A radiation-aware multi-cell coordinated transmission method and system supporting sustainable urllc traffic

Abstract

The application discloses a radiation-aware multi-cell cooperative transmission method and system supporting sustainable URLLC service, relates to the technical field of wireless communication, and comprises the following steps: constructing a system model; based on the system model, a long-term random optimization problem facing multiple time slots is constructed with the target of minimizing system long-term time average transmission power; a virtual queue is introduced to convert the long-term random optimization problem into a single time slot deterministic optimization sub-model in each scheduling period; the association relationship between each cellular user and a multi-antenna base station is determined through centralized global cooperative matching; each multi-antenna base station independently solves the beamforming and subcarrier allocation strategy of itself in a distributed manner, and the dynamic repeated execution in continuous scheduling periods is implemented to realize multi-cell cooperative data transmission. The application realizes joint resource scheduling and long-term stability guarantee across cells in view of the problems of strong burstiness, interference sensitivity and high centralized optimization cost of URLLC service.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and in particular to a radiation-sensing multi-cell cooperative transmission method and system that supports sustainable URLLC services. Background Technology

[0002] With the rapid development of mission-critical applications such as the Industrial Internet, the Internet of Vehicles, and intelligent manufacturing, wireless networks need to provide ultra-reliable, low-latency communication capabilities stably over long periods in complex environments. To meet stringent latency and reliability requirements, existing systems generally employ short packet transmission mechanisms with limited code lengths to shorten channel occupancy time. However, this approach inevitably leads to a higher probability of decoding errors, causing a decrease in system reliability during actual operation. Furthermore, while single-cell architectures can reduce interference through dedicated spectrum or simple power control, their capacity and reliability are difficult to guarantee in scenarios with increased user density and drastic fluctuations in service load, failing to meet the continuous requirements of Ultra-reliable Low-Latency Communication (URLLC) services.

[0003] To improve overall performance, multi-cell cooperative transmission has become an important development direction. By sharing spatial resources and coordinating scheduling among multiple base stations, more efficient load sharing and interference suppression can be achieved during momentary service congestion. However, existing cooperative methods mostly rely on centralized control, requiring the acquisition of channel state and interference information for the entire network, which leads to a rapid increase in computational complexity and backhaul overhead, making it difficult to meet real-time requirements. Furthermore, in cases where multiple cell coverage areas overlap, the lack of an effective radiation constraint mechanism results in excessively high leakage radiation from base stations in idle areas, further exacerbating the difficulty of interference management and weakening the overall efficiency of cross-cell cooperation.

[0004] On the other hand, wireless networks exhibit significant time-varying characteristics in actual operation, with user arrival, traffic volume, and channel conditions all showing continuous random fluctuations. Much existing research remains focused on static optimization models within a single time slot, failing to describe the queuing, transmission, and delay evolution of tasks across multiple consecutive time slots. Consequently, it is difficult to balance system power consumption, queue stability, and reliability over long periods. Especially in URLLC scenarios, the lack of dynamic management of data queue evolution can easily lead to system backlog and instability, resulting in task latency violating service requirements.

[0005] Based on the aforementioned practical needs and technical limitations, existing multi-cell URLLC systems urgently require a distributed cooperative transmission scheme capable of operating in dynamic network environments. This scheme would enable base stations to cooperate with low overhead without relying on centralized processing, and mitigate mutual interference between coverage areas by reasonably constraining spatial radiation. Simultaneously, a dynamic optimization framework is needed to achieve multi-timeslot-oriented resource allocation under random service arrivals and channel variations, thereby improving energy efficiency while maintaining long-term service stability and the reliable performance of URLLC. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a radiation-sensing multi-cell cooperative transmission method and system that supports sustainable URLLC services.

[0007] To solve the above technical problems, the technical solution of the present invention is as follows:

[0008] A radiation-aware multi-cell cooperative transmission method supporting sustainable URLLC services includes the following steps:

[0009] Step S10: Construct the system model, including:

[0010] A multi-cell downlink communication model is constructed, which includes a cloud server, multiple multi-antenna base stations and multiple cellular users with URLLC requirements, and the downlink transmission from the multi-antenna base stations to the cellular users is based on multiple orthogonal subcarriers;

[0011] Construct a task queue model that dynamically updates the queue backlog status based on the amount of user data arriving and the actual amount of data leaving.

[0012] A radiation footprint model is constructed, which discretizes the total service area into multiple sub-regions, extracts idle sub-regions that do not carry active users, and evaluates the expected radiation signal power in the idle sub-regions based on channel statistical characteristics.

[0013] Step S20: Based on the system model, with the goal of minimizing the long-term average transmit power of the system, a long-term stochastic optimization problem oriented towards multiple time slots is constructed. The optimization problem includes constraints on the maximum transmit power of the base station, subcarrier allocation, radiated power in idle areas, queue stability, and transmission rate.

[0014] Step S30: Introduce a virtual queue to handle the queue stability constraint, and transform the long-term stochastic optimization problem into a single-slot deterministic optimization sub-model within each scheduling cycle;

[0015] Step S40: Based on the single-slot deterministic optimization sub-model, determine the association between each cellular user and the multi-antenna base station through centralized global cooperative matching;

[0016] Step S50: Each of the multi-antenna base stations independently solves its own beamforming and subcarrier allocation strategy in a distributed manner according to the association relationship, and dynamically repeats the execution within a continuous scheduling period to achieve multi-cell collaborative data transmission.

[0017] As a preferred embodiment of the radiation-aware multi-cell cooperative transmission method supporting sustainable URLLC services according to the present invention, wherein: the construction of the multi-cell downlink communication model in step S10 includes:

[0018] The total number of base stations in a multi-cell downlink communication system is defined as follows: The total number of users with URLLC requirements is The number of antennas in each base station is The number of orthogonal carriers in each cell is , define the first The first multi-antenna base station The number of subcarriers is ,make This represents the channel coefficient of user k on this subcarrier during the scheduling period t, and the value of user k in the t-th subcarrier. The received signal on each subcarrier is represented as follows: ,in, This indicates the number of subcarriers within the scheduling period t. From base station To users Beamforming vector, and These represent the transmitted signal and additive white Gaussian noise, respectively.

[0019] Set binary constraints for base station selection and subcarrier allocation The corresponding signal-to-noise ratio is: ,in, This indicates that user terminal k is on a subcarrier The total interference power received by the device;

[0020] Under the condition of finite block length, user terminal k within the scheduling period t, subcarrier The transmission rate on it is: ,in, Used to characterize channel dispersion, the function Q(·) represents the Gaussian Q-function. Its inverse function, The channel block length, Let be the decoding error probability of user terminal k.

[0021] As a preferred embodiment of the radiation-aware multi-cell cooperative transmission method supporting sustainable URLLC services according to the present invention, wherein: the task queue model construction in step S10 includes:

[0022] The service load requested by the user terminal enters its corresponding cache queue in a first-in-first-out manner, forming a user data queue. The user data queue This represents the amount of data waiting to be served at user terminal k at time t;

[0023] Each user's data queue is initially empty, and at time... Data queue backlog for user k The update relationship is as follows: ,in, This represents the amount of data actually transmitted and sent by user terminal k within the scheduling period t. This represents the amount of data generated by user terminal k within the scheduling period t;

[0024] The queue stability constraint satisfies: ,in, This represents the statistical expectation relative to random channel conditions and the arrival process of the task.

[0025] As a preferred embodiment of the radiation-aware multi-cell cooperative transmission method supporting sustainable URLLC services according to the present invention, wherein: the construction of the radiation footprint model in step S10 includes:

[0026] The total service area is divided into contiguous rectangular sub-regions, and an index set for all sub-regions is created. Within the scheduling period t, a set of sub-region indexes containing the active cellular users is determined based on their location coordinates. , the index set With the set The difference set is defined as the set of free sub-regions that do not carry cellular users. ,Right now ;

[0027] For any set of free subregions subregion Multi-antenna base stations To the sub-region In subcarrier The propagation channel model is as follows: ,in, Indicates the guide vector. express A complex vector of dimension 1 Here, A is the small-scale fading coefficient, and A is the path loss constant. For log-normal shadowing fading variables, The path decay exponent. and The multi-antenna base stations are respectively and the sub-region The coordinates of the center point;

[0028] Based on the statistical characteristics of the propagation channel from the multi-antenna base station to the idle sub-region, the expected value of the radiated signal power generated by each multi-antenna base station in the idle sub-region is calculated as follows: ,in, Represents the statistical signal coefficients. To The conjugate transpose of . Indicates subcarrier From base station To users The beamforming vector.

[0029] As a preferred embodiment of the radiation-aware multi-cell cooperative transmission method supporting sustainable URLLC services according to the present invention, in step S20, the long-term stochastic optimization problem oriented towards multiple time slots is constructed with the objective of minimizing the long-term average transmit power of the system as the goal.

[0030] Based on the premise of jointly optimizing base station selection, subcarrier allocation, and transmit beamforming vector, an objective function is constructed to minimize the long-term average transmit power of the system: ,in, The total transmit power of all base stations during the scheduling period t is expressed as: ;

[0031] Define a set of joint constraints, the set of constraints including:

[0032] The user-subcarrier unique access constraint is used to limit a single cellular user to access at most a single subcarrier of a single multi-antenna base station, and its expression is: ;

[0033] The maximum transmit power constraint of the base station is used to limit the transmit power of each of the multi-antenna base stations to no more than its corresponding maximum available power. Its expression is: ;

[0034] The idle area radiated power constraint is used to limit the set of idle sub-regions that do not carry the cellular users. any subregion within Its cumulative signal radiated power is expected to be no greater than a specified threshold. Its expression is: ;

[0035] The transmission rate constraint is used to ensure that the amount of task data requested by a new request meets the data transmission requirements within the specified scheduling period, and its expression is: ,in, Indicates user terminal At any moment The amount of data generated Indicates user terminal At any moment The amount of data transferred internally, and ;

[0036] The queue stability constraint is used to determine the actual queue backlog in the communication network. To keep it within the average threshold, the expression is: ,in, This is a preset threshold for the cumulative queue size.

[0037] As a preferred embodiment of the radiation-aware multi-cell cooperative transmission method supporting sustainable URLLC services according to the present invention, wherein: in step S30, a virtual queue is introduced to handle the queue stability constraint, and the long-term stochastic optimization problem is transformed into a single-slot deterministic optimization sub-model within each scheduling cycle, including:

[0038] remember For the virtual power queue introduced for user k at time t under long-term average constraints, the virtual queue at the next time step is: ;

[0039] The quadratic Lyapunov function for measuring the size of the virtual queue is constructed as follows: , where, definition The concatenated vector accumulated for the total queue;

[0040] Calculate the expected change of the Lyapunov function over a scheduling period as the Lyapunov drift. : And introduce control parameter V to control Lyapunov drift. With the objective function transmit power By weighing the trade-offs, the upper bound expression that needs to be minimized in each scheduling period t is derived as follows: ,in, It is a non-negative constant. These are control parameters used to make trade-offs between drift and target. This is a constant term under scheduling period t;

[0041] Add to the constraints of the optimization problem The max function in the upper bound expression is eliminated, and based on the dynamic changes in queue length and environment, the transmission rate constraint spanning multiple scheduling cycles is adaptively decomposed into a minimum throughput constraint for a single cycle. : ,in, For users At any moment The maximum number of transmission time slots allowed for the task. Indicates the start time of the transmission for this task;

[0042] By removing the constant term from the upper bound expression and minimizing the remaining deterministic objective function, the original stochastic optimization problem is transformed into a continuous single-slot online deterministic optimization sub-model, whose objective function expression is: And the constraint is And the maximum transmit power constraint, subcarrier allocation constraint, and idle area radiated power constraint of the base station. This represents the control parameter V used to balance drift and target, and the total transmit power of all base stations during the scheduling period t. The product of.

[0043] As a preferred embodiment of the radiation-sensing multi-cell cooperative transmission method supporting sustainable URLLC services according to the present invention, wherein: in step S40, determining the association relationship between each cellular user and the multi-antenna base station through centralized global cooperative matching based on the single-slot deterministic optimization sub-model includes:

[0044] The symbolic representation of time t is omitted, and the binary correlation indicator variable is defined as follows: The Used to characterize cellular user k and multi-antenna base station Given the association states between them, and with the matching constraint of dividing all cellular users into L non-overlapping user groups, the base station selection objective function aimed at minimizing the overall system cost is constructed as follows: ,in, The virtual power queue introduced for user k For users Data queue backlog For user terminals In subcarrier The transmission rate on ;

[0045] Construct a cluster utility function that reflects the tradeoff between power consumption and data rate. Based on this, a transitive strict preference order is established for the cellular user for the multi-antenna base station, and an association switching determination criterion is set: cellular user k switches from the multi-antenna base station if and only if... service cluster Switch to multi-antenna base station service cluster Subsequently, when the total utility of the two service clusters involved in the handover is significantly greater than the total utility before the handover, the cellular user k triggers an association switching operation, and the determination criterion satisfies: ;

[0046] Based on the preset upper limit of the switching ratio Limit the number of cellular users participating in association switching, and obtain the maximum statistical channel gain from each cellular user to all multi-antenna base stations. With minimum statistical channel gain Calculate the absolute value of the corresponding statistical channel gain difference. Sort them in ascending order and extract the first few from the ascending sort. A set of cellular user plans is generated for each element, and the associated exchange is iteratively executed within the set of cellular user plans according to the associated exchange judgment criteria until the matching is stable, where K is the total number of cellular users.

[0047] As a preferred embodiment of the radiation-sensing multi-cell cooperative transmission method supporting sustainable URLLC services according to the present invention, wherein: in step S50, each of the multi-antenna base stations independently solves its own beamforming and subcarrier allocation strategy in a distributed manner according to the association relationship, including:

[0048] Based on a non-convex optimization problem involving a composite objective function and discrete scheduling constraints, a compact lower bound convex function of the logarithmic rate function is constructed using a continuous convex approximation algorithm. A first auxiliary variable is introduced using a quadratic transformation method to construct an upper bound convex function of the logarithmic rate function. A substitution function is established through successive approximations to achieve affine form transformation, and a second auxiliary variable is introduced using a difference method to construct the lower bound of the substitution function.

[0049] Non-smoothness in subcarrier allocation constraints Norm terms are introduced to establish a smooth approximation model using a continuous smooth function. Based on the inherent non-convexity of the smooth function, a quadratic upper bound function containing a dynamic penalty coefficient is constructed to equivalently define the smooth function.

[0050] The original objective function, transmission rate constraint, and subcarrier allocation constraint in the non-convex optimization problem are replaced with the derived compact lower bound convex function, upper bound convex function, and quadratic upper bound function, respectively. This reconstructs a convex optimization problem in which the objective function is separable convex and all constraints are convex and quadratic in structure. The interior point method is then used to iteratively solve the convex optimization problem.

[0051] As a preferred embodiment of the radiation-aware multi-cell cooperative transmission method supporting sustainable URLLC services according to the present invention, wherein: the calculation expression of the first auxiliary variable is: ;

[0052] The expression for calculating the second auxiliary variable is: ;

[0053] The introduction of a continuous smooth function to establish a smooth approximation model includes: using a logarithmic smooth function to establish a smooth approximation model, which is equivalently defined as follows: ,in, To control the non-negativity of the approximation accuracy of smoothness;

[0054] The expression for the upper bound function of the quadratic form is: ,in, , , .

[0055] The present invention also provides a radiation-aware multi-cell cooperative transmission system supporting sustainable URLLC services, for implementing the method described in any of the above claims, comprising:

[0056] The model building module is used to build system models, including multi-cell downlink communication models, task queue models, and radiation footprint models.

[0057] The long-term optimization problem construction module is used to construct a long-term stochastic optimization problem for multiple time slots based on the system model, with the goal of minimizing the long-term average transmit power of the system. The optimization problem includes constraints such as the maximum transmit power of the base station, the unique access constraint between the user and the subcarrier, the radiated power constraint in the idle area, the queue stability constraint, and the transmission rate constraint.

[0058] The online problem dimensionality reduction and transformation module is used to introduce a virtual queue to handle the queue stability constraints and transform the long-term stochastic optimization problem into a single-slot deterministic optimization sub-model within each scheduling cycle.

[0059] The global matching module is used to determine the association between each cellular user and the multi-antenna base station through centralized global cooperative matching based on the single-slot deterministic optimization sub-model.

[0060] The distributed resource solving and transmission module is used to independently solve its own beamforming and subcarrier allocation strategy in a distributed manner according to the association relationship, and dynamically repeat the execution in a continuous scheduling period to realize multi-cell collaborative data transmission.

[0061] The beneficial effects of this invention are:

[0062] (1) In existing systems, when multiple cells overlap, the invalid transmissions of each base station in the idle area will result in extremely high leakage radiation, exacerbating the difficulty of interference management. This invention innovatively constructs a radiation footprint model, which adaptively controls the power distribution of the signal in space by accurately evaluating and strictly limiting the expected power of the radiated signal in the idle sub-area based on channel statistical characteristics. This mechanism not only cuts off invalid electromagnetic radiation from the spatial source and significantly reduces uncontrollable cross-cell mutual interference in the overlapping coverage area, but also establishes a green dynamic resource management mechanism with the goal of minimizing long-term power consumption, greatly improving the sustainability of the overall network.

[0063] (2) To address the problem that URLLC services are highly bursty and existing static optimization methods cannot describe the long-term queuing delay evolution, this invention adopts Lyapunov stochastic optimization theory to cleverly reduce the long-term stochastic optimization problem spanning multiple time slots into a single-time-slot online deterministic optimization sub-model that depends on the current state. In particular, this invention proposes a "divide and conquer" adaptive constraint decomposition strategy, which hard-decomposes the transmission rate constraint across scheduling cycles into the minimum throughput constraint (minimum passing grade) of the current time slot. This effectively avoids long-term queue backlog and network congestion caused by burst traffic, and provides absolutely reliable delay guarantee for URLLC tasks while minimizing the time-averaged transmit power.

[0064] (3) When handling multi-cell user-base station association, existing centralized multi-cell cooperation suffers from extremely high backhaul overhead and computational latency due to the need to obtain global instantaneous channel state information. This invention models base station selection as a many-to-one finite matching game problem, guiding association switching by constructing a utility function that reflects the trade-off between power consumption and data rate. Simultaneously, it innovatively specifies a planned handover set arranged in ascending order based on the difference between the maximum and minimum statistical channel gain, and strictly limits the proportion of users allowed to handover. This mechanism avoids frequent handovers between user terminals and base stations, significantly reducing the signaling overhead and backhaul burden required to obtain global information, perfectly meeting the stringent requirements of URLLC scenarios for extremely fast response.

[0065] (4) To address the complex problem of mixed-integer nonlinear programming involving composite objective functions and discrete scheduling constraints, this invention proposes a highly original distributed dimensionality reduction solution architecture. To address the interference coupling generated by transmit beamforming in the rate function, dual auxiliary variables are introduced through continuous convex approximation, quadratic transformation, and difference methods, perfectly decoupling the numerator and denominator of the multi-user interference term; to address the non-smoothness caused by discrete scheduling of subcarrier allocation... Norm constraints are applied, and a smooth approximation model is established using a logarithmic smoothing function. Furthermore, a quadratic upper bound function with dynamic penalty coefficients is constructed. By driving the penalty coefficient of the beamforming vector, which approaches zero, to its maximum value during iteration, sparse updates are forced. This distributed solution mechanism not only approximates the global optimum but also endows each base station with strong independent computing capabilities, greatly improving the solution efficiency and feasibility of multi-cell collaborative transmission. Attached Figure Description

[0066] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0067] Figure 1 A flowchart illustrating the distributed cooperative multi-time-slot wireless resource allocation method based on radiation sensing provided by this invention.

[0068] Figure 2 This is a system model diagram constructed in the radiation-sensing-based distributed cooperative multi-time-slot wireless resource allocation method provided by the present invention;

[0069] Figure 3 A schematic diagram showing the change of data queue length over time and different control parameters;

[0070] Figure 4 This is a schematic diagram showing the variation of time-averaged transmit power with time and different control parameters.

[0071] Figure 5 This is a graph showing the relationship between time-averaged transmit power and maximum transmission delay.

[0072] Figure 6 Another flowchart of the distributed cooperative multi-time slot wireless resource allocation method based on radiation sensing provided by the present invention. Detailed Implementation

[0073] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0074] See Figure 1 This application provides a distributed cooperative multi-timeslot radio resource allocation method based on radiation sensing. This method addresses the problems of high burstiness, interference sensitivity, and high cost of centralized optimization in URLLC services, achieving joint resource scheduling and long-term stability assurance across cells. The method specifically includes the following steps:

[0075] Step S10: Construct the system model, including constructing a multi-cell downlink communication model, a task queue model, and a radiation footprint model. The construction methods for each model are as follows:

[0076] Constructing a multi-cell downlink communication model:

[0077] First, a multi-cell space division multiple access (SDMA) downlink network model is established, including one cloud server, L multi-antenna base stations, and K cellular users with specific URLLC requirements. The cloud server has a built-in caching module, and each user terminal obtains service data from the cloud server through a connected base station. Data transmission connections are established between the cloud server and each base station via wired fiber optic links. Each base station is configured with... Each user terminal is equipped with a single antenna. In this network, a unified frequency reuse method is used to improve spectrum utilization efficiency; downlink transmission from the base station to the user terminal is based on… The process is performed using orthogonal subcarriers, with each subcarrier having a bandwidth of [missing information]. In this application, the scheduling period (SP) is defined as a fixed-length time interval during which service quality assessment, resource allocation planning, and data transmission operations are periodically performed. It is typically assumed that the channel coherence time is longer than the duration of a single frame; therefore, channel state information remains largely valid across multiple short frames and can be used for subsequent short-term resource scheduling. The allowable service transmission delay may be longer than the coherence time.

[0078] Then, define the first The first multi-antenna base station The number of subcarriers is ,make Let represent the channel coefficient of user terminal k on this subcarrier during the scheduling period t. User terminal k in the t-th... The received signal on each subcarrier can be represented as: ,in, This indicates the number of subcarriers within the scheduling period t. From base station To users Beamforming vector, and Let represent the transmitted signal and additive white Gaussian noise, respectively. Considering the synchronization error between base stations and the uncertainty of multi-base station connections, binary constraints are set for base station selection and subcarrier allocation. This limits the connectivity of user terminals on specific base stations and subcarriers. The corresponding signal-to-interference-plus-noise ratio (SINR) can be expressed as: ,in, This indicates that user terminal k is on a subcarrier The total interference power received. This interference power includes intra-cell interference generated by other users within the same cell, and inter-cell interference caused by parallel transmissions from other base stations, i.e. .

[0079] Under the condition of finite block length, user terminal k within the scheduling period t, subcarrier transmission rate (Unit: bits / s / Hz) can be expressed as: ,in, Used to characterize channel discreteness, function Represents the Gaussian Q-function. Its inverse function, The channel block length, Let be the decoding error probability of user terminal k. Let the transmission duration of each scheduling cycle be denoted as . Then the user terminal During the scheduling period The number of bits transmitted within is: .

[0080] Construct a task queue model:

[0081] By constructing a queue scheduling model for dynamically tracking cache status, real-time management and control of data queues can be achieved. Service loads requested by user terminals enter their corresponding cache queues in a first-in, first-out (FIFO) manner, thus forming the actual queues. This queue represents user terminals. At any moment The amount of data waiting to be served. By default, each queue is empty at the initial moment, that is... At that moment Data queue backlog for user k Updated to: ,in, This represents the amount of data actually transmitted and sent by user terminal k within the scheduling period t. This represents the amount of data generated by user terminal k within the scheduling period t, and its maximum transmission duration is defined as follows: Specifically, if the current transmission and Correspondingly, there is And its transmission delay The following conditions must be met: , .

[0082] Considering the stability requirements of dynamic systems, each user terminal must ensure that the long-term average arrival rate of its queue does not exceed the long-term average departure rate, i.e. ,in, This represents the statistical expectation relative to random channel conditions and the task arrival process. Furthermore, according to Little's law, the average latency is proportional to the average queue length. This principle regulates latency by limiting queue length, thereby ensuring that system latency remains finite and within an acceptable range.

[0083] Constructing a radiation footprint model:

[0084] To accurately determine the radiation footprint of all base stations within the service area, the propagation channel of each base station across the entire service area is modeled. Specifically, the total service area is divided into continuous rectangular sub-regions, each with side lengths of [missing information]. and Meters. The number of sub-regions along the x-axis is... The number of sub-regions along the y-axis is ,in and This indicates the size of the service area. The coordinates of the diagonal points of the service area are set to the origin (0, 0) and... Then, the set of indices for all sub-regions is represented as .

[0085] Given a multi-antenna base station Location and sub-regions center point Subsequently, multi-antenna base stations can be further developed. To sub-region In subcarrier Propagation channel modeling on ,in, Indicates the guide vector. express A complex vector of dimension 1 Here, A is the small-scale fading coefficient, and A is the path loss constant. For log-normal shadowing fading variables, The path decay exponent. and The multi-antenna base stations are respectively and the sub-region The coordinates of the center point. To improve spectrum utilization efficiency and reduce ineffective electromagnetic radiation in areas with no service demand, the radiation intensity of base stations in sub-areas that do not contain active user terminals of this cell are subject to key control. First, Defined as the set of sub-region indices containing user terminals within a scheduling period t. Correspondingly, This is used to represent the set of sub-regions that do not carry user terminals during this period. Based on the above definition, for any sub-region belonging to... subregion For base stations in scheduling period t, subcarrier The propagation signal radiating upwards into this sub-region is modeled and represented as: .

[0086] To reduce signaling overhead, the radiation footprint is managed using the statistical characteristics of the channel. For those belonging to... subregion The base station is on the subcarrier The expected power of the radiated signal on the surface can be expressed as: ,in, Represents the statistical signal coefficients. To The conjugate transpose of . Indicates subcarrier From base station To users The beamforming vector.

[0087] Step S20: Based on the system model, with the goal of minimizing the long-term average transmit power of the system, a long-term stochastic optimization problem oriented towards multiple time slots is constructed. The optimization problem includes constraints on the maximum transmit power of the base station, subcarrier allocation, radiated power in idle areas, queue stability, and transmission rate.

[0088] Specifically, a model is constructed for the transmit power optimization problem of a multi-cell system, making... Indicates the scheduling period The total transmit power of all base stations within the area is expressed as: .

[0089] Since each user is only allowed to access one subcarrier under any base station, then The following constraints shall apply: .

[0090] Multi-antenna base station The transmission power must not exceed its maximum usable power. , represented as: .

[0091] The cumulative signal radiated power in the idle sub-region must not exceed a specified threshold. ,Right now: .

[0092] To ensure the mission For a transmission to be completed within a specified scheduling period, its transmission rate should meet the following conditions: ,in, Indicates the start time of the transmission for this task. Indicates user terminal At any moment The amount of data generated Indicates user terminal At any moment The amount of data transferred internally, and The stability of a communication network is guaranteed by the following constraints: Controlled within a tolerable average threshold: ,in, This is a preset threshold for the cumulative queue size.

[0093] This application jointly optimizes base station selection, subcarrier allocation, and beamforming to minimize the long-term time-averaged transmit power of the system. The objective function can be expressed as follows: .

[0094] Step S30: Introduce a virtual queue to handle the queue stability constraint, and transform the long-term stochastic optimization problem into a single-slot deterministic optimization sub-model within each scheduling cycle.

[0095] Specifically, this involves implementing an online transformation of a multi-slot discrete optimization problem. To handle long-term time-averaged constraints related to queue buffers, [the following is implemented] to address the constraints. A virtual queue was introduced, and the record... For the virtual power queue introduced for user k at time t under long-term average constraints, the virtual queue at the next time step is: .

[0096] To simplify the notation, the definition will be... The concatenated vector representing the total queue size. A quadratic Lyapunov function used to measure the size of the virtual queue. Expressed as: .

[0097] Accordingly, Lyapunov drift is defined as the expected change in the Lyapunov function over a scheduling period, i.e.: .

[0098] in, These are control parameters used to control drift. With objective function A trade-off must be struck between these factors. According to Lyapunov optimization theory, wireless resource allocation should achieve the following in each scheduling cycle t: The upper bound is minimized. This upper bound can be expressed as: ,in, Let be a non-negative constant, and its expression is: ,and, It is a constant under scheduling period t, which is determined by the current system state, as shown below: .

[0099] Because the right-hand side formula in the above upper bound expression exists Its expression is mathematically nondifferentiable. If ,but Equivalent to Therefore, if constraints are added... If so, the max function in the upper bound expression can be eliminated.

[0100] In addition, constraints The problem becomes complex due to its spanning multiple scheduling cycles. Therefore, a "divide and conquer" strategy is adopted, adaptively decomposing the constraint into minimum throughput constraints for each single cycle based on changes in queue length and environmental dynamics. Specifically: ,in, For users At any moment The maximum number of transmission time slots allowed for the task. This indicates the start time of the task's transmission.

[0101] By removing the constant term from the upper bound expression, the deterministic problem under the t-th SP can be written as: And the constraint is and the maximum transmit power constraint of the base station Subcarrier allocation constraints and radiated power constraints in the idle region , This represents the control parameter V used to balance drift and target, and the total transmit power of all base stations during the scheduling period t. The product of . Here, the original stochastic optimization problem is transformed into a continuous online problem.

[0102] Step S40: Based on the single-slot deterministic optimization sub-model, determine the association between each cellular user and the multi-antenna base station through centralized global cooperative matching.

[0103] Specifically, a method for globally partitioning the user set is first presented to determine which base station is responsible for the data transmission of each user. To this end, auxiliary variables are introduced. Used to represent the association between a base station and a user terminal, where, when Time indicates user terminal Associated with multi-antenna base stations ,otherwise Based on this, the base station selection subproblem can be expressed as: ,in, The virtual power queue introduced for user k For users Data queue backlog For user terminals In subcarrier The transmission rate on satisfy and It can be observed that the above problem is a many-to-one matching problem involving two finite and disjoint sets, with the goal of dividing all user terminals into L non-overlapping user groups.

[0104] During the matching process, each user terminal maintains a transitive, strict preference order for the base station set L. Correspondingly, each multi-antenna base station... It will also have a corresponding preference order for the user set K. (Symbol) This indicates that user terminal k strictly prefers base station Instead This preference relationship is transitive: if and Then there is To address the aforementioned peer effect and solve this user terminal clustering game problem, we first define exchange matching as follows: ,in, This indicates that when user terminal k is in cluster The utility obtained by this cluster during the process. That is, if user terminal k is associated with the base station... The total utility obtained at that time is better than that associated with the base station. The utility of time, then user terminal k will prefer matching Instead Specifically, user terminals switch between different clusters based on a trade-off between power consumption and data rate, and the user terminal's choice of base station depends on changes in cooperation utility. A further explanation of the strict matching relationship is as follows: .

[0105] To reduce the signaling overhead incurred by user terminals during base station access handover, the permitted handover ratio for user terminals is stipulated to be no more than [a certain percentage]. Furthermore, the upper-level centralized control unit only collects statistical channel gain information after every few scheduling cycles. Subsequently, the user terminal... The maximum and minimum statistical channel gains to each base station are denoted as follows: and Calculate and sort the differences in ascending order. The first in this sorted list The set of user terminals planned for switching consists of 12 elements, denoted as _____. By employing comparison and swap operations, the user terminal compares the utility of all available options and decides whether to perform a swap based on a predetermined priority relationship.

[0106] Step S50: Each multi-antenna base station independently solves its own beamforming and subcarrier allocation strategy in a distributed manner according to the correlation relationship, and dynamically repeats the execution within a continuous scheduling period to achieve multi-cell collaborative data transmission.

[0107] Specifically, step S40 above provides a distributed solution mechanism for wireless resource allocation. After the base station selection result is determined, each base station can independently complete the resource allocation design for its own cell in a distributed manner. Among them, the... The optimization problem to be solved for a multi-antenna base station can be expressed as: ,in Due to the existence of the composite objective function and discrete scheduling constraints, the above-mentioned... The optimization problem for multi-antenna base stations exhibits a non-convex structure, making it difficult to guarantee a globally optimal solution. To address this issue, a low-complexity method is proposed to iteratively find a stable solution.

[0108] For ease of expression, Defined as the rate term. In this logarithmic function In the middle, the transmitted beam is formed Will pass The interference term in the denominator, and other user terminals This introduces coupling. Therefore, by constructing tight upper and lower bound approximations, the optimization variables can be effectively decoupled, and the original non-convex terms can be transformed into convex forms. Based on this, the following results are presented.

[0109] First of all, let In the i-th iteration of the continuous convex approximation algorithm The solution result is then It can be approximated by the following compact lower bound convex function: ,in, as well as .at the same time, The upper bound of can be approximated by the following convex function, namely ,in, Furthermore, auxiliary variables are introduced through a quadratic transformation method. Its form is: .

[0110] Then, due to about Since it is concave, it can be transformed into an affine form by establishing a substitution function through successive approximations. Furthermore, a difference method is proposed for constructing... The lower bound is determined, and the numerator and denominator are decoupled. Specifically, for any feasible... , An upper bound can be constructed by the following formula: ,in, The gradient is constant. Meanwhile, A tight lower bound can be approximated as: ,here Calculated as .

[0111] Combining the above formulas, the equation is... The objective function and rate-related constraints can both be reconstructed into convex upper bound functions and convex constraints. Before that, the last non-convex constraint must be processed.

[0112] The final step is to introduce a smooth function to approximate the result. Normative terms, thus handling constraints The non-smoothness of the material. Its form can be rewritten as... .

[0113] When parameter When the value is sufficiently small, the function can be considered as A precise and smooth approximation of the norm. Due to the nonconvexity introduced by the logarithmic function, this term can be defined as follows: ,in, , ,here In the i-th iteration, if If it is close to zero, then It will become extremely large, which will prompt Update along the zero direction in the next iteration.

[0114] By replacing the non-convex objective function, subcarrier allocation constraints, and transmission rate constraints with the upper and lower bounds derived above, the original optimization problem is reconstructed, making the objective function a separable convex form, and all constraints convex and quadratic in structure. Therefore, the reconstructed problem is essentially a convex optimization problem, which can be solved efficiently using the interior-point method.

[0115] After acquiring channel state information and related network status, the centralized control unit first performs the association between user terminals and base stations, and allocates the main computing tasks to the corresponding base stations. The proposed online cooperative scheduling scheme is divided into two key stages: (1) centralized base station selection based on a many-to-one finite matching algorithm; (2) each designated base station performs cooperative beamforming and subcarrier allocation in a distributed manner. The update strategies of all base stations will be continuously broadcast until the system reaches a stable state or the upper limit of the number of iterations.

[0116] This example verifies the algorithm's performance through simulation experiments. It is assumed that... Each user is Each base station provides service. The path loss varies with distance. ,in The distance from the user to the base station is expressed in kilometers. Shadow fading is represented by a log-normal random variable with a mean of zero and a variance of 8 dB. The small-scale fading coefficients follow a Rayleigh distribution with unit variance.

[0117] Experiment (A): The changing trends of system queue backlog and long-term average transmit power as the number of scheduling time slots increases.

[0118] Experiment (B): Under different system parameters, the maximum transmission delay of the task is increased from 3 to 10, and its performance is compared with that of existing similar algorithms.

[0119] The experimental parameters are shown in Table 1.

[0120] Table 1

[0121]

[0122] This example will consider two performance metrics, namely:

[0123] 1. The changes in system queue backlog and long-term average transmit power under different parameters as the number of transmission time slots increases.

[0124] 2. Performance differences in time-averaged transmit power between algorithms with similar parameters.

[0125] The simulation results of this example are as follows:

[0126] 1. The proposed method changes with the number of transmission time slots under different parameters regarding the degree of system queue backlog and time-averaged transmit power.

[0127] like Figure 3 As shown, the total queue length increases rapidly in the initial stage, then converges smoothly to a stable state. This phenomenon indicates that RA-MBC possesses good dynamic resource scheduling capabilities and can effectively maintain network stability. Furthermore, the queue stability value is related to the parameters... There is a positive correlation. This is because the larger... It will place more emphasis on reducing transmission power consumption, thereby reducing transmission power and data throughput, leading to a rapid accumulation of queue length.

[0128] exist Figure 4In the initial stage of scheduling, the time-averaged transmit power shows an upward trend, then gradually converges to a stable value. This is because, in the initial stage of system operation, the base stations have not yet formed a stable cooperative scheduling strategy, and to ensure timely transmission of initial data, the system tends to increase transmit power. As the scheduling process continues, the interference structure between base stations tends to stabilize, and the RA-MBC scheme proposed in this invention gradually establishes a stable cooperative strategy.

[0129] 2. Comparison of time-averaged transmit power with similar algorithms under different parameters

[0130] like Figure 5 As shown, with the relaxation of the maximum tolerable transmission delay, the time-averaged total transmit power of each algorithm shows a decreasing trend, and the rate of decrease gradually slows down. This is because: firstly, the amount of data to be transmitted per time slot decreases as the delay is relaxed; secondly, the increase in the delay upper limit also promotes data queue accumulation, thereby increasing the required transmit power to some extent. Furthermore, the figure further demonstrates the advantages of radiation-sensing-based cooperative transmission design in power control.

[0131] In addition, this application also provides a radiation-sensing multi-cell cooperative transmission system supporting sustainable URLLC services to implement the above method. The system includes a model building module, a long-term optimization problem building module, an online problem dimensionality reduction and transformation module, a global matching module, and a distributed resource solving and transmission module.

[0132] Specifically, the model building module is used to build system models, including multi-cell downlink communication models, task queue models, and radiation footprint models.

[0133] The long-term optimization problem construction module is used to construct a long-term stochastic optimization problem for multiple time slots based on the system model, with the goal of minimizing the long-term average transmit power of the system. The optimization problem includes constraints on the maximum transmit power of the base station, the unique access constraint between the user and the subcarrier, the radiated power constraint in the idle area, the queue stability constraint, and the transmission rate constraint.

[0134] The online problem dimensionality reduction and transformation module is used to introduce a virtual queue to handle the queue stability constraints, and to transform the long-term stochastic optimization problem into a single-slot deterministic optimization sub-model within each scheduling cycle.

[0135] The global matching module is used to determine the association between each cellular user and the multi-antenna base station through centralized global cooperative matching based on the single-slot deterministic optimization sub-model.

[0136] The distributed resource solving and transmission module is used to independently solve its own beamforming and subcarrier allocation strategy in a distributed manner according to the association relationship, and dynamically repeats the execution in a continuous scheduling cycle to achieve multi-cell collaborative data transmission.

[0137] Therefore, the technical solution of this application achieves source-end suppression of cross-cell interference by constructing adjustable radiation characteristics in the spatial domain, and reduces the computational and backhaul burden caused by centralized optimization based on a distributed resource scheduling mechanism. At the same time, by utilizing a dynamic optimization framework, the system can maintain long-term queue stability and controlled energy consumption when facing random service arrivals and time-varying channel conditions, thereby improving the overall network sustainability and operational efficiency while ensuring URLLC service quality.

[0138] In addition to the above embodiments, the present invention may have other implementation methods; all technical solutions formed by equivalent substitution or equivalent transformation fall within the protection scope claimed by the present invention.

Claims

1. A radiation-aware multi-cell cooperative transmission method supporting sustainable URLLC services, characterized in that: Includes the following steps: Step S10: Construct the system model, including: A multi-cell downlink communication model is constructed, which includes a cloud server, multiple multi-antenna base stations and multiple cellular users with URLLC requirements, and the downlink transmission from the multi-antenna base stations to the cellular users is based on multiple orthogonal subcarriers; Construct a task queue model that dynamically updates the queue backlog status based on the amount of user data arriving and the actual amount of data leaving. A radiation footprint model is constructed, which discretizes the total service area into multiple sub-regions, extracts idle sub-regions that do not carry active users, and evaluates the expected radiation signal power in the idle sub-regions based on channel statistical characteristics. Step S20: Based on the system model, with the goal of minimizing the long-term average transmit power of the system, a long-term stochastic optimization problem oriented towards multiple time slots is constructed. The optimization problem includes the maximum transmit power constraint of the base station, the unique access constraint of the user-subcarrier, the radiated power constraint of the idle area, the queue stability constraint, and the transmission rate constraint. Step S30: Introduce a virtual queue to handle the queue stability constraint, and transform the long-term stochastic optimization problem into a single-slot deterministic optimization sub-model within each scheduling cycle; Step S40: Based on the single-slot deterministic optimization sub-model, determine the association between each cellular user and the multi-antenna base station through centralized global cooperative matching; Step S50: Each of the multi-antenna base stations independently solves its own beamforming and subcarrier allocation strategy in a distributed manner according to the association relationship, and dynamically repeats the execution within a continuous scheduling period to achieve multi-cell collaborative data transmission.

2. The radiation-sensing multi-cell cooperative transmission method supporting sustainable URLLC services according to claim 1, characterized in that: The construction of the multi-cell downlink communication model in step S10 includes: The total number of base stations in a multi-cell downlink communication system is defined as follows: The total number of users with URLLC requirements is The number of antennas in each base station is The number of orthogonal carriers in each cell is , define the first The first multi-antenna base station The number of subcarriers is ,make This represents the channel coefficient of user k on this subcarrier during the scheduling period t, and the value of user k in the t-th subcarrier. The received signal on each subcarrier is represented as follows: ,in, This indicates the number of subcarriers within the scheduling period t. The beamforming vector from the base station to the user, and These represent the transmitted signal and additive white Gaussian noise, respectively. Set binary constraints for base station selection and subcarrier allocation The corresponding signal-to-noise ratio is: ,in, This indicates that user terminal k is on a subcarrier The total interference power received by the device; Under the condition of finite block length, user terminal k within the scheduling period t, subcarrier The transmission rate on it is: ,in, Used to characterize channel dispersion, the function Q(·) represents the Gaussian Q-function. Its inverse function, The channel block length, Let be the decoding error probability of user terminal k.

3. The radiation-sensing multi-cell cooperative transmission method supporting sustainable URLLC services according to claim 2, characterized in that: The step S10 of constructing the task queue model includes: The service load requested by the user terminal enters its corresponding cache queue in a first-in-first-out manner, forming a user data queue. The user data queue This represents the amount of data waiting to be served at user terminal k at time t; Each user's data queue is initially empty, and at time... Data queue backlog for user k The update relationship is as follows: ,in, This represents the amount of data actually transmitted and sent by user terminal k within the scheduling period t. This represents the amount of data generated by user terminal k within the scheduling period t; The queue stability constraint satisfies: ,in, This represents the statistical expectation relative to random channel conditions and the arrival process of the task.

4. The radiation-sensing multi-cell cooperative transmission method supporting sustainable URLLC services according to claim 3, characterized in that: The step S10 of constructing the radiation footprint model includes: The total service area is divided into contiguous rectangular sub-regions, and an index set for all sub-regions is created. Within the scheduling period t, a set of sub-region indexes containing the active cellular users is determined based on their location coordinates. , the index set With the set The difference set is defined as the set of free sub-regions that do not carry cellular users. ,Right now ; For any set of free subregions subregion Multi-antenna base stations To the sub-region In subcarrier The propagation channel model is as follows: ,in, Indicates the guide vector. express A complex vector of dimension 1 Here, A is the small-scale fading coefficient, and A is the path loss constant. For log-normal shadowing fading variables, The path decay exponent. and The multi-antenna base stations are respectively and the sub-region The coordinates of the center point; Based on the statistical characteristics of the propagation channel from the multi-antenna base station to the idle sub-region, the expected value of the radiated signal power generated by each multi-antenna base station in the idle sub-region is calculated as follows: ,in, Represents the statistical signal coefficients. To The conjugate transpose of . Indicates subcarrier From base station To users The beamforming vector.

5. The radiation-sensing multi-cell cooperative transmission method supporting sustainable URLLC services according to claim 4, characterized in that: In step S20, the long-term stochastic optimization problem for multiple time slots is constructed with the objective of minimizing the system's long-term average transmit power as the objective. Based on the premise of jointly optimizing base station selection, subcarrier allocation, and transmit beamforming vector, an objective function is constructed to minimize the long-term average transmit power of the system: ,in, The total transmit power of all base stations during the scheduling period t is expressed as: ; Define a set of joint constraints, the set of constraints including: The user-subcarrier unique access constraint is used to limit a single cellular user to access at most a single subcarrier of a single multi-antenna base station, and its expression is: ; The maximum transmit power constraint of the base station is used to limit the transmit power of each of the multi-antenna base stations to no more than its corresponding maximum available power. Its expression is: ; The idle area radiated power constraint is used to limit the set of idle sub-regions that do not carry the cellular users. any subregion within Its cumulative signal radiated power is expected to be no greater than a specified threshold. Its expression is: ; The transmission rate constraint is used to ensure that the amount of task data requested by a new request meets the data transmission requirements within the specified scheduling period, and its expression is: ,in, For users At any moment The maximum number of transmission time slots allowed for the task. Indicates the start time of the transmission for this task. Indicates user terminal At any moment The amount of data generated Indicates user terminal At any moment The amount of data transferred internally, and ; The queue stability constraint is used to determine the actual queue backlog in the communication network. To keep it within the average threshold, the expression is: ,in, This is a preset threshold for the cumulative queue size.

6. The radiation-sensing multi-cell cooperative transmission method supporting sustainable URLLC services according to claim 5, characterized in that: In step S30, a virtual queue is introduced to handle the queue stability constraint, transforming the long-term stochastic optimization problem into a single-slot deterministic optimization sub-model within each scheduling cycle, including: remember For the virtual power queue introduced for user k at time t under long-term average constraints, the virtual queue at the next time step is: ; The quadratic Lyapunov function for measuring the size of the virtual queue is constructed as follows: , where, definition The concatenated vector accumulated for the total queue; Calculate the expected change of the Lyapunov function over a scheduling period as the Lyapunov drift. : and introduce control parameters To drift in Lyapunov With the objective function transmit power By weighing the trade-offs, the upper bound expression that needs to be minimized in each scheduling period t is derived as follows: ,in, It is a non-negative constant. These are control parameters used to make trade-offs between drift and target. This is a constant term under scheduling period t; Add to the constraints of the optimization problem The max function in the upper bound expression is eliminated, and based on the dynamic changes in queue length and environment, the transmission rate constraint spanning multiple scheduling cycles is adaptively decomposed into a minimum throughput constraint for a single cycle. : ; By removing the constant term from the upper bound expression and minimizing the remaining deterministic objective function, the original stochastic optimization problem is transformed into a continuous single-slot online deterministic optimization sub-model, whose objective function expression is: And the constraint is And the maximum transmit power constraint, subcarrier allocation constraint, and idle area radiated power constraint of the base station. This represents the control parameter V used to balance drift and target, and the total transmit power of all base stations during the scheduling period t. The product of.

7. The radiation-sensing multi-cell cooperative transmission method supporting sustainable URLLC services according to claim 6, characterized in that: In step S40, determining the association between each cellular user and the multi-antenna base station through centralized global cooperative matching based on the single-slot deterministic optimization sub-model includes: The symbolic representation of time t is omitted, and the binary correlation indicator variable is defined as follows: The Used to characterize cellular user k and multi-antenna base station Given the association states between them, and with the matching constraint of dividing all cellular users into L non-overlapping user groups, the base station selection objective function aimed at minimizing the overall system cost is constructed as follows: ,in, The virtual power queue introduced for user k For users Data queue backlog For user terminals In subcarrier The transmission rate on ; Construct a cluster utility function that reflects the tradeoff between power consumption and data rate. Based on this, a transitive strict preference order is established for the cellular user for the multi-antenna base station, and an association switching determination criterion is set: cellular user k switches from the multi-antenna base station if and only if... service cluster Switch to multi-antenna base station service cluster Subsequently, when the total utility of the two service clusters involved in the handover is significantly greater than the total utility before the handover, the cellular user k triggers an association switching operation, and the determination criterion satisfies: ; Based on the preset upper limit of the switching ratio Limit the number of cellular users participating in association switching, and obtain the maximum statistical channel gain from each cellular user to all multi-antenna base stations. With minimum statistical channel gain Calculate the absolute value of the corresponding statistical channel gain difference. Sort them in ascending order and extract the first few from the ascending sort. A set of cellular user plans is generated for each element, and the associated exchange is iteratively executed within the set of cellular user plans according to the associated exchange judgment criteria until the matching is stable, where K is the total number of cellular users.

8. The radiation-sensing multi-cell cooperative transmission method supporting sustainable URLLC services according to claim 1, characterized in that: In step S50, each of the multi-antenna base stations independently solves its own beamforming and subcarrier allocation strategy in a distributed manner according to the association relationship, including: Based on a non-convex optimization problem involving a composite objective function and discrete scheduling constraints, a compact lower bound convex function of the logarithmic rate function is constructed using a continuous convex approximation algorithm. A first auxiliary variable is introduced using a quadratic transformation method to construct an upper bound convex function of the logarithmic rate function. A substitution function is established through successive approximations to achieve affine form transformation, and a second auxiliary variable is introduced using a difference method to construct the lower bound of the substitution function. Non-smoothness in subcarrier allocation constraints Norm terms are introduced to establish a smooth approximation model using a continuous smooth function. Based on the inherent non-convexity of the smooth function, a quadratic upper bound function containing a dynamic penalty coefficient is constructed to equivalently define the smooth function. The original objective function, transmission rate constraint, and subcarrier allocation constraint in the non-convex optimization problem are replaced with the derived compact lower bound convex function, upper bound convex function, and quadratic upper bound function, respectively. This reconstructs a convex optimization problem in which the objective function is separable convex and all constraints are convex and quadratic in structure. The interior point method is then used to iteratively solve the convex optimization problem.

9. The radiation-sensing multi-cell cooperative transmission method supporting sustainable URLLC services according to claim 8, characterized in that: The expression for calculating the first auxiliary variable is: ; The expression for calculating the second auxiliary variable is: ; The introduction of a continuous smooth function to establish a smooth approximation model includes: using a logarithmic smooth function to establish a smooth approximation model, which is equivalently defined as follows: ,in, To control the non-negativity of the approximation accuracy of smoothness; The expression for the upper bound function of the quadratic form is: ,in, , , .

10. A radiation-aware multi-cell cooperative transmission system supporting sustainable URLLC services, for implementing the method as described in any one of claims 1 to 9, characterized in that: include: The model building module is used to build system models, including multi-cell downlink communication models, task queue models, and radiation footprint models. The long-term optimization problem construction module is used to construct a long-term stochastic optimization problem for multiple time slots based on the system model, with the goal of minimizing the long-term average transmit power of the system. The optimization problem includes constraints such as the maximum transmit power of the base station, the unique access constraint between the user and the subcarrier, the radiated power constraint in the idle area, the queue stability constraint, and the transmission rate constraint. The online problem dimensionality reduction and transformation module is used to introduce a virtual queue to handle the queue stability constraints and transform the long-term stochastic optimization problem into a single-slot deterministic optimization sub-model within each scheduling cycle. The global matching module is used to determine the association between each cellular user and the multi-antenna base station through centralized global cooperative matching based on the single-slot deterministic optimization sub-model. The distributed resource solving and transmission module is used to independently solve its own beamforming and subcarrier allocation strategy in a distributed manner according to the association relationship, and dynamically repeat the execution in a continuous scheduling period to realize multi-cell collaborative data transmission.

Images