Serial interference cancellation sequence configuration method and device based on probability iterative optimization

Abstract

The invention discloses a serial interference cancellation sequence configuration method and device based on probability iterative optimization, and relates to the technical field of wireless communication, and the method comprises the following steps: initializing a device set and algorithm parameters; discretizing decoding positions in the same frequency spectrum through binary coding, and compressing search dimensions; generating a candidate SIC sequence sample set based on the adaptive probability distribution; performing mapping and conflict processing on the samples, and screening feasible samples; evaluating the performance of a feasible sample by using a complex system model, and mapping the performance into a single evaluation scalar; screening elite samples and iteratively updating probability distribution; and judging probability distribution convergence, and outputting an optimal SIC sequence. Through a probability-driven sample generation and conflict processing mechanism, the method does not depend on an explicit mathematical model, can adapt to a complex system model, remarkably reduces the calculation complexity, meets the parallel optimization requirements in a multi-device and multi-spectrum-cluster scene, and further reduces the calculation complexity.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, specifically to a serial interference cancellation sequence configuration method and apparatus based on probabilistic iterative optimization. Background Technology

[0002] Non-orthogonal multiple access (NOMA) technology, as a key candidate technology for fifth-generation (5G) and future wireless communication systems, significantly improves spectral efficiency by allowing multiple users to transmit on the same spectrum resources and utilizing serial interference cancellation (SIC) technology at the receiver for multi-user detection. In NOMA systems, the channel conditions of different user signals differ, and the SIC receiver decodes and cancels user signals sequentially according to a specific decoding order. Optimizing the uplink decoding order is crucial for minimizing the total system transmit power, ensuring user fairness, and improving overall system performance. Traditional SIC order optimization methods mainly rely on enumeration methods or gradient-based optimization algorithms. Enumeration methods search for the global optimum by exhaustively searching all possible decoding orders; however, as the number of users in the system increases, the number of possible decoding orders grows exponentially, leading to a sharp increase in computational complexity, making it difficult to meet the real-time requirements of multi-user, multi-spectrum cluster scenarios. On the other hand, while gradient-based optimization algorithms are computationally efficient, their performance heavily depends on the explicit mathematical expression of the objective function or system model. In practical communication systems, due to complex factors such as channel estimation errors, hardware nonlinearity, and multi-domain coupling, the system model is often a "black box" that is difficult to model accurately. This makes traditional gradient optimization methods difficult to apply due to the lack of differentiable models, and they may even fail to converge to a satisfactory solution. Therefore, existing technologies face the dual challenges of high computational complexity and reliance on explicit mathematical models when dealing with the SIC sequence optimization problem in NOMA systems. They struggle to meet the low-power consumption and high real-time transmission requirements of complex black-box systems while ensuring global optimality. Summary of the Invention

[0003] To address the shortcomings mentioned in the background art, the present invention aims to provide a serial interference cancellation sequence configuration method and apparatus based on probabilistic iterative optimization.

[0004] Firstly, the objective of this invention can be achieved through the following technical solution: a serial interference cancellation sequence configuration method based on probabilistic iterative optimization, the method comprising the following steps: Receive NOMA system initialization parameters, wherein the NOMA system initialization parameters include the set of devices in the same spectrum NOMA group, the number of binary encoding bits, the probability distribution parameter matrix, the candidate sample size, the number of elite samples, the parameter smoothing update factor, and the convergence judgment threshold. Multiple sets of independent binary sequence samples of candidate SIC order are generated based on the probability distribution parameter matrix. Decoding position conflict detection is performed based on the independent binary sequence samples of multiple sets of candidate SIC order, and valid binary sequence samples are selected. The effective binary sequence samples are input into a pre-established complex black-box simulation model. The overall system performance of the effective binary sequence samples is evaluated based on a multi-dimensional evaluation function to obtain a single-dimensional performance evaluation scalar. The number of elite samples is filtered based on a single-dimensional performance evaluation scalar to obtain an elite sample set. The probability distribution parameter matrix is ​​updated based on the elite sample set. The convergence status is judged based on the updated probability distribution parameter matrix. If convergence is not achieved, multiple independent binary sequence samples of candidate SIC order are generated again based on the updated probability distribution parameter matrix, and the probability distribution parameter matrix is ​​updated again. After convergence is achieved, the optimal SIC order is output.

[0005] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: setting the NOMA system initialization parameters, including: Initialize the same-spectrum NOMA group containing The system comprises several wireless devices, including IoT sensing terminals and mobile user equipment, all sharing the same frequency band spectrum resources; and determines the number of binary encoding bits for the decoding priority position of each device. and satisfy ,initialization Wienouri probability distribution parameter matrix Each element in the matrix Representing the The device code The probability of a bit being 1; setting algorithm parameters: candidate sample size for each iteration. Elite sample size Parameter smoothing update factor Convergence threshold .

[0006] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the process of generating multiple sets of independent binary sequence samples of candidate SIC orders based on the probability distribution parameter matrix, comprising: Based on the probability distribution of the current iteration round, the binary parameter bit probability set For each binary bit of the SIC binary code Perform Bernoulli sampling: with probability Will Set to 1, with probability Set to 0, and finally generate Group of independent binary sequence samples Each set of samples corresponds to a decoding order of a NOMA group; the binary sequence is converted into the actual SIC order of the NOMA group through the discretization mapping rules defined by the SIC encoder mapper. For the actual SIC order of any device, the SIC encoder mapper follows the following formula for coordinate mapping. MERGEFORMAT (1).

[0007] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the process of detecting decoding position conflicts based on independent binary sequence samples of multiple sets of candidate SIC orders, comprising: For the generated Group SIC sequential samples, remove feasible samples with duplicate decoding positions, that is, if the sample contains If the sample fails, delete it; otherwise, keep the remaining feasible samples. Group samples are used as valid binary sequence samples.

[0008] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the process of inputting the valid binary sequence sample into a pre-established complex black-box simulation model, comprising: The pre-established complex black-box simulation model receives the candidate SIC order. As input, a multi-domain coupled model of environmental simulation is used to output a set of real-time performance indicators of the system at that location. The multi-dimensional evaluation function analyzes the black-box output. We extract key performance indicators directly related to the optimization objective and map them to a single scalar evaluation value. As follows: MERGEFORMAT (2).

[0009] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the process of filtering the number of elite samples based on a single-dimensional performance evaluation scalar to obtain an elite sample set, including: Elite sample selection driven by probability distribution: generated from the current iteration round From the candidate location samples, the location with the best system evaluation value is selected based on importance sampling and distribution matching. Group as elite sample set , representing the best-performing SIC order candidate under the current probability distribution; Iterative Update of Probability Distribution: Based on the minimization of the Kullback-Leibler divergence criterion of the probability distribution optimization method, and by maximizing the likelihood probability of elite samples, the probability of each bit of the random SIC sample generator is updated according to the following formula based on the binary bit values ​​of the elite samples. ; MERGEFORMAT (3) Smoothing factor Used to control the weight balance between historical probability distribution and elite sample statistics.

[0010] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the process of determining the convergence state based on the updated probability distribution parameter matrix, comprising: Calculate the maximum change in the probability distribution after the update. If it is less than the preset threshold If the probability distribution has converged, the iteration is terminated and the current optimal deployment position is output; otherwise, the probability distribution has not converged. The sample generation and update process is repeated, serving as the input parameter for the next iteration; after the algorithm converges, the randomly generated optimal binary sequence is used... The optimal SIC order is calculated by inputting the SIC encoder mapper.

[0011] Secondly, in order to achieve the above objectives, the present invention discloses a serial interference cancellation sequence configuration device based on probabilistic iterative optimization, comprising: The parameter acquisition module is used to receive NOMA system initialization parameters, wherein the NOMA system initialization parameters include the set of devices in the same spectrum NOMA group, the number of binary encoding bits, the probability distribution parameter matrix, the candidate sample size, the number of elite samples, the parameter smoothing update factor, and the convergence judgment threshold. The sample filtering module is used to generate multiple sets of independent binary sequence samples of candidate SIC order based on the probability distribution parameter matrix, perform decoding position conflict detection based on the independent binary sequence samples of multiple sets of candidate SIC order, and filter to obtain valid binary sequence samples. The performance evaluation module is used to input valid binary sequence samples into a pre-established complex black-box simulation model, evaluate the overall system performance of the valid binary sequence samples based on a multi-dimensional evaluation function, and obtain a single-dimensional performance evaluation scalar. The optimal order configuration module is used to filter the number of elite samples based on a single-dimensional performance evaluation scalar to obtain an elite sample set. The probability distribution parameter matrix is ​​updated based on the elite sample set. The convergence status is determined based on the updated probability distribution parameter matrix. If convergence has not been achieved, multiple independent binary sequence samples of candidate SIC orders are generated again based on the updated probability distribution parameter matrix, and the probability distribution parameter matrix is ​​updated again. After convergence is achieved, the optimal SIC order is output.

[0012] In another aspect of the present invention, in order to achieve the above-mentioned objective, a terminal device is disclosed, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. The memory stores the computer program capable of running on the processor, and when the processor loads and executes the computer program, it employs the serial interference cancellation sequence configuration method based on probabilistic iterative optimization as described above.

[0013] In another aspect of the present invention, in order to achieve the above objective, a computer-readable storage medium is disclosed, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is loaded and executed by a processor, the serial interference cancellation sequence configuration method based on probabilistic iterative optimization as described above is employed.

[0014] The beneficial effects of this invention are: This invention utilizes a probability-driven sample generation and conflict resolution mechanism, which does not rely on explicit mathematical models. It can adapt to complex system models, significantly reduce computational complexity, and meet the parallel optimization requirements in multi-device and multi-spectral cluster scenarios, further reducing computational complexity. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2 This is a schematic diagram of the nonorthogonal multiple access interference cancellation order optimization method based on distributed probability update of the present invention; Figure 3 This is a schematic diagram of the SIC sequential sample generation process based on sample generation probability according to the present invention; Figure 4 This is a schematic diagram of the sample generation and parameter update process based on parallel processing in this invention; Figure 5 This is a specific test convergence diagram of the present invention; Figure 6 This is a schematic diagram of the device structure of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Example 1: like Figure 1 As shown, the sequential configuration method for eliminating serial interference based on probabilistic iterative optimization includes the following steps: S101: Receive NOMA system initialization parameters, wherein the NOMA system initialization parameters include the set of devices in the same spectrum NOMA group, the number of binary encoding bits, the probability distribution parameter matrix, the candidate sample size, the number of elite samples, the parameter smoothing update factor, and the convergence judgment threshold. The NOMA system initialization parameters are set, including: Initialize the same-spectrum NOMA group containing The system comprises several wireless devices, including IoT sensing terminals and mobile user equipment, all sharing the same frequency band spectrum resources; and determines the number of binary encoding bits for the decoding priority position of each device. and satisfy ,initialization Wienouri probability distribution parameter matrix Each element in the matrix Representing the The device code The probability of a bit being 1; setting algorithm parameters: candidate sample size for each iteration. Elite sample size Parameter smoothing update factor Convergence threshold .

[0018] S102: Generate multiple sets of independent binary sequence samples of candidate SIC order based on the probability distribution parameter matrix, perform decoding position conflict detection based on the multiple sets of independent binary sequence samples of candidate SIC order, and filter to obtain valid binary sequence samples. The process of generating multiple sets of independent binary sequence samples of candidate SIC orders based on the probability distribution parameter matrix includes: Based on the probability distribution of the current iteration round, the binary parameter bit probability set For each binary bit of the SIC binary code Perform Bernoulli sampling: with probability Will Set to 1, with probability Set to 0, and finally generate Group of independent binary sequence samples Each set of samples corresponds to a decoding order of a NOMA group; the binary sequence is converted into the actual SIC order of the NOMA group through the discretization mapping rules defined by the SIC encoder mapper. For the actual SIC order of any device, the SIC encoder mapper follows the following formula for coordinate mapping. MERGEFORMAT (4).

[0019] any device The SIC sequential encoding mapping is as follows A binary sequence of bits, and all The devices are connected in parallel as The space of discrete networks, i.e. .

[0020] The process of decoding position collision detection based on independent binary sequence samples from multiple candidate SIC orders includes: For the generated Group SIC sequential samples, remove feasible samples with duplicate decoding positions, that is, if the sample contains If the sample fails, delete it; otherwise, keep the remaining feasible samples. Group samples are used as valid binary sequence samples.

[0021] Specifically, the rules for decoding position conflict detection based on independent binary sequence samples from multiple candidate SIC orders are as follows: For any randomly generated sample If there exist elements at any two positions and have If two devices choose the same decoding order, then that sample is deleted. And retain other valid samples.

[0022] S103: Input the valid binary sequence samples into the pre-established complex black-box simulation model, evaluate the overall system performance of the valid binary sequence samples based on the multi-dimensional evaluation function, and obtain a single-dimensional performance evaluation scalar. The process of inputting valid binary sequence samples into a pre-established complex black-box simulation model includes: The pre-established complex black-box simulation model receives the candidate SIC order. As input, a multi-domain coupled model of environmental simulation is used to output a set of real-time performance indicators of the system at that location. The multi-dimensional evaluation function analyzes the black-box output. We extract key performance indicators directly related to the optimization objective and map them to a single scalar evaluation value. As follows: MERGEFORMAT (5).

[0023] The specific implementation process of the pre-established complex black-box simulation model and multi-dimensional evaluation function includes: The input to the black box is the SIC decoding order of any sample; the black box internally loads a set of environmental parameter vectors describing the current communication scenario in real time. This vector mainly includes channel state information, system configuration parameters, network dynamic parameters, etc. The black box internally pre-defines a parameterized evolution function. This function will decode the input in the correct order. and current environmental parameters and adjustable system parameters As a black-box system variable. Function The core of this system consists of a series of communication theory models, such as energy consumption models, transmission delay models, weighted rate models, etc., used to calculate various performance indicators. (Function) The final output is a performance metric vector. .

[0024] Next, a normalization function is defined for each key performance indicator (such as energy consumption, transmission latency, system throughput, user fairness, etc.). The normalized metric is mapped to the interval [0,1], where 1 represents the optimal value. Next, based on the specific needs and application scenarios of the communication system (e.g., whether it's an energy-efficient IoT scenario or a vehicle-to-everything (V2X) scenario prioritizing low latency and high reliability), a corresponding weight coefficient is assigned to each normalized metric. These weights reflect the importance of different performance dimensions in the overall evaluation. Finally, a weighted summation formula (e.g., overall score = w1 × normalized throughput + w2 × (1 - normalized latency) + w3 × (1 - normalized energy consumption) + ...) is used to calculate the overall score for each dimension. The higher this score, the better the overall performance of the adopted SIC decoding order under the current communication environment parameters.

[0025] Note that this invention does not have specific requirements for the black box and the evaluation function. Since the optimization method of this invention is based on probabilistic updates, the SIC decoding order can be optimized for any simulable system. Moreover, the more complex the system, the lower the complexity of the method of this invention.

[0026] S104: Based on a single-dimensional performance evaluation scalar, the number of elite samples is filtered to obtain an elite sample set. The probability distribution parameter matrix is ​​updated based on the elite sample set. The convergence status is determined based on the updated probability distribution parameter matrix. If convergence is not achieved, multiple independent binary sequence samples of candidate SIC orders are generated again based on the updated probability distribution parameter matrix, and the probability distribution parameter matrix is ​​updated again. After convergence is achieved, the optimal SIC order is output.

[0027] The process of filtering the number of elite samples based on a single-dimensional performance evaluation scalar to obtain the elite sample set includes: Elite sample selection driven by probability distribution: generated from the current iteration round From the candidate location samples, the location with the best system evaluation value is selected based on importance sampling and distribution matching. Group as elite sample set , representing the best-performing SIC order candidate under the current probability distribution; Iterative Update of Probability Distribution: Based on the minimization of the Kullback-Leibler divergence criterion of the probability distribution optimization method, and by maximizing the likelihood probability of elite samples, the probability of each bit of the random SIC sample generator is updated according to the following formula based on the binary bit values ​​of the elite samples. ; MERGEFORMAT (6) Smoothing factor Used to control the weight balance between historical probability distribution and elite sample statistics.

[0028] The process of determining the convergence state based on the updated probability distribution parameter matrix includes: Calculate the maximum change in the probability distribution after the update. If it is less than the preset threshold If the probability distribution has converged, the iteration is terminated and the current optimal deployment position is output; otherwise, the probability distribution has not converged. The sample generation and update process is repeated, serving as the input parameter for the next iteration; after the algorithm converges, the randomly generated optimal binary sequence is used... The optimal SIC order is calculated by inputting the SIC encoder mapper.

[0029] Specifically, the present invention will be further illustrated below through embodiments: (1) System Initialization: This step sets the basic framework for the entire optimization process. First, the communication scenario unit to be optimized is identified, namely a co-spectrum NOMA user group, which contains a group consisting of... A set consisting of wireless devices (or users) is denoted as . Next, we will perform crucial parameter settings: for... The decoding sequence of each device is encoded to determine the required number of binary code bits. This operation aims to compress the search space. Then, the core engine of the algorithm—the Bernoulli probability parameter matrix—is initialized. This is OK A matrix of columns, where each element Representative equipment The binary encoding of the first The initial probability is 1. To ensure the globality of the search, all The initial values ​​are all set to 0.5. Finally, configure the algorithm's running parameters: total number of candidate samples. (Number of schemes generated per generation), number of elite samples (The number of excellent solutions used to guide the search direction) Smooth update factor (Control update speed, Convergence threshold (Conditions for determining the termination of the algorithm).

[0030] (2) Search Space Compression and Binary Encoding: The core of this step is to compress the search space of a naturally discrete sorting problem (where the search space is...). This can be transformed into a more manageable binary optimization problem. Specifically, it involves: [details about the device set]. Each device in Assign a unique identifier, which is a string of length 1. The binary sequence is represented as Each of them It is a binary variable that takes the value 0 or 1. Then, all... The binary identifiers of each device are concatenated sequentially to form a complete, dimensionless array. joint binary vector This vector represents an encoding scheme for the decoding order, thus mapping the complex permutation and combination problem into a structured binary space.

[0031] (3) Probability-driven candidate sample generation: This step is one of the core actions of probability distribution optimization, aiming to generate a batch of candidate solutions based on the current probability distribution. For example... Figure 2 As shown, this process can be intuitively understood as follows: the table on the left represents the current probability parameter matrix. It is a The grid consists of cells, each containing a probability value. The sample generator generates a sample for each probability parameter. Perform an independent Bernoulli trial (think of it as a biased cointoss) to The probability of producing "1" is... The probability of this is "0". Through this bit-by-bit sampling, a value is generated for each device. Bit-by-bit binary encoding ultimately forms a complete binary sample (i.e.) Figure 2 (A sample column on the right). Repeat this process. You can get it in one go. There are 10 independent candidate samples. Each sample is a binary representation of a possible SIC order.

[0032] (4) Feasible Sample Filtering and Conflict Handling: Since the sampling in step (3) is independent, conflicts may occur when the generated binary samples are mapped back to the actual decoding order (i.e., multiple devices are mapped to the same sequential position). This step is responsible for filtering infeasible solutions. The specific operation is: First, for each sample, each device... of The binary code is obtained through the formula. MERGEFORMAT (7) Convert to a decimal integer This integer represents the device. The candidate decoding positions (priority, with smaller numbers indicating higher priority) are then examined. Next, the candidate decoding position sets for all devices in this sample are checked. If there are any two different devices and satisfy If a sample fails to pass the collision detection, it is determined to have a "decoding position conflict" and is an invalid SIC sequence (because only one user's signal can be decoded at a time), and must be discarded. All samples that pass the collision detection are retained as feasible samples and proceed to the next evaluation step.

[0033] (5) Black-box system performance evaluation and scalarization: This step is responsible for evaluating the performance of each feasible SIC sequence. The feasible sequence obtained in step (4) (i.e., a set of...) This signal is used as input and passed to a "complex black box system." This system is typically a high-fidelity communication link-level or system-level simulation platform that integrates complex channel models, signal processing flows, and SiC receiver models. It is capable of simulating real-world environments and outputting a series of system performance observations. (For example, the system's total transmit power, overall throughput, user fairness index, etc.). Then, the "evaluation function". It will analyze these observations and aggregate them into a single scalar evaluation value based on the optimization objective (such as minimizing total power). (For example, = -Total power, so that a larger value indicates better performance). This scalar value... This uniquely characterizes the quality of the SIC sequence.

[0034] (6) Elite Sample Selection and Probability Distribution Update: This step is the learning and guiding stage for probability distribution optimization. First, elite sample selection: based on the evaluation value calculated in step (5). Sort all feasible samples and select the top performers. These samples form an "elite sample set." These samples represent the most promising solutions that can be found under the current probability distribution. Then, the probability distribution is updated: based on this elite sample set, the probability parameter matrix is ​​updated. The updated formula is: (3) This formula means that the new probability value is a weighted average of the old probability value and the statistical frequency of that bit being "1" in the elite sample. (Smoothing factor) This is used to balance historical information with new current knowledge. Through this update, the algorithm "learns" the binary bit patterns of elite samples, making the next iteration more inclined to generate new samples that are similar to elite samples and have better performance.

[0035] (7) Convergence Judgment: This step determines whether the optimization process terminates. Calculate the probability parameter matrix. Maximum change before and after this update (5) This change reflects the activity level of the algorithm's learning process. If (Convergence threshold) indicates that the probability distribution has stabilized and the algorithm has found the region of optimal solution, then the process continues to step (8); otherwise, the updated... As input, return to step (3) to begin a new round of the “generate-evaluate-update” cycle.

[0036] (8) Optimal Solution Output: After the algorithm is determined to be convergent in step (7), it enters the output stage. The optimal solution can be derived from the final probability distribution. A sample generated by mid-samples, or an evaluation value from an elite sample set. The highest-performing sample. The binary vector of that optimal sample. The mapping rules in step (4) are converted into the final decoding order. This order is the approximately globally optimal serial interference cancellation order calculated by this optimization method for the current NOMA group. This order can be directly used to configure the uplink transmission SIC receiver in the baseband signal processing unit to achieve efficient interference cancellation and signal recovery.

[0037] like Figure 4 As shown, in the algorithm proposed in this invention, the four parts—SIC sample sequential sample generator, conflict sample remover, communication black-box simulation model, and sample evaluator—can be processed synchronously by a parallel processor. The parallel processing system consists of… The architecture consists of several independently running parallel processing units. Each unit sequentially executes SIC sequential sample generation (e.g., generating a binary sequence 1010), conflict sample removal, black-box simulation, and sample evaluation steps. Finally, an elite sample filter selects elite samples and updates the probability distribution parameters. This architecture reduces the time consumption of a single iteration from... Reduce to This significantly improves the real-time performance of deployment location optimization while ensuring the accuracy of binary encoding and the rigor of multi-objective constraint verification.

[0038] To illustrate the feasibility of this invention, this application uses a randomly generated communication environment (i.e., a black-box system). The performance of this invention was tested under (randomly generated) conditions. All parameter settings are shown in Table 1 below: Table 1 Figure 5 The convergence process in this communication environment is demonstrated, showing that the optimization method proposed in this invention can achieve convergence in just a few dozen iterations, with lower computational complexity.

[0039] Example 2: To achieve the above objective, such as Figure 6 As shown, based on Embodiment 1, this invention discloses a serial interference cancellation sequence configuration device based on probabilistic iterative optimization, comprising: The parameter acquisition module 11 is used to receive NOMA system initialization parameters, wherein the NOMA system initialization parameters include the set of devices in the same spectrum NOMA group, the number of binary encoding bits, the probability distribution parameter matrix, the candidate sample size, the number of elite samples, the parameter smoothing update factor, and the convergence judgment threshold. The sample filtering module 12 is used to generate multiple sets of independent binary sequence samples of candidate SIC order based on the probability distribution parameter matrix, perform decoding position conflict detection based on the independent binary sequence samples of multiple sets of candidate SIC order, and filter to obtain valid binary sequence samples. The performance evaluation module 13 is used to input valid binary sequence samples into a pre-established complex black-box simulation model, evaluate the overall system performance of the valid binary sequence samples based on a multi-dimensional evaluation function, and obtain a single-dimensional performance evaluation scalar. The optimal sequence configuration module 14 is used to filter the number of elite samples based on a single-dimensional performance evaluation scalar to obtain an elite sample set. The probability distribution parameter matrix is ​​updated based on the elite sample set. The convergence status is judged based on the updated probability distribution parameter matrix. If convergence is not achieved, multiple independent binary sequence samples of candidate SIC order are generated again based on the updated probability distribution parameter matrix, and the probability distribution parameter matrix is ​​updated again. After convergence is achieved, the optimal SIC order is output.

[0040] Based on the same inventive concept, this invention also provides a computer device, comprising: one or more processors, and a memory for storing one or more computer programs; the programs include program instructions, and the processor executes the program instructions stored in the memory. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, used to implement one or more instructions, specifically for loading and executing one or more instructions stored in a computer storage medium to implement the above-described method.

[0041] It should be further explained that, based on the same inventive concept, the present invention also provides a computer storage medium storing a computer program, which, when executed by a processor, performs the above-described method. This storage medium can be any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In the present invention, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0042] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0043] The foregoing has shown and described the basic principles, main features, and advantages of this disclosure. Those skilled in the art should understand that this disclosure is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of this disclosure. Various changes and modifications can be made to this disclosure without departing from its spirit and scope, and all such changes and modifications fall within the scope of this disclosure as claimed.

Claims

1. A method for configuring the order of serial interference cancellation based on probability iterative optimization, characterized in that, The method comprises the following steps: Receiving NOMA system initialization parameters, wherein the NOMA system initialization parameters comprise a device set of a same-frequency spectrum NOMA group, a number of binary coded bits, a probability distribution parameter matrix, a candidate sample size, an elite sample number, a parameter smoothing update factor, and a convergence determination threshold; Based on the probability distribution parameter matrix, a plurality of groups of candidate SIC order independent binary sequence samples are generated, decoding position conflict detection is performed based on the plurality of groups of candidate SIC order independent binary sequence samples, and valid binary sequence samples are screened out; The valid binary sequence samples are input into a pre-established complex black box simulation model, the system comprehensive performance of the valid binary sequence samples is evaluated based on a multi-dimensional evaluation function, and a single-dimensional performance evaluation scalar is obtained; Based on the single-dimensional performance evaluation scalar, the elite sample number is screened, an elite sample set is obtained, the probability distribution parameter matrix is updated based on the elite sample set, the convergence state is judged based on the updated probability distribution parameter matrix, if the convergence has not been reached, a plurality of groups of candidate SIC order independent binary sequence samples are generated again based on the updated probability distribution parameter matrix, and the probability distribution parameter matrix is updated again, until the convergence is reached, and the optimal SIC order is output.

2. The method of claim 1, wherein, The setting of the NOMA system initialization parameters comprises: Initialize the same-spectrum NOMA group containing The system comprises several wireless devices, including IoT sensing terminals and mobile user equipment, all sharing the same frequency band spectrum resources; the binary encoding bits for the decoding priority position of each device are determined. and satisfy ,initialization Wienouri probability distribution parameter matrix Each element in the matrix Representing the The device code The probability of a bit being 1; setting algorithm parameters: candidate sample size for each iteration. Elite sample size Parameter smoothing update factor Convergence threshold .

3. The method of claim 1, wherein, The process of generating a plurality of groups of candidate SIC order independent binary sequence samples based on the probability distribution parameter matrix comprises: According to the distribution probability binary parameter bit probability set of the current iteration round , for each binary bit of the SIC binary code , Bernoulli sampling is performed: with probability , set to 1, and with probability , set to 0, finally generating a group of independent binary sequence samples , each group of samples corresponding to the decoding order of a NOMA group; through the discrete mapping rule defined by the SIC encoding mapper, the binary sequence is converted into the actual SIC order of the NOMA group, for the actual SIC order of any device, wherein the SIC encoding mapper follows the following formula for coordinate mapping; \ MERGEFORMAT (1).

4. The method of claim 1, wherein, The process of performing decoding position conflict detection based on the plurality of groups of candidate SIC order independent binary sequence samples comprises: For the generated set of SIC ordered samples, eliminate feasible samples that have decoding position duplicates, i.e. if a sample has , then delete the sample; keep the remaining feasible set of samples as valid binary sequence samples.

5. The method of claim 1, wherein, The process of inputting the valid binary sequence samples into the pre-established complex black box simulation model comprises: Pre-established complex black-box simulation model receives candidate SIC order As input, multi-domain coupled model through environmental simulation, output real-time performance indicator set of system under this position ; multi-dimensional evaluation function extracts key performance indicators directly related to optimization target from analytical black-box output , and maps them into a single scalar evaluation value, as follows: \ MERGEFORMAT (2).

6. The method of claim 1, wherein, The process of screening the elite sample number based on the single-dimensional performance evaluation scalar to obtain the elite sample set comprises: Distribution probability driven elite sample screening: from the current iteration round generated group candidate position samples, according to importance sampling and distribution matching, screen out the system evaluation value optimal group as an elite sample set , representing the optimal SIC order candidate under the current probability distribution; Distribution probability probability distribution iterative update: according to the minimum Kullback-Leibler divergence criterion of the distribution probability optimization method, the binary bit probability of the random SIC sample generator is updated according to the following formula by maximizing the likelihood probability of the elite sample based on the binary bit value of the elite sample ; \ MERGEFORMAT (3) where the smoothing factor For controlling the balance of the weight between the historical probability distribution and the elite sample statistics.

7. The method of claim 1, wherein, The process of judging the convergence state based on the updated probability distribution parameter matrix comprises: Computing the maximum change of the updated probability distribution If less than a preset threshold , the probability distribution is determined to have converged, the iteration is terminated, and the current optimal deployment position is output; otherwise, the probability distribution has not converged, and is taken as the input parameter of the next round of iteration, and the sample generation and updating process is repeated; after the algorithm converges, the randomly generated optimal binary sequence is input into the SIC encoding mapper to calculate the optimal SIC order.

8. Apparatus for configuring order of serial interference cancellation based on probabilistic iterative optimization, employing the method for configuring order of serial interference cancellation based on probabilistic iterative optimization according to any one of claims 1 to 7, characterized in that It comprises: The parameter acquisition module is used for receiving NOMA system initialization parameters, wherein the NOMA system initialization parameters comprise a device set of a same-frequency spectrum NOMA group, a number of binary coded bits, a probability distribution parameter matrix, a candidate sample size, an elite sample number, a parameter smoothing update factor, and a convergence determination threshold; The sample screening module is used for generating a plurality of groups of candidate SIC order independent binary sequence samples based on the probability distribution parameter matrix, performing decoding position conflict detection based on the plurality of groups of candidate SIC order independent binary sequence samples, and screening valid binary sequence samples; The performance evaluation module is used for inputting the valid binary sequence samples into a pre-established complex black box simulation model, evaluating the system comprehensive performance of the valid binary sequence samples based on a multi-dimensional evaluation function, and obtaining a single-dimensional performance evaluation scalar. The sequential optimal configuration module is configured to filter the number of elite samples based on a single dimension performance evaluation scalar to obtain an elite sample set, update a probability distribution parameter matrix based on the elite sample set, judge a convergence state based on the updated probability distribution parameter matrix, if the convergence state is not reached, generate again a plurality of groups of candidate SIC sequential independent binary sequence samples based on the updated probability distribution parameter matrix, and update again the probability distribution parameter matrix until the convergence state is reached, and output an optimal SIC sequence.

9. A terminal device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, The memory stores a computer program capable of running on the processor, and the processor loads and executes the computer program, and adopts the serial interference cancellation order configuration method based on the probability iterative optimization in any one of claims 1 to 7.

10. A computer-readable storage medium having stored therein a computer program, characterized in that, The computer program is loaded and executed by the processor, and adopts the serial interference cancellation order configuration method based on the probability iterative optimization in any one of claims 1 to 7.

Images