Resource allocation method and system for c-rsma assisted digital twin network joint device pairing

Abstract

The application discloses a resource allocation method and system for C-RSMA auxiliary digital twin network joint device pairing, and belongs to the technical field of wireless communication and digital twin fusion. In view of the problems that the channel difference of a digital twin network is large, data demand is unbalanced, and data redundancy and high system time delay are caused by the fact that the prior art does not combine modeling accuracy constraints, a C-RSMA digital twin uplink transmission system model is first constructed, near-end-far-end device pairing is completed through a two-stage strategy, the minimum transmission data amount of the device is determined in combination with the modeling accuracy constraint, and then the joint optimization of power allocation, time allocation factor and computing resource is completed through an alternating optimization framework, so that the overall generation time delay of the digital twin system is minimized. The application can significantly improve the transmission reliability of edge devices and the system resource utilization rate, and is suitable for a large-scale multi-device access digital twin commercial scene.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication and digital twin convergence technology, specifically involving the cross-application of cooperative rate split multiple access technology and industrial digital twin networks in 6G scenarios, and particularly involving a resource allocation method and system for C-RSMA-assisted digital twin network joint device pairing. Background Technology

[0002] With the deep integration of 6G technology with vertical industries such as the Industrial Internet, intelligent transportation, and smart grids, digital twins have become a core supporting technology for realizing the digital mapping and intelligent management of all elements of physical entities. In a digital twin network, physical terminal devices need to continuously upload real-time status data to the digital twin server on the base station side to support the high-precision construction and dynamic updating of the twin. Therefore, the real-time performance and reliability of uplink data transmission directly determine the core performance boundaries of the digital twin system.

[0003] In commercial scenarios of digital twins with large-scale access of multiple devices, terminal devices naturally have two core characteristics: First, there are significant differences in channel conditions. Near-end devices close to the base station have good channel quality, while far-end devices located at the cell edge suffer from severe channel fading and have limited transmission performance. Second, the demand for twin modeling data is uneven. Different types of physical entities have different requirements for the accuracy of twin modeling, resulting in significant differences in the amount of state data uploaded.

[0004] Existing multiple access and resource allocation schemes are ill-suited to the core characteristics of the aforementioned digital twin scenarios: traditional TDMA and OFDMA schemes have limited spectral efficiency and severely inadequate transmission performance of edge devices, failing to meet the low latency requirements for real-time twin updates; Cooperative Non-Orthogonal Multiple Access (C-NOMA) imposes strict constraints on user channel gain ranking, resulting in insufficient robustness to interference management and significant performance degradation in multi-user scenarios; while Cooperative Rate Split Multiple Access (C-RSMA) offers flexible interference management capabilities and potential for edge device performance improvement, existing solutions are designed solely for traditional communication systems, focusing on traditional communication metrics such as throughput and spectral efficiency, completely neglecting the modeling accuracy constraints unique to digital twin systems. This makes it impossible to adaptively control data transmission volume according to twin service requirements, easily leading to redundant data transmission and additional system latency; furthermore, most existing solutions only optimize communication resources in a single dimension, failing to achieve global joint optimization of device pairing, power, time, and computing resources, thus failing to achieve optimal synergy between "modeling accuracy, system latency, and resource utilization," making it difficult to support the large-scale commercial deployment of digital twin technology. Summary of the Invention

[0005] To address the aforementioned issues, this invention aims to propose a resource allocation method and system for C-RSMA-assisted digital twin network joint device pairing. This method deeply integrates the advantages of C-RSMA's cooperative transmission and flexible interference management with the modeling accuracy constraints of digital twins. It resolves the core pain points in existing technologies, such as "disconnect between communication performance optimization and twin service requirements, insufficient multi-dimensional resource collaboration, and limited transmission performance of edge devices." While strictly ensuring the minimum accuracy requirements of twin modeling, it minimizes the overall generation latency of the digital twin system, while significantly improving system resource utilization and edge device transmission reliability.

[0006] To achieve the above objectives, the technical solution of the present invention is implemented as follows: A resource allocation method for joint device pairing in a C-RSMA-assisted digital twin network is applied to a digital twin uplink transmission system. The system includes a base station, multiple physical terminal devices, and a digital twin server. The total system bandwidth is divided into... Each subcarrier can be assigned a pair of devices consisting of a near-end device and a far-end device, comprising the following steps: S1. Determine the target application system: Clarify the hardware composition, subcarrier allocation rules, and device access rules of the digital twin uplink transmission system, and identify the optimization target of this method; S2. For the above target application system, construct a digital twin network uplink transmission system model based on Cooperative Rate Split Multiple Access (C-RSMA), define the terminal device set, channel model, transmission delay and computation delay model, and clarify the system optimization objectives and full-process constraints to provide quantitative rules and optimization boundaries for subsequent steps. S3. Based on the channel model constructed in S2 and the native data requirements of terminal devices, a two-stage device pairing strategy is implemented: first, the nearest device with the optimal channel gain is matched for each subcarrier; then, the remote device is assigned to each subcarrier in combination with the native data requirements of the remote device and the channel gain of inter-device cooperation, thus completing the one-to-one binding of subcarriers and devices and generating a global device pairing set. S4. Based on the device pairing set generated by S3, and combined with the digital twin modeling accuracy constraints defined by S2, the minimum amount of data to be transmitted for each terminal device to meet the minimum modeling requirements is calculated through the data accuracy model, and the business boundary for optimizing transmission and computing resources is determined. S5. Based on the device pairing results of S3 and the minimum data transmission volume of S4, initialize the initial values ​​of three types of optimization variables: power allocation, time allocation factor, and computing resource allocation. At the same time, set the iterative convergence judgment conditions to complete the preparatory work for alternating optimization. S6. Using the initial values ​​of S5 or the time allocation factor and computational resource allocation results from the previous iteration, based on the system model of S2, the device pairing results of S3, and the minimum data transmission volume of S4, the non-convex power allocation problem is transformed into a convex one by using a stepwise convex approximation method. The optimal power allocation result for this round is obtained by solving the problem through convex optimization. S7, the optimal power allocation result of the current round output by fixed S6, the initial value of S5 initialization, or the computational resource allocation result of the previous iteration output, based on the system model of S2, the device pairing result of S3 and the minimum transmission data volume of S4, the optimal time allocation factor of the current round is obtained by using the golden section search method. S8, the current round's optimal power allocation result output by fixed S6, the current round's optimal time allocation factor output by S7, the system model of S2, the device pairing result of S3, and the minimum data transmission amount of S4, transform the computational resource allocation problem into a convex optimization problem and solve it to obtain the current round's optimal computational resource allocation result; S9. Update the parameters of the current round of optimization results output by S6-S8, and repeat the alternating optimization steps from S6 to S8 until the system delay change between two adjacent iterations meets the convergence judgment condition set by S5. The iteration terminates and outputs the globally optimal resource allocation scheme that adapts to the target application system of S1.

[0007] Furthermore, in step S2, the system optimization objective is to minimize the overall generation delay of the digital twin system, and the objective function expression is:

[0008] in, For the device pairing set, the first The device pairs corresponding to each subcarrier are represented as follows: , and These represent the far-end and near-end devices allocated to subcarrier k, respectively. For equipment The actual amount of data transmitted. Assigning variables to power It is the transmission power of the remote device. It is the power of the near-end device sending its own split sub-messages. It is the power of the near-end device forwarding information from the far-end device. As a time allocation factor, For the computing resources allocated to the device, It is a subcarrier The set, and These represent the actual amount of data transmitted by the remote and near-end devices, respectively. and These represent the rates of the remote and near-end devices, respectively. and These represent the computation latency of the remote device and the local device, respectively.

[0009] Furthermore, in step S2, the overall process constraints include: , among which, among which For the complete collection of terminal devices, For equipment The amount of native state data generated; ,in For the transmission power of the remote device, This is the maximum transmit power of the remote device; ,in , The transmit power for the near-end device to send its own split sub-messages. The transmission power for near-end devices to relay data from far-end devices. This refers to the maximum transmit power of the near-end device;

[0010] ,in For the first The time allocation factor of each subcarrier represents the time proportion of the two-stage transmission of C-RSMA. , For the first Each subcarrier corresponds to a device pair, ensuring that only one pair of devices is allocated to each subcarrier; Ensure that all terminal devices complete subcarrier allocation; ,in For equipment The accuracy of digital twin modeling, The minimum modeling accuracy threshold preset for the system; ,in To be assigned to the device Computing resources This represents the maximum total computing resources available for the digital twin server.

[0011] Furthermore, in step S2: the C-RSMA uplink transmission is divided into two stages: the first stage time proportion is... Remote devices send status data The signal is simultaneously received by the base station and the paired near-end device; the second phase time proportion is The near-end device splits its own data into two sub-messages. and , and the number of remote devices after decoding According to the joint encoding, the base station completes data decoding through serial interference cancellation; the uplink transmission rate expression between the near-end device and the far-end device is:

[0012]

[0013] in, and These are the uplink transmission rates for the near-end device and the far-end device, respectively. and It is a sub-message of the base station decoding the near-end device. and rate, It is the near-end device decoding the sub-message of the remote device. rate, The base station directly decodes the sub-messages of the remote device. rate, It is a remote device sub-message The rate at which the base station decodes the data after it has been forwarded by the near-end equipment; the expression is as follows:

[0014]

[0015]

[0016]

[0017]

[0018] in, The bandwidth of a single subcarrier; The signal-to-noise ratio of the link from the remote user to the base station; The signal-to-noise ratio of the link from the remote user to the near user; Decode the first part of the sub-message near the user for the base station Signal-to-noise ratio; For base station decoding of near-end user sub-messages and forwarding of far-end user sub-messages Signal-to-noise ratio; Decode the second part of the sub-message near the user for the base station The signal-to-noise ratio, of which and For remote users and nearby users To the base station on subcarrier Channel gain; For remote users To nearby users In subcarrier The channel gain is above, and all of the above channels follow Rayleigh fading. Indicates the noise variance; The expression for the time delay calculation model is:

[0019] in, The computational resource requirement coefficient per bit of data.

[0020] Furthermore, the two-stage device pairing strategy in step S3 specifically includes: S31. Initialize the set of unallocated devices. ,make , For the complete collection of terminal devices; S32, Channel gain between all unallocated devices and subcarriers Construct the channel gain matrix and remove the smallest value. Each channel gain value This refers to the total number of subcarriers and the corresponding combination of devices and subcarriers. Records are unassignable sets If the combination of device and subcarrier Belonging to the set If so, the device will not participate in the allocation of that subcarrier; S33. Sort all subcarriers in descending order according to the number of channel gains deleted in each subcarrier, and select the near-end device for each subcarrier in turn; for the current subcarrier In the unassigned equipment set In the process, the device that satisfies the undeleted constraint and has the maximum channel gain is selected as the near-end device. : And update the set of unassigned devices: Repeat the above process until all subcarriers have been allocated to near-end devices; S34. Select the remaining unassigned devices as a candidate remote device set, and determine their data requirements accordingly. Sort in descending order; S35. Initialize the set of subcarriers for unassigned remote devices. For each remote device after sorting From the set Select to increase cooperative link gain The largest subcarrier is allocated, that is:

[0021] It is a remote device Select the optimal subcarrier and put the device As the remote device of this subcarrier Simultaneously update the subcarrier set ; S36. After all devices are allocated, the device pairing set corresponding to each subcarrier is obtained. .

[0022] Furthermore, in step S4, the data accuracy model is: Where 'a' is the accuracy control parameter; combined with minimum modeling accuracy constraints The minimum data transmission volume for each terminal device is derived as follows: .

[0023] Furthermore, the power allocation optimization in step S6 specifically includes: S61. Fixed time allocation factor and computational resource allocation results, introducing auxiliary variables. With target variable By performing equivalent substitutions on the power variable, new optimization variables can be constructed. The original minimization of maximum latency problem is transformed into... To optimize the equivalence problem of the objective; S62. Transform the non-convex rate constraint expression into a differential convex DC structure, perform a first-order Taylor expansion to linearize the concave function terms, and construct an approximate function for the lower bound of the rate. S63. Substitute the approximate rate expression into the constraint conditions to complete the transformation from a non-convex problem to a convex optimization problem. Solve the problem using a convex optimization solver to obtain the optimal power allocation result for this round.

[0024] Furthermore, the time allocation factor optimization in step S7 specifically includes: S71. The results of fixed power allocation and computational resource allocation decompose the global optimization problem into... A factor regarding the time allocation of a single subcarrier Independent subproblems; S72. For each univariate subproblem, the golden section search method is adopted. Test points are selected in the interval [0,1] according to the golden ratio, and the corresponding objective function value is calculated. The search range is narrowed through interval shrinking iteration until the interval length is less than the preset accuracy threshold. The optimal time allocation factor of the subcarrier is then output. .

[0025] Furthermore, the computational resource allocation optimization in step S8 specifically includes: With fixed power allocation results and time allocation factors, auxiliary variables are introduced. This represents the maximum generation delay of the system. The original computational resource allocation problem is transformed into a convex optimization problem, which is solved by a convex optimization solver to obtain the optimal computational resource allocation result for this round.

[0026] To achieve the above objectives, the present invention also provides a resource allocation system for C-RSMA-assisted digital twin network joint device pairing, for performing the above-described method, comprising: The system confirmation module is used to identify the hardware composition, subcarrier allocation rules, and device access rules of the target digital twin uplink transmission system, and to clarify the optimization targets. The model building module is used to build a C-RSMA-based digital twin network uplink transmission system model, define the terminal device set, channel model, transmission delay and computation delay model, and clarify the system optimization objectives and full-process constraints. The device pairing module is used to execute a two-stage device pairing strategy, complete the one-to-one binding of subcarriers and device pairs, and generate a global device pairing set. The data volume optimization module is used to calculate the minimum amount of data to be transmitted for each terminal device to meet the minimum modeling requirements by combining modeling accuracy constraints and data accuracy models. The initialization module is used to initialize the initial values ​​of optimization variables for power allocation, time allocation factor, and computing resource allocation, and to set the conditions for iterative convergence. The alternating optimization module includes a power optimization unit, a time optimization unit, and a computing resource optimization unit, which are used to sequentially optimize power allocation, time allocation factor, and computing resources in stages. The iterative convergence module updates the parameters of the output of the alternating optimization module, repeatedly triggers the alternating optimization process until the convergence judgment condition is met, and outputs the globally optimal resource allocation scheme that adapts to the target system.

[0027] Beneficial effects: (1) Deep integration of business and communication, eliminating data redundancy from the root: This invention is the first to deeply bind C-RSMA technology with the accuracy constraints of digital twin modeling. By adaptively determining the minimum amount of data transmitted by the device through the data accuracy model, under the premise of strictly ensuring the minimum accuracy requirements of twin modeling, it completely eliminates invalid redundant data transmission, greatly reduces the system transmission and calculation delay, and solves the technical defect of "communication optimization and twin business disconnection" in the existing technology.

[0028] (2) Adapting to the characteristics of the scenario and greatly improving the performance of edge devices: The present invention designs a two-stage device pairing strategy that takes into account channel gain, cooperative link quality and device data requirements, to achieve optimal matching between near-end and far-end devices, perfectly adapting to the characteristics of large channel differences and unbalanced data requirements of multiple devices in the digital twin scenario, significantly improving the transmission reliability and rate of far-end devices at the cell edge, and solving the problem of limited edge device performance in the existing solution.

[0029] (3) Multi-dimensional resource joint optimization to achieve optimal global performance: This invention constructs a multi-dimensional joint optimization framework of "device pairing - power allocation - time allocation - computing resource allocation", and designs a special solution algorithm for the characteristics of different optimization variables. It approaches the global optimal solution through alternating iteration, which greatly improves the system resource utilization efficiency compared with the existing single-dimensional resource optimization scheme.

[0030] (4) The algorithm has good convergence and is suitable for commercial application scenarios: The solution algorithm proposed in this invention has good convergence and can converge to a stable solution within a finite number of iterations, which is suitable for commercial scenarios of digital twins with large-scale access of multiple devices. Simulation results show that the system generation delay of this invention is significantly better than that of TDMA, C-NOMA, and traditional unoptimized C-RSMA under different transmission power, number of devices, modeling accuracy and computing resources. In particular, the performance advantage is more prominent in scenarios with low transmission power, high accuracy requirements and large-scale device access. Attached Figure Description

[0031] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a flowchart illustrating the steps of the resource allocation method for C-RSMA-assisted digital twin network joint device pairing according to an embodiment of the present invention. Figure 2 This is an algorithm flowchart of the resource allocation method for C-RSMA-assisted digital twin network joint device pairing according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the digital twin uplink transmission system in the resource allocation method for C-RSMA-assisted digital twin network joint device pairing according to an embodiment of the present invention; Figure 4The graphs show the overall generation latency versus the maximum power limit of the remote user for the digital twin systems of the proposed C-RSMA, Time Division Multiple Access (Benchmark TDMA), Cooperative Non-Orthogonal Multiple Access without Time Allocation Optimization and Pairing Strategy Optimization (Benchmark C-NOMA), and Cooperative Rate Division Multiple Access without Time Allocation Optimization and Pairing Strategy Optimization (Benchmark C-RSMA). Figure 5 Generate latency versus device count graphs for the overall digital twin systems of Proposed C-RSMA, Benchmark TDMA, Benchmark C-NOMA, and Benchmark C-RSMA; Figure 6 Generate curves of latency versus minimum accuracy requirements for the overall digital twin systems of Proposed C-RSMA, Benchmark TDMA, Benchmark C-NOMA, and Benchmark C-RSMA. Figure 7 Generate a graph showing the overall latency versus total computing resources for the digital twin systems of Proposed C-RSMA, Benchmark TDMA, Benchmark C-NOMA, and Benchmark C-RSMA. Detailed Implementation

[0032] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0033] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0034] Definitions: C-RSMA stands for Cooperative Rate-Splitting Multiple Access, and its Chinese standard translation is Cooperative Rate-Splitting Multiple Access (also known as Cooperative Rate-Splitting Multiple Access). It is a new type of wireless multiple access technology that evolved from Rate-Splitting Multiple Access (RSMA) by introducing an inter-device cooperative relay mechanism.

[0035] The core logic is to enable near-end devices with good channel conditions to assist far-end devices with poor channel conditions in forwarding data. At the same time, through rate splitting coding, efficient parallel transmission of multiple users on the same subcarrier is achieved, taking into account both spectrum efficiency and transmission reliability of edge devices.

[0036] Example 1 See Figures 1-7A resource allocation method for joint device pairing in a C-RSMA-assisted digital twin network is applied to a digital twin uplink transmission system. The system includes a base station, multiple physical terminal devices, and a digital twin server. The total system bandwidth is divided into... Each subcarrier can be assigned a pair of devices consisting of a near-end device and a far-end device, comprising the following steps: S1. Determine the target application system: Clarify the hardware composition, subcarrier allocation rules, and device access rules of the digital twin uplink transmission system, and identify the optimization target of this method; S2. For the above target application system, construct a digital twin network uplink transmission system model based on Cooperative Rate Split Multiple Access (C-RSMA), define the terminal device set, channel model, transmission delay and computation delay model, and clarify the system optimization objectives and full-process constraints to provide quantitative rules and optimization boundaries for subsequent steps. S3. Based on the channel model constructed in S2 and the native data requirements of terminal devices, a two-stage device pairing strategy is implemented: first, the nearest device with the optimal channel gain is matched for each subcarrier; then, the remote device is assigned to each subcarrier in combination with the native data requirements of the remote device and the channel gain of inter-device cooperation, thus completing the one-to-one binding of subcarriers and devices and generating a global device pairing set. S4. Based on the device pairing set generated by S3, and combined with the digital twin modeling accuracy constraints defined by S2, the minimum amount of data to be transmitted for each terminal device to meet the minimum modeling requirements is calculated through the data accuracy model, and the business boundary for optimizing transmission and computing resources is determined. S5. Based on the device pairing results of S3 and the minimum data transmission volume of S4, initialize the initial values ​​of three types of optimization variables: power allocation, time allocation factor, and computing resource allocation. At the same time, set the iterative convergence judgment conditions to complete the preparatory work for alternating optimization. S6. Using the initial values ​​of S5 or the time allocation factor and computational resource allocation results from the previous iteration, based on the system model of S2, the device pairing results of S3, and the minimum data transmission volume of S4, the non-convex power allocation problem is transformed into a convex one by using a stepwise convex approximation method. The optimal power allocation result for this round is obtained by solving the problem through convex optimization. S7, the optimal power allocation result of the current round output by fixed S6, the initial value of S5 initialization, or the computational resource allocation result of the previous iteration output, based on the system model of S2, the device pairing result of S3 and the minimum transmission data volume of S4, the optimal time allocation factor of the current round is obtained by using the golden section search method. S8, the current round's optimal power allocation result output by fixed S6, the current round's optimal time allocation factor output by S7, the system model of S2, the device pairing result of S3, and the minimum data transmission amount of S4, transform the computational resource allocation problem into a convex optimization problem and solve it to obtain the current round's optimal computational resource allocation result; S9. Update the parameters of the current round of optimization results output by S6-S8, and repeat the alternating optimization steps from S6 to S8 until the system delay change between two adjacent iterations meets the convergence judgment condition set by S5. The iteration terminates and outputs the globally optimal resource allocation scheme that adapts to the target application system of S1.

[0037] It should be noted that, see Figure 3 The digital twin uplink transmission system consists of one base station (BS), U physical devices (PDs), and a corresponding digital twin network (DNT). The base station acts as the central server, responsible for data reception and digital twin model construction. A physical network (PN) is composed of physical devices. The set of devices is denoted as . Each physical device Generate state data The total system bandwidth is , divided into There are orthogonal subcarriers, each with a bandwidth of . The set of subcarriers is denoted as A pairwise allocation mechanism is adopted, where every two physical devices are paired and allocated to the same subcarrier, denoted as the first pair. The device pairs corresponding to each subcarrier are Among them, equipment The channel conditions are better than those of the equipment. ,Right now For remote devices, For near-end devices.

[0038] First, each pair of devices transmits data on its corresponding subcarrier. The transmission time is divided into two phases, with the following time proportions: and ,in Assigning time as a factor. In the first phase, remote devices... With power Send its status data The signal is simultaneously detected by the base station and near-end devices. Receive; in the second phase, near-end equipment The decoding obtained in the first stage The information is re-encoded and combined with its own data. Joint transmission is performed according to the C-RSMA mechanism. Among them, Split into two sub-messages and And they are transmitted at different powers, while The message is not split. The base station decodes the received signal, and the decoding order is as follows: .

[0039] Furthermore, to ensure the accuracy of digital twin modeling, the actual amount of data transmitted by each device is defined as follows: The corresponding modeling accuracy is expressed as ,in These are precision control parameters.

[0040] After the terminal devices complete the upload of status data, the server processes the data from each terminal device to generate the corresponding digital twin model. This process will generate computational latency.

[0041] Therefore, this embodiment introduces an inter-device cooperative transmission mechanism and rate split multiple access technology, enabling data from remote devices to be forwarded collaboratively by near-end devices. Combined with digital twin accuracy constraints, the amount of transmitted data is adaptively controlled. This improves the data transmission reliability and system resource utilization efficiency of weak channel devices, reduces redundant data transmission overhead, and effectively reduces the overall generation latency of the digital twin system in scenarios with large differences in channel conditions and uneven data demand.

[0042] In summary, this embodiment forms a complete low-latency optimization process by constructing a C-RSMA model, pairing devices in two stages, determining the minimum data volume based on accuracy, and iteratively optimizing power / time / computing resources, thus ensuring twin accuracy and minimizing the overall system generation latency.

[0043] In a specific example, in step S2, the system optimization objective is to minimize the overall generation delay of the digital twin system, and the objective function expression is:

[0044] in, For the device pairing set, the first The device pairs corresponding to each subcarrier are represented as follows: , and These represent the far-end and near-end devices allocated to subcarrier k, respectively. For equipment The actual amount of data transmitted. Assigning variables to power It is the transmission power of the remote device. It is the power at which the near-end device transmits its own information. It is the power of the near-end device forwarding information from the far-end device. As a time allocation factor, For the computing resources allocated to the device, It is a carrier wave The set, and These represent the actual amount of data transmitted by the remote and near-end devices, respectively. and These represent the rates of the remote and near-end devices, respectively. and These represent the computation latency of the remote device and the local device, respectively.

[0045] This embodiment establishes a mathematical model with "maximum transmission delay + maximum computation delay" as the objective function to achieve accurate quantification of system delay and ensure that the optimization direction is consistent and calculable.

[0046] In a specific example, the overall process constraints in step S2 include: , among which, among which For the complete collection of terminal devices, For equipment The amount of native state data generated; ,in For the transmission power of the remote device, This is the maximum transmit power of the remote device; ,in , The transmit power for the near-end device to send its own split sub-messages. The transmission power for near-end devices to relay data from far-end devices. This refers to the maximum transmit power of the near-end device;

[0047] ,in For the first The time allocation factor of each subcarrier represents the time proportion of the two-stage transmission of C-RSMA. , For the first Each subcarrier corresponds to a device pair, ensuring that only one pair of devices is allocated to each subcarrier; Ensure that all terminal devices complete subcarrier allocation; ,in For equipment The accuracy of digital twin modeling, The minimum modeling accuracy threshold preset for the system; ,in To be assigned to the device Computing resources This represents the maximum total computing resources available for the digital twin server.

[0048] It should be noted that in this embodiment... Ensure that the amount of data uploaded by the device does not exceed the amount of data it generates. At the same time, it cannot be negative; The transmission power of remote devices is limited to the maximum power. At the same time, it cannot be negative; The total power used by near-end devices for their own data and for forwarding data is limited to not exceeding their maximum power. ; Ensure that the power allocation variable for near-end devices is non-negative; The time allocation factor is between 0 and 1, representing the ratio of the two-stage transmission time. Ensure that each subcarrier can only be assigned to two devices; Ensure that all devices are assigned to a subcarrier; Ensure that the digital twin modeling accuracy of each device meets the minimum requirements. ; Ensure that the total computing resources allocated to all devices do not exceed the upper limit. .

[0049] This embodiment ensures that the optimization is legal and feasible by setting nine types of constraints, including data volume, power, time, subcarrier, accuracy, and computing resources, and meets the requirements of hardware limitations and twin accuracy.

[0050] In a specific example, in step S2: the C-RSMA uplink transmission is divided into two stages: the first stage time proportion is... Remote devices send status data The signal is simultaneously received by the base station and the paired near-end device; the second phase time proportion is The near-end device splits its own data into two sub-messages. and , and the decoded remote device data Joint encoding and transmission are performed, and the base station completes data decoding through serial interference cancellation; the uplink transmission rate expression between the near-end device and the far-end device is:

[0051] in, and These are the uplink transmission rates for the near-end device and the far-end device, respectively. and It is a sub-message of the base station decoding the near-end device. and rate, It is the near-end device decoding the sub-message of the remote device. rate, The base station directly decodes the sub-messages of the remote device. rate, It is a remote device sub-message The rate at which the base station decodes the data after it has been forwarded by the near-end equipment; the expression is as follows:

[0052]

[0053]

[0054]

[0055]

[0056] in, The bandwidth of a single subcarrier; The signal-to-noise ratio of the link from the remote user to the base station; The signal-to-noise ratio of the link from the remote user to the near user; Decode the first part of the sub-message near the user for the base station Signal-to-noise ratio; For base station decoding of near-end user sub-messages and forwarding of far-end user sub-messages Signal-to-noise ratio; Decode the second part of the sub-message near the user for the base station The signal-to-noise ratio, of which and For remote users and nearby users To the base station on subcarrier Channel gain; For remote users To nearby users In subcarrier The channel gain is above, and all of the above channels follow Rayleigh fading. This represents the noise variance.

[0057] The expression for the time delay calculation model is:

[0058] in, The computational resource requirement coefficient per bit of data.

[0059] This embodiment employs a two-stage C-RSMA cooperative transmission, defines a rate and computational delay model, improves the transmission reliability of weak devices, increases spectrum efficiency, and reduces latency for edge users.

[0060] In a specific example, the two-stage device pairing strategy in step S3 specifically includes: S31. Initialize the set of unallocated devices. ,make , For the complete collection of terminal devices; S32, Channel gain between all unallocated devices and subcarriers Construct the channel gain matrix and remove the smallest value. Each channel gain value This refers to the total number of subcarriers and the corresponding combination of devices and subcarriers. Records are unassignable sets If the combination of device and subcarrier Belonging to the set If so, the device will not participate in the allocation of that subcarrier; S33. Sort all subcarriers in descending order according to the number of channel gains deleted in each subcarrier, and select the near-end device for each subcarrier in turn; for the current subcarrier In the unassigned equipment set In the process, the device that satisfies the undeleted constraint and has the maximum channel gain is selected as the near-end device. : And update the set of unassigned devices: Repeat the above process until all subcarriers have been allocated to near-end devices; S34. Select the remaining unassigned devices as a candidate remote device set, and determine their data requirements accordingly. Sort in descending order; S35. Initialize the set of subcarriers for unassigned remote devices. For each remote device after sorting From the set Select to increase cooperative link gain The largest subcarrier is allocated, that is:

[0061] It is a remote device Select the optimal subcarrier and put the device As the remote device of this subcarrier Simultaneously update the subcarrier set ; S36. After all devices are allocated, the device pairing set corresponding to each subcarrier is obtained. .

[0062] In this embodiment, the near-end device is selected first, and then the far-end device is matched according to the cooperative link gain to achieve optimal device pairing, improve cooperative gain, and balance channel differences and data demand.

[0063] In a specific example, in step S4, the data accuracy model is: Where 'a' is the accuracy control parameter; combined with minimum modeling accuracy constraints The minimum data transmission volume for each terminal device is derived as follows: .

[0064] This embodiment derives the minimum data transmission volume of the device based on accuracy constraints, thereby reducing redundant data and lowering the transmission and computation burden while meeting the modeling accuracy requirements.

[0065] In a specific example, the power allocation optimization in step S6 includes: S61. Fixed time allocation factor and computational resource allocation results, introducing auxiliary variables. With target variable By performing equivalent substitutions on the power variable, new optimization variables can be constructed. The original minimization of maximum latency problem is transformed into... To optimize the equivalence problem of the objective; S62. Transform the non-convex rate constraint expression into a differential convex DC structure, perform a first-order Taylor expansion to linearize the concave function terms, and construct an approximate function for the lower bound of the rate. S63. Substitute the approximate rate expression into the constraint conditions to complete the transformation from a non-convex problem to a convex optimization problem. Solve the problem using a convex optimization solver to obtain the optimal power allocation result for this round.

[0066] In the specific implementation: Factors are allocated at a given time. and computing resource allocation In the case of studying power allocation With optimization, the problem now becomes:

[0067] Introducing auxiliary variables and target variable And by making equivalent substitutions to the power variables, new optimization variables are constructed. The original problem of minimizing the maximum system latency is transformed into... To optimize the equivalent problem of the objective, the constraints include: (1) (2) The original question is transformed into a question about , and Optimization issues:

[0068] To handle the nonconvexity of (2), This can be represented as a difference convex function (DC) structure, which is the form of the difference between two logarithmic functions, thus transforming the original non-convex rate function into a concave function minus a concave function form: (3) in, , , , , ; The DC structure is processed using a successive convex approximation method. At the current iteration point, the concave functions within it are approximated. Perform a first-order Taylor expansion and linearization. This represents the current iteration number. (4) Substituting (4) into (3) will construct its lower bound approximate function. Approximate expressions for the rates of remote and near users are obtained, respectively. (5) (6) S55, Substituting the approximate rate expression into the original constraint conditions, the original constraint (2) becomes: (7) S56, using the above approximation, the subproblem can be transformed into:

[0069] As can be seen, the subproblem is a convex problem, which can be solved using any convex optimization solver (such as CVX).

[0070] This embodiment transforms the non-convex power problem into a convex optimization problem by using the successive convex approximation (SCA), which efficiently solves for the optimal power, reduces interference, and improves the transmission rate.

[0071] In a specific example, the time allocation factor optimization in step S7 specifically includes: S71. The fixed power allocation result and the computational resource allocation result decompose the global optimization problem into K time allocation factors for a single subcarrier. Independent subproblems; S72. For each univariate subproblem, the golden section search method is adopted. Test points are selected in the interval [0,1] according to the golden ratio, and the corresponding objective function value is calculated. The search range is narrowed through interval shrinking iteration until the interval length is less than the preset accuracy threshold. The optimal time allocation factor of the subcarrier is then output. .

[0072] It should be noted that, since the devices are independent of each other, the optimization problem can be decomposed into multiple problems related to individual time allocation factors. Solve the subproblems separately:

[0073] Since this problem is non-convex and involves only a single variable, a one-dimensional golden section search method is used to solve it. Two test points are selected within the interval according to the golden ratio. Then, the objective function values ​​corresponding to the two test points are calculated, which represent the maximum completion delay under the current time allocation factor. Based on the relationship between the objective function values, the interval with the better objective function value is retained, while the other interval is discarded. This process of interval shrinkage and test point update is repeated to continuously narrow the search range. When the length of the search interval is less than a preset precision threshold, the search process is terminated, and the time allocation factor that minimizes the objective function value within the current interval is taken as the optimal solution.

[0074] This embodiment optimizes the two-stage time ratio of C-RSMA through the golden ratio search to obtain the optimal time allocation and further shorten the transmission latency.

[0075] In a specific instance, the computational resource allocation optimization in step S8 includes: With fixed power allocation results and time allocation factors, auxiliary variables are introduced. This represents the maximum generation delay of the system. The original computational resource allocation problem is transformed into a convex optimization problem, which is solved by a convex optimization solver to obtain the optimal computational resource allocation result for this round.

[0076] In the specific implementation, auxiliary variables are introduced. This represents the maximum completion delay of the system, transforming the original problem into the following equivalent form:

[0077] As can be seen, this problem is a convex problem, which can be solved using any convex optimization solver (such as CVX).

[0078] This embodiment transforms the allocation of computing resources into a convex optimization problem, rationally allocating computing power and reducing the latency of digital twin model generation.

[0079] Simulation verification: The specific implementation was simulated using MATLAB software. The specific parameters are set as shown in Table 1.

[0080] Table 1 Simulation Parameter Table

[0081] Figure 4 The curves showing the relationship between the overall generation latency of the digital twin system and the maximum transmit power of the remote terminal equipment are illustrated. It can be seen that as the maximum transmit power increases, the latency of each scheme gradually decreases. This is because increasing the transmit power improves the data transmission rate, thereby shortening the communication latency. In contrast, the method proposed in this invention achieves the lowest latency across all power ranges. This is because it effectively improves the system resource utilization efficiency through joint optimization of terminal pairing and resource allocation. The Benchmark C-RSMA scheme suffers performance degradation due to its fixed-rate splitting factor and insufficient pairing strategy flexibility; the Benchmark C-NOMA scheme further deteriorates in latency performance due to its poor resource allocation adaptability. Especially in the low transmit power region, the performance advantage of the method proposed in this invention is more significant, indicating its stronger capabilities in interference management and resource coordination.

[0082] Figure 5 The curves showing the relationship between the overall generation latency of the digital twin system and accuracy requirements are illustrated. It can be seen that as accuracy requirements increase, the latency of all schemes increases. This is because higher accuracy requirements require the transmission of more data, thus increasing the communication burden. The method of this invention consistently achieves the lowest latency under different accuracy requirements, and its advantage is more pronounced in the high-precision region. This indicates that the proposed method can effectively adapt to the requirements of high-reliability data transmission. In contrast, the Benchmark C-RSMA scheme has a slightly higher latency, highlighting the importance of the adaptive rate splitting mechanism; while the Benchmark C-NOMA scheme, due to its lack of flexible resource scheduling capabilities, experiences a significant performance degradation under high accuracy requirements.

[0083] Figure 6The graph shows the relationship between the overall generation latency of the digital twin system and the number of terminal devices. It can be seen that the latency of all schemes increases with the increase in the number of terminals. The method of this invention maintains the lowest latency performance across all user scales. In contrast, the Benchmark C-RSMA scheme, due to its pairing strategy being based solely on channel conditions and failing to consider the differences in data volume between terminals, suffers from resource allocation mismatch. This problem becomes more pronounced as the number of terminals increases, leading to performance degradation and widening the gap with the method of this invention.

[0084] Figure 7 The graph shows the relationship between the overall generation latency of the digital twin system and the total computing resources of the system. It can be seen that as the available computing resources increase, the latency of each scheme gradually decreases, which is because the increased computing power effectively reduces task processing time. The method of this invention achieves optimal performance under all computing resource conditions, further verifying the effectiveness of the proposed joint optimization method in computing resource allocation.

[0085] Example 2 To achieve the above objectives, this embodiment also provides a resource allocation system for C-RSMA-assisted digital twin network joint device pairing, used to perform the above method, including: The system confirmation module is used to identify the hardware composition, subcarrier allocation rules, and device access rules of the target digital twin uplink transmission system, and to clarify the optimization targets. The model building module is used to build a C-RSMA-based digital twin network uplink transmission system model, define the terminal device set, channel model, transmission delay and computation delay model, and clarify the system optimization objectives and full-process constraints. The device pairing module is used to execute a two-stage device pairing strategy, complete the one-to-one binding of subcarriers and device pairs, and generate a global device pairing set. The data volume optimization module is used to calculate the minimum amount of data to be transmitted for each terminal device to meet the minimum modeling requirements by combining modeling accuracy constraints and data accuracy models. The initialization module is used to initialize the initial values ​​of optimization variables for power allocation, time allocation factor, and computing resource allocation, and to set the conditions for iterative convergence. The alternating optimization module includes a power optimization unit, a time optimization unit, and a computing resource optimization unit, which are used to sequentially optimize power allocation, time allocation factor, and computing resources in stages. The iterative convergence module updates the parameters of the output of the alternating optimization module, repeatedly triggers the alternating optimization process until the convergence judgment condition is met, and outputs the globally optimal resource allocation scheme that adapts to the target system.

[0086] The resource allocation system for C-RSMA-assisted digital twin network joint device pairing described in this embodiment has the same advantages over the prior art as the resource allocation method for C-RSMA-assisted digital twin network joint device pairing described above, and will not be repeated here.

[0087] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A resource allocation method for joint device pairing in a C-RSMA-assisted digital twin network, applied to a digital twin uplink transmission system, wherein the system includes a base station, multiple physical terminal devices, and a digital twin server, and the total system bandwidth is divided into... Each subcarrier has orthogonal subcarriers, and each subcarrier can be assigned a pair of devices consisting of a near-end device and a far-end device, characterized in that... Includes the following steps: S1. Determine the target application system: Clarify the hardware composition, subcarrier allocation rules, and device access rules of the digital twin uplink transmission system, and identify the optimization target of this method; S2. For the above target application system, construct a digital twin network uplink transmission system model based on Cooperative Rate Split Multiple Access (C-RSMA), define the terminal device set, channel model, transmission delay and computation delay model, and clarify the system optimization objectives and full-process constraints to provide quantitative rules and optimization boundaries for subsequent steps. S3. Based on the channel model constructed in S2 and the native data requirements of terminal devices, a two-stage device pairing strategy is implemented: first, the nearest device with the optimal channel gain is matched for each subcarrier; then, the remote device is assigned to each subcarrier in combination with the native data requirements of the remote device and the channel gain of inter-device cooperation, thus completing the one-to-one binding of subcarriers and devices and generating a global device pairing set. S4. Based on the device pairing set generated by S3, and combined with the digital twin modeling accuracy constraints defined by S2, the minimum amount of data to be transmitted for each terminal device to meet the minimum modeling requirements is calculated through the data accuracy model, and the business boundary for optimizing transmission and computing resources is determined. S5. Based on the device pairing results of S3 and the minimum data transmission volume of S4, initialize the initial values ​​of three types of optimization variables: power allocation, time allocation factor, and computing resource allocation. At the same time, set the iterative convergence judgment conditions to complete the preparatory work for alternating optimization. S6. Using the initial values ​​of S5 or the time allocation factor and computational resource allocation results from the previous iteration, based on the system model of S2, the device pairing results of S3, and the minimum data transmission volume of S4, the non-convex power allocation problem is transformed into a convex one by using a stepwise convex approximation method. The optimal power allocation result for this round is obtained by solving the problem through convex optimization. S7, the optimal power allocation result of the current round output by fixed S6, the initial value of S5 initialization, or the computational resource allocation result of the previous iteration output, based on the system model of S2, the device pairing result of S3 and the minimum transmission data volume of S4, the optimal time allocation factor of the current round is obtained by using the golden section search method. S8, the current round's optimal power allocation result output by fixed S6, the current round's optimal time allocation factor output by S7, the system model of S2, the device pairing result of S3, and the minimum data transmission amount of S4, transform the computational resource allocation problem into a convex optimization problem and solve it to obtain the current round's optimal computational resource allocation result; S9. Update the parameters of the current round of optimization results output by S6-S8, and repeat the alternating optimization steps from S6 to S8 until the system delay change between two adjacent iterations meets the convergence judgment condition set by S5. The iteration terminates and outputs the globally optimal resource allocation scheme that adapts to the target application system of S1.

2. The resource allocation method for C-RSMA-assisted digital twin network joint device pairing according to claim 1, characterized in that, In step S2, the system optimization objective is to minimize the overall generation delay of the digital twin system, and the objective function expression is: in, For the device pairing set, the first The device pairs corresponding to each subcarrier are represented as follows: , and These represent the far-end and near-end devices allocated to subcarrier k, respectively. For equipment The actual amount of data transmitted. Assigning variables to power It is the transmission power of the remote device. It is the transmit power of the near-end device when sending its own split sub-messages. It is the power of the near-end device forwarding data from the far-end device. As a time allocation factor, For the computing resources allocated to the device, It is a subcarrier The set, and These represent the actual amount of data transmitted by the remote and near-end devices, respectively. and These represent the rates of the remote and near-end devices, respectively. and These represent the computation latency of the remote device and the local device, respectively.

3. The resource allocation method for C-RSMA-assisted digital twin network joint device pairing according to claim 2, characterized in that, In step S2, the overall process constraints include: , among which, among which For the complete collection of terminal devices, For equipment The amount of native state data generated; ,in For the transmission power of the remote device, This is the maximum transmit power of the remote device; ,in , The transmit power for the near-end device to send its own split sub-messages. The transmission power for near-end devices to relay data from far-end devices. This refers to the maximum transmit power of the near-end device; ,in For the first The time allocation factor of each subcarrier represents the time proportion of the two-stage transmission of C-RSMA. , For the first Each subcarrier corresponds to a device pair, ensuring that only one pair of devices is allocated to each subcarrier; Ensure that all terminal devices complete subcarrier allocation; ,in For equipment The accuracy of digital twin modeling, The minimum modeling accuracy threshold preset for the system; ,in To be assigned to the device Computing resources This represents the maximum total computing resources available for the digital twin server.

4. The resource allocation method for C-RSMA-assisted digital twin network joint device pairing according to claim 2, characterized in that, In step S2: the C-RSMA uplink transmission is divided into two stages: the first stage accounts for a certain time percentage. Remote devices send status data The signal is simultaneously received by the base station and the paired near-end device; the second phase time proportion is The near-end device splits its own data into two sub-messages. and , and the decoded remote device data Joint encoding and transmission are performed, and the base station completes data decoding through serial interference cancellation; the uplink transmission rate expression between the near-end device and the far-end device is: in, and These are the uplink transmission rates for the near-end device and the far-end device, respectively. and It is a sub-message of the base station decoding the near-end device. and rate, It is the near-end device decoding the sub-message of the remote device. rate, The base station directly decodes the sub-messages of the remote device. rate, It is a remote device sub-message The rate at which the base station decodes the data after it has been forwarded by the near-end equipment; the expression is as follows: in, The bandwidth of a single subcarrier; The signal-to-noise ratio of the link from the remote user to the base station; The signal-to-noise ratio of the link from the remote user to the near user; Decode the first part of the sub-message near the user for the base station Signal-to-noise ratio; For base station decoding of near-end user sub-messages and forwarding of far-end user sub-messages Signal-to-noise ratio; Decode the second part of the sub-message near the user for the base station The signal-to-noise ratio, of which and For remote users and nearby users To the base station on subcarrier Channel gain; For remote users To nearby users In subcarrier The channel gain is above, and all of the above channels follow Rayleigh fading. Indicates the noise variance; The expression for the time delay calculation model is: in, The computational resource requirement coefficient per bit of data.

5. The resource allocation method for C-RSMA-assisted digital twin network joint device pairing according to claim 1, characterized in that, The two-stage device pairing strategy in step S3 specifically includes: S31. Initialize the set of unallocated devices. ,make , For the complete collection of terminal devices; S32, Channel gain between all unallocated devices and subcarriers Construct the channel gain matrix and remove the smallest value. Each channel gain value This refers to the total number of subcarriers and the corresponding combination of devices and subcarriers. Records are unassignable sets If the combination of device and subcarrier Belonging to the set If so, the device will not participate in the allocation of that subcarrier; S33. Sort all subcarriers in descending order according to the number of channel gains deleted in each subcarrier, and select the near-end device for each subcarrier in turn; for the current subcarrier In the unassigned equipment set In the process, the device that satisfies the undeleted constraint and has the maximum channel gain is selected as the near-end device. : And update the set of unassigned devices: Repeat the above process until all subcarriers have been allocated to near-end devices; S34. Select the remaining unassigned devices as a candidate remote device set, and determine their data requirements accordingly. Sort in descending order; S35. Initialize the set of subcarriers for unassigned remote devices. For each remote device after sorting From the set Select to increase cooperative link gain The largest subcarrier is allocated, that is: It is a remote device Select the optimal subcarrier and put the device As the remote device of this subcarrier Simultaneously update the subcarrier set ; S36. After all devices are allocated, the device pairing set corresponding to each subcarrier is obtained. .

6. The resource allocation method for C-RSMA-assisted digital twin network joint device pairing according to claim 1, characterized in that, In step S4, the data accuracy model is: ,in For accuracy control parameters; combined with minimum modeling accuracy constraints The minimum data transmission volume for each terminal device is derived as follows: 。 7. The resource allocation method for C-RSMA-assisted digital twin network joint device pairing according to claim 1, characterized in that, The power allocation optimization in step S6 specifically includes: S61. Fixed time allocation factor and computational resource allocation results, introducing auxiliary variables. With target variable By performing equivalent substitutions on the power variable, new optimization variables can be constructed. The original minimization of maximum latency problem is transformed into... To optimize the equivalence problem of the objective; S62. Transform the non-convex rate constraint expression into a differential convex DC structure, perform a first-order Taylor expansion to linearize the concave function terms, and construct an approximate function for the lower bound of the rate. S63. Substitute the approximate rate expression into the constraint conditions to complete the transformation from a non-convex problem to a convex optimization problem. Solve the problem using a convex optimization solver to obtain the optimal power allocation result for this round.

8. The resource allocation method for C-RSMA-assisted digital twin network joint device pairing according to claim 1, characterized in that, Step S7, time allocation factor optimization, specifically includes: S71. The results of fixed power allocation and computational resource allocation decompose the global optimization problem into... A factor regarding the time allocation of a single subcarrier Independent subproblems; S72. For each univariate subproblem, the golden section search method is adopted. Test points are selected in the interval [0,1] according to the golden ratio, and the corresponding objective function value is calculated. The search range is narrowed through interval shrinking iteration until the interval length is less than the preset accuracy threshold. The optimal time allocation factor of the subcarrier is then output. .

9. The resource allocation method for C-RSMA-assisted digital twin network joint device pairing according to claim 1, characterized in that, Step S8, the optimization of computing resource allocation, specifically includes: With fixed power allocation results and time allocation factors, auxiliary variables are introduced. This represents the maximum generation delay of the system. The original computational resource allocation problem is transformed into a convex optimization problem, which is solved by a convex optimization solver to obtain the optimal computational resource allocation result for this round.

10. A resource allocation system for C-RSMA-assisted digital twin network joint device pairing, characterized in that, For performing the method according to any one of claims 1-9, comprising: The system confirmation module is used to identify the hardware composition, subcarrier allocation rules, and device access rules of the target digital twin uplink transmission system, and to clarify the optimization targets. The model building module is used to build a C-RSMA-based digital twin network uplink transmission system model, define the terminal device set, channel model, transmission delay and computation delay model, and clarify the system optimization objectives and full-process constraints. The device pairing module is used to execute a two-stage device pairing strategy, complete the one-to-one binding of subcarriers and device pairs, and generate a global device pairing set. The data volume optimization module is used to calculate the minimum amount of data to be transmitted for each terminal device to meet the minimum modeling requirements by combining modeling accuracy constraints and data accuracy models. The initialization module is used to initialize the initial values ​​of optimization variables for power allocation, time allocation factor, and computing resource allocation, and to set the conditions for iterative convergence. The alternating optimization module includes a power optimization unit, a time optimization unit, and a computing resource optimization unit, which are used to sequentially optimize power allocation, time allocation factor, and computing resources in stages. The iterative convergence module updates the parameters of the output of the alternating optimization module, repeatedly triggers the alternating optimization process until the convergence judgment condition is met, and outputs the globally optimal resource allocation scheme that adapts to the target system.

Images