, increasing the total number of users from 7 to 25.

[0229] The simulation results of this example are as follows:

[0230] 1. Comparison of the method in this embodiment with the baseline method in terms of transmission rate at different transmit powers:

[0231] From the appendix Figure 3 As can be seen, the throughput of all schemes exhibits a strictly monotonically increasing relationship with the maximum transmit power. Furthermore, the distributed optimization framework proposed in this embodiment, combined with a multi-carrier, multi-layer sub-information access mechanism, demonstrates significantly better throughput performance than existing benchmark schemes.

[0232] Analysis shows that the traditional twin-message RSMA scheme can essentially be regarded as a special case of the CDMM-RSMA architecture in this embodiment. The scheme in this embodiment realizes dynamic power allocation between different sub-messages through a low-complexity power allocation algorithm, thereby expanding the maximum achievable rate domain. Experimental results show that the cooperative radio resource allocation mechanism proposed in this embodiment has two core advantages: first, it can achieve efficient utilization of system resources; second, compared with the traditional independent single-cell multi-access scheme, it shows a significant performance improvement in system flexibility.

[0233] 2. Comparison of computational complexity between the method in this embodiment and the baseline method under different numbers of users:

[0234] From the appendix Figure 4 As can be seen, the computational complexity of the CDMM-RSMA algorithm designed in this embodiment is significantly lower than that of the Riemannian Gradient-Based RSMA (RG-RSMA) scheme and the Meta-Heuristic Non-Orthogonal Multiple Access (MH-NOMA) scheme, especially in large-scale access scenarios; for example, when the number of UTs K=25, the computational complexity can be reduced by up to about 100 times. This is due to the fact that the algorithm proposed in this embodiment makes full use of the (semi-)closed update mechanism, avoiding the large number of iterative operations required by the interior point method or genetic algorithm.

[0235] This embodiment of the method achieves joint optimization of user association, power allocation, subcarrier scheduling, decoding order, and receive beamforming in a multi-cell, multi-carrier system by dividing the message stream of each user terminal into multiple sub-messages and allocating differentiated rates. This maximizes system performance and rate, not only constructing a novel uplink multi-dimensional resource allocation theoretical framework, but also verifying the superiority of RSMA technology in complex interference environments through multi-domain collaborative optimization of space-frequency-power-decoding, thereby effectively improving the throughput performance of cellular systems and having a significant advantage in reducing computational complexity.

[0236] In addition to the embodiments described above, the present invention may have other implementations. All technical solutions formed by equivalent substitution or equivalent transformation fall within the protection scope claimed by the present invention.

Claims

1. A multi-base station cooperative radio resource allocation method based on uplink rate segmentation, characterized in that: Includes the following steps: S1. Establish the uplink RSMA transmission model; S2. Establish the optimization problem and constraints of a multi-cell uplink communication system; S3. Perform a global partitioning of the user set and determine which base station each user uploads to; S4. Solve the wireless resource allocation scheme in a distributed manner; S5. Construct a multi-cell distributed cooperation framework based on uplink RSMA; In step S1, the uplink RSMA transmission model includes I multi-antenna base stations and K cellular users, and is equipped with There are N receiving antennas, each single-antenna user communicates with a specific base station of the receiving antenna, and the entire spectrum is divided into N orthogonal subcarriers; In uplink RSMA, each user terminal splits its message into M parts and transmits them simultaneously to the access base station; Subsequently, the base station employs serial interference cancellation technology to decode all users' sub-messages according to a predetermined decoding order; the sets of users, carriers, sub-messages, base stations, and antennas are defined sequentially as follows: , , , and ; Let the m-th sub-information of user k be (k,m), and the n-th subcarrier of base station i be (i,n). Let... The signal to be transmitted on (i,n) represents the sub-information (k,m). （1）； in, Let represent the (k,m)th sub-information on the (i,n)th subcarrier, and satisfy . ; express The transmission power; Define the channel information of user k on the (i,n)th subcarrier as follows: Then the received signal of base station i on carrier n is expressed as: （2）； in, Indicates receiving beamforming, Indicates power as Gaussian white noise; base station Continuous interference cancellation is used to eliminate all sub-information in the decoded received signal, and the entire multi-cell system coexists. The number of sub-information items, where the number of sub-information items exists on the (i,n)th subcarrier. satisfy ; definition For all possible decoding orders of all sub-information, and optimize the decoding order variables. ,in This represents the decoding order of the sub-information on the (i,n)th carrier, i.e. And the decoding order of the m-th sub-information of user k on the (i,n)-th carrier is The base station decodes and removes code in order of priority, prioritizing those with higher priority. The sub-message is treated as interference, while the remaining sub-messages are treated as interference; therefore, the sub-messages... The signal-to-interference-plus-noise ratio is expressed as: （3）； in, This represents the channel information of user j on the (i,n)th carrier. This represents the transmit power of the (j,l)th sub-information on the (i,n)th subcarrier; This indicates that the decoding order on subcarrier (i,n) is later than... The set of remaining sub-messages, Let represent the intra-cell interference power suffered by sub-information (k,m) on the (i,n)th subcarrier, and （4）； Therefore, based on formula (3), the spectral efficiency of user k can be written as: （5）； in, Indicates sub-message Signal-to-interference-plus-noise ratio; In step S2, it is set that each user can only select one base station to access, and the transmission power is... The following formula must be satisfied: （6）； Where sgn(.) represents the symbol function; each sub-information can only occupy a single subcarrier, and this constraint is expressed as: （7）； Assume the maximum transmit power of each base station is The non-negativity constraint for power allocation is expressed as: （8）； The receive beamforming at the base station is represented by a unit constant modulus constraint: （9）； Minimum transmission rate per user Must meet: （10）； in, Represents the minimum transmission rate threshold for the k-th user; Defined Let be a predefined nonnegative value used to determine the splitting ratio of the m-th sub-message, and with the following constraints: （11）； （12）； The selection constraint for the decoding order is expressed as: （13）； By jointly optimizing base station selection, decoding sequence, carrier allocation, and power control to maximize system performance and rate, the objective function is expressed as follows: ; In step S3, auxiliary variables are selected and defined for the base station. ,in This indicates that the k-th user has accessed the i-th base station. The constraint shown in equation (6) is rewritten as follows: （14）； A matching game is performed, using a priority list of base stations and user terminals as input data, and outputting matching relationships; in the initial stage, the associated objects of all user terminals are set to the nearest base station; make This represents the average channel norm of the k-th user terminal across all subcarriers at the i-th base station. In each interaction of the game, the central unit comprehensively considers... The selection of base stations for users is based on the number of users in the current cell; the selection of base stations for user terminals to be accessed is designed based on the following criteria: （15）； in, Let i represent the set of users within base station i. The user cooperation vector is obtained according to the above formula. .

2. The multi-base station cooperative radio resource allocation method based on uplink rate segmentation according to claim 1, characterized in that: In step S4, when a given base station is selected... In a distributed manner, each base station is responsible for its own resource allocation design; firstly, continuous slack variables are introduced. The constraints shown in equation (7) are rewritten as follows: （16）； When receiving beamforming Given the decoding order, the joint optimization problem of power control and carrier allocation for the i-th base station is expressed as: （17）； The objective function for the above problem can be further expressed as: （18）； in, , , This indicates that the decoding order on carrier (i,n) is not less than The sub-information set, i.e. ; Allocate higher data rates to higher-priority sub-messages; this is achieved by utilizing priority indexes and introducing slack variables. The problem shown in equation (17) can be equivalently rewritten as: （19）； Solving the above problem lies in optimizing the coupling between variables and Non-concavity; for linear products The log-exponential reconstruction method is used to reconstruct the data. and Rewritten as and Then Further rewritten as: （20）； because and about and It is a joint convex function, and the continuous convex approximation method is used iteratively to approximate it. By approximation, we have: （21）； and Represented as: （22）； in, express about and The constant gradient; By combining formulas (20) to (22) to replace The problem shown in equation (19) is approximated as a convex lower bound maximization problem. Its optimal solution is unique. The problem is solved by iteratively updating until the lower bound no longer tightens, using the Lagrange dual decomposition theory.

3. The multi-base station cooperative radio resource allocation method based on uplink rate segmentation according to claim 2, characterized in that: In step S4, the problem shown in equation (19) is solved using the Lagrange dual decomposition theory, and the corresponding Lagrange function is expressed as: （23）； in, Let the Lagrange multipliers be represented; the objective function of the dual problem is expressed as: （24）； in, satisfy The dual problem is simplified to: （25）； in, Based on the Caro-Kun-Tucker conditions, the dual problem is solved using an iterative algorithm based on gradient descent.

4. The multi-base station cooperative radio resource allocation method based on uplink rate segmentation according to claim 3, characterized in that: In step S4, the dual problem is solved using an iterative algorithm based on gradient descent, according to the Caro-Kuhn-Tucker condition. First, the Lagrangian function is calculated. The gradient is: （26）； （27）； in, ; information The iterative process of the transmit power is expressed as: （28）； Among them, (e) and These represent the iteration index and step size, respectively. Then, the carrier allocation is represented as: （29）； The (k,m)th piece of information was assigned the maximum The corresponding carrier (i,n) is: （30）； The minimization of the dual problem for each base station, as shown in equation (25), is solved using the subgradient method; the Lagrange dual variables are updated according to the following formula: （31）； （32）； （33）； Where (u) is the iteration index, Indicates positive step size; projection operator In the above formula, it is defined as a feasible lower bound.

5. The multi-base station cooperative radio resource allocation method based on uplink rate segmentation according to claim 4, characterized in that: In step S4, the receiving beamforming method is designed, first defining... , and ,Will Rewritten as: （34）； This leads to the following optimization problem: （35）； By utilizing the Lagrange duality transformation, the objective function of the above equation can be rewritten as: （36）； Introducing auxiliary variables ,definition , ; By reintroducing the auxiliary variable and employing a quadratic transformation, the above equation can be rewritten as: （37）； in, express The conjugate of, and The optimal solution is obtained by taking the first-order partial derivative of the above equation, that is: （38）； Substituting equation (38) into the problem shown in equation (35), regarding The problem is represented in the following form: （39）； in, , The problem above can be solved using the alternating direction multiplier method.

6. The multi-base station cooperative radio resource allocation method based on uplink rate segmentation according to claim 5, characterized in that: In step S4, the problem shown in equation (39) is solved using the alternating direction multiplier method. First, for Introducing auxiliary variables Rewriting the problem shown in equation (39) in the form of an augmented Lagrangian function, we have: （40）； in, Indicates the possible values ​​of beamforming variables; and These represent the real and imaginary parts of a Lagrange multiplier, respectively. express Indicator functions; The problem is decomposed into multiple local problems using the alternating direction multiplier method, and the update of each variable is represented as follows: （41）； （42）； （43）； in, t represents the iteration index. The step size; because The convexity will The update solution is as follows: （44）； Then, by removing irrelevant beam vectors, equation (42) is derived as follows: （45）； in, Indicates (.) in Projection on; Finally, based on the joint optimization of channel gain and rate split ratio of the user terminal, the optimal decoding order is mathematically represented by a preset decoding order optimization criterion. The decoding order optimization criterion is as follows: for the first... The approximate optimal decoding order of all sub-messages on each subcarrier is as follows: Sort in descending order.

Images