A low-orbit satellite internet of things beam management method based on information age perception

Abstract

The application claims a low-orbit satellite Internet of Things (SIoT) beam management method based on information age (AoI) awareness, belonging to the technical field of wireless communication. In view of the problems of uneven time and space distribution of SIoT business, strong interference of multi-satellite coverage overlap, low utilization rate of on-board resources, and difficulty in guaranteeing the freshness of real-time business information, first, the time and space characteristics of business are represented by calculating the rate threshold according to the cell business information, and the AoI evolution model is constructed to quantify the information freshness; combined with the rate threshold and channel gain, a weighted AoI evaluation model is constructed to output the service priority. Then, based on the service priority, an improved genetic algorithm is proposed to realize the beam pattern design, and a weighted AoI beam closing algorithm is proposed to improve the resource utilization rate by closing the low-yield beam. Finally, the successive convex approximation method is used for inter-beam power allocation and rate matching to meet the business rate requirements of different cells, and the SIoT beam management of minimizing AoI is completed, which guarantees the information freshness and improves the on-board resource utilization rate.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology. Specifically, it relates to a low-Earth orbit satellite Internet of Things (IoT) beam management method based on information age perception. Background Technology

[0002] As the scale of IoT business continues to grow, a large number of terminals are being deployed in urban fringe areas, oceans, mountainous regions, and remote areas. Traditional terrestrial communication networks, limited by infrastructure construction costs, geographical environment, and coverage capabilities, struggle to provide stable, continuous, and low-latency connectivity services for these areas. Low-Earth orbit (LEO) satellite IoT, with its advantages of wide coverage, flexible deployment, and low transmission latency, has become an important technological approach to support wide-area real-time sensing and remote control.

[0003] Existing multi-beam satellite systems typically use static continuous illumination to allocate onboard power and bandwidth resources to each coverage area. This approach is ill-suited to the uneven spatiotemporal distribution of IoT services, the instantaneous surge of services in hotspot areas, and the dynamic changes in coverage across multiple satellites. Beam-hopping technology can illuminate different beams in different time slots, enabling on-demand reallocation of onboard resources. However, when multiple adjacent beams are activated simultaneously or when the coverage areas of different satellites overlap, interference between users and between satellites is significantly enhanced.

[0004] Rate Splitting Multiple Access (RSMA) can split messages into public and private parts, enabling flexible interference management by partially decoding and partially treating them as noise. Meanwhile, real-time IoT services prioritize information freshness over pure throughput. Existing research largely focuses on single-satellite scenarios or prioritizes throughput and fairness, failing to adequately consider the dynamic topology of multiple LEO satellites, beam hopping, and the coupling relationship between RSMA transmission and information freshness. Therefore, how to achieve collaborative beam management in a multi-LEO satellite IoT architecture, and design a beam management scheme that balances multi-satellite collaboration, interference suppression, resource utilization, and information freshness, has become a critical issue that urgently needs to be addressed. Summary of the Invention

[0005] This invention aims to solve the problems of the prior art mentioned above. It proposes a low-Earth orbit satellite Internet of Things (IoT) beam management method based on information age perception. The technical solution of this invention is as follows:

[0006] A low-Earth orbit satellite Internet of Things (IoT) beam management method based on information age perception includes the following steps:

[0007] S1: Construct a multi-low-orbit satellite multi-beam satellite IoT scenario that includes multiple low-orbit satellites, fixed Earth cells, IoT terminals, and gateway stations, and establish satellite visibility relationships, beam-cell service relationships, and beam transition cycle models.

[0008] S2: Construct a satellite-to-cell user channel model based on the shadowed Rice channel, and construct an Age of Information (AoI) evolution model and a cell rate threshold model;

[0009] S3: Employs single-layer RSMA downlink multicast transmission, divides each beam message into public messages and private messages, and establishes a model for public and private flow transmission power, public rate matching, and cell reachability rate.

[0010] S4: With the goal of minimizing the average AoI of all fixed Earth cells within a scheduling cycle, construct a collaborative beam management model that jointly optimizes beam hopping patterns, common power allocation, private power allocation, and common rate matching, and set constraints on the maximum number of satellite beams, satellite visibility, total power budget, single beam power budget, and maximum AoI of the cell.

[0011] S5: For the power allocation and common rate matching subproblems under single time slot fixed beam pattern, the successful transmission variables in AoI evolution are continuously relaxed, and the common power, private power and common rate matching are solved by the successive convex approximation method.

[0012] S6: For the beam pattern design subproblem, an improved priority-based genetic algorithm is used for candidate beam pattern search. Local beam state coding is employed. A fitness function and mutation weights are constructed based on the previous time slot AoI, rate satisfaction, channel gain, and rate threshold. Through forced serving cell priority allocation, initial feasible population generation, elite retention, parent pool sampling, satellite-level light / heavy mutation, and deduplication sorting, the output beam pattern is determined. Candidate beam patterns;

[0013] S7: Forward Successive convex approximation (SCA) complex evaluation is performed on each candidate beam pattern, and a beam shut-off algorithm based on weighted AoI is executed on the candidate patterns. The low-weight beams are shut off first, and the low-yield beams are removed if the average AoI does not deteriorate after shutting off.

[0014] S8: Under the time-slot-by-time hierarchical solution framework, execute steps S5 to S7 sequentially according to the time slot order. After completing the solution of the current time slot, update the AoI state and enter the next time slot. Continue until all time slots are solved, and output the beam transition pattern, common power allocation result, private power allocation result and common rate matching result for the entire scheduling cycle.

[0015] Furthermore, the multi-low-Earth orbit satellite multi-beam satellite Internet of Things system model in step S1 specifically includes the following:

[0016] Let the set of LEO satellites and the set of fixed Earth cells be denoted as follows: and ,definition Fixed Earth Community The collection of IoT users, the sum of the number of IoT users in all communities is ;satellite The available beam set is ,definition The maximum number of beams for the satellite, a binary variable. Used to identify beams Whether by satellite generate;

[0017] The beam transition period contains one Time slot, time slot index is binary variables Used to indicate satellites In the time slot Is it visible in the community? This variable is determined by the elevation angle of the cell center; if it is less than the set minimum elevation angle, the cell will not have visible satellites; binary variable. and Indicates in time slot Beam Earth-fixed communities Service relationships and satellites Earth-fixed communities Service relationship; Indicates the community In the The formula for determining whether a time slot is served is as follows:

[0018]

[0019] Furthermore, the channel model and AoI evolution model in step S2 are specifically as follows:

[0020] (1) Channel model:

[0021] Earth Fixed Community The first in IoT users and generated beams In the The channel gain for each time slot is:

[0022]

[0023] in, For receiving antenna gain, The angle between the beam axis and the direction of the receiving antenna, i.e., the offset angle, is a random shaded Rice variable. Obeying the parameter is the direct component Shadow fading component Multipath components The distribution of shadow Rice, For users in the first Each time slot and generated beam satellite The distance between them For carrier wavelength, The specific formula for calculating the transmit antenna gain is as follows:

[0024]

[0025] in, The angle between the beam axis and the direction of the receiving antenna, i.e., the offset angle. It is a first-order Bessel function of the first kind. Where is the antenna radius. For carrier wavelength, The peak gain of the antenna is calculated using the following formula:

[0026]

[0027] Define noise power as The specific expression is:

[0028]

[0029] in, Boltzmann's constant, Noise temperature of the receiving system It is the bandwidth allocated to the user's link.

[0030] (2) AoI evolution model

[0031] The condition for a successful state update is equivalent to a rate threshold constraint. Successfully transferred binary variables. To characterize the result of signal transmission; if the rate is achievable Exceeding the rate threshold Then the first The first of the IoT communities The signal of the IoT user in the first Each time slot can be successfully transmitted, using Indicate; otherwise, ;

[0032] Definition of the first The first of the IoT communities The number of IoT users in the first AoI expression for each time slot:

[0033]

[0034] in, ;

[0035] In the The average AoI of all cells in each time slot is:

[0036]

[0037] Furthermore, the first The average AoI of each IoT cell over the entire scheduling period is:

[0038]

[0039] Furthermore, in step S3, single-layer RSMA downlink multicast transmission is adopted, and the relevant signal model and rate model are as follows:

[0040] The signal model employs single-layer RSMA downlink multicast transmission, dividing all IoT users in each cell into a group. Each time slot A multicast message was sent to In the group of cells illuminated by beams, by The index divides each message into a public part and a private part, i.e. The common parts of each group are merged and encoded into a common stream shared by all groups, i.e. The private portion is independently encoded into a private stream for each group, i.e. ;

[0041] Beam The transmitted signal is ,in and This indicates the beamwidth allocated to the transmission power of public and private messages, respectively, let... Let the transmission power vector of the public message be... Let the transmission power vector of the private message be represented. Then, the transmission power matrices of the public and private messages during the entire scheduling period are defined as follows: and When each user receives a signal, the public message is first decoded, and the private message is treated as interference.

[0042] In the Each time slot, collection users in The signal-to-interference-plus-noise ratio of the public message is:

[0043]

[0044] user The achievable rate for public messages is:

[0045]

[0046] Due to common flow Shared by all beams, to ensure successful decoding for every user. The achievable rate of the system's common flow is:

[0047]

[0048] make ,in Indicates in time slot The contribution rate of each beam to the reachable rate of the system's common flow; the rate matching matrix of the common message during the entire hopping beam period is expressed as follows: Subsequently, the system uses serial interference cancellation to subtract the common stream, each user decodes its corresponding private stream, and treats the private streams of other users as interference;

[0049] Therefore, in the Each time slot, user The signal-to-interference-plus-noise ratio of the private message is:

[0050]

[0051] in Indicates the first Each time slot service cell The corresponding beam, For users with corresponding channel gain of this beam The achievable rate for private messages is:

[0052]

[0053] To ensure that all users within a multicast group can successfully decode their required private signals, beamforming The private message rate is:

[0054]

[0055] No. Each time slot, beam The total rate is:

[0056]

[0057] Earth Fixed Community In the The achievable rate for each time slot is:

[0058]

[0059] Users in each cell have the same actual data rate. Therefore, the AoI evolution model for a cell can be expressed as follows:

[0060]

[0061] in Successful transmission of binary variables in the cell is represented as:

[0062]

[0063] in This is because users within the same cell have the same service, so the cell rate threshold is the same.

[0064] Throughout the scheduling cycle The average AoI for all cells is:

[0065]

[0066] Furthermore, the cooperative beam management model constructed in step S4 aims to minimize the average AoI of all cells throughout the entire scheduling cycle, and is expressed as:

[0067]

[0068] stC1:

[0069] C2:

[0070] C3:

[0071] C4:

[0072] C5:

[0073] C6:

[0074] C7:

[0075] C8:

[0076] C1 represents satellite In the time slot The number of communities served does not exceed C2 indicates that any cell is served by at most one beam in the same time slot; C3 indicates the visibility variable. Service relationship variables , C4 represents binary variables; C4 indicates that the common power, private power, and common rate matching are all non-negative; C5 indicates that the cell AoI does not exceed the preset upper limit. C6 indicates that the total transmission power of a single satellite does not exceed the total power of the satellite. C7 indicates that the sum of common rate matches does not exceed the system's achievable common flow rate. C8 indicates that the transmit power of a single beam does not exceed the beam power limit. ;

[0077] Furthermore, in step S5, the power allocation and common rate matching subproblem under a single-slot fixed beam pattern is solved using a successive convex approximation method, specifically including:

[0078] (1) Rewrite the objective function

[0079] To express the objective function as a linear summation, we first rewrite the piecewise update of equation (17) into a single expression:

[0080]

[0081] Secondly, for The relaxation process was performed, and the specific relaxation results are as follows:

[0082]

[0083] At the same time, a new constraint is added here to ensure the effective utilization of beam resources:

[0084]

[0085] At this point, the original conditions are relaxed to This facilitates subsequent solutions, especially for the relaxed portion. It can be written as the following expression:

[0086]

[0087] (2) SCA solution

[0088] First, we introduce an auxiliary variable. ,in Indicates beam The lower bound of the total rate leads to the following inequality:

[0089]

[0090] according to By definition, the above constraints can be equivalent to:

[0091]

[0092] because Even with the existence of , the above constraints remain non-convex. Therefore, we introduce an auxiliary variable here. and ,in This represents the lower bound of the SINR of a user's private messages, while Therefore, the above constraints can be rewritten as:

[0093]

[0094]

[0095] Then, the SCA method is used to make And order ,but It is a difference-of-convex (DC) function. The first-order Taylor approximation is as follows:

[0096]

[0097] in yes In the The value in the next iteration. Therefore, the above constraint can be approximated as:

[0098]

[0099] Next, taking the natural logarithm of constraint (30) yields its equivalent convex form:

[0100]

[0101] Similarly, for constraint C7, an auxiliary variable is introduced. ,make Indicates user The lower bound of SINR for public messages, C7 can be rewritten as:

[0102]

[0103]

[0104] Similarly, constraint (32) can also be converted into DC form, let and ,but The first-order Taylor approximation is shown below:

[0105]

[0106] Therefore, constraint (34) can be approximated as:

[0107]

[0108] Then, perform the same operation as above on constraint (33):

[0109]

[0110] Finally, constraints C4, C6, C7, (23) can be rewritten as:

[0111]

[0112]

[0113]

[0114]

[0115] Therefore, for a single time slot It can be equivalently converted to :

[0116]

[0117] stC1:

[0118] C2:

[0119] C3:

[0120] C4:

[0121] C5:

[0122] C6:

[0123] C7:

[0124] C8:

[0125] C9:

[0126] Through the above steps Transformation of convex functions The optimal solution for this time slot can be obtained by using the CVX tool to solve for the optimal values ​​of each parameter.

[0127] Furthermore, the priority-based improved genetic algorithm in step S6 includes:

[0128] (1) Encoding and initialization: Let Indicates the first Satellites in each time slot The Local beam service status, among which This indicates that the beam serves a fixed cell on Earth. , This indicates that the beam is off. Based on this, following a fixed order of satellite number and local beam number, the service status of all local beams within a time slot is concatenated into a single entity. , No. Each chromosome corresponds to a complete single-slot beam pattern; the current slot population is denoted as... During generation and mutation, constraints on satellite visibility, single-cell single-service, and maximum number of beams per satellite are met. During initialization, cells whose AoI (Aspect-Oriented Indicator) in the previous time slot has reached the threshold are prioritized for allocation, and idle beams are supplemented from the set of visible but unserved cells; unallocated beams remain off.

[0129] (2) Fitness assessment: For each chromosome in the population The corresponding beam pattern, during the genetic algorithm evaluation phase, uses a fixed common power, private power, and common rate average allocation to estimate the achievable rate of each cell. Based on the aforementioned continuously relaxed AoI evolution model, the single-slot average relaxed AoI of this pattern is calculated and directly used as the fitness function.

[0130]

[0131] in Candidate chromosomes Corresponding residential area Single-slot relaxation AoI;

[0132] (3) Selection and mutation: Each generation is sorted in ascending order of fitness, elite individuals are retained, and new individuals are generated by sampling from the pool of high-quality parents; mutation is triggered on a satellite-by-satellite basis, mild mutation reselects one local beam, and severe mutation reselects multiple local beams; new serving cells are selected by roulette based on weights constructed from the previous time slot AoI, channel gain and rate threshold, and the genetic algorithm does not use crossover operation;

[0133] (4) Candidate output: After iterating to the maximum number of generations or satisfying the convergence condition, sort the population by fitness and remove duplicate chromosomes, retaining A candidate beam pattern is used as input for subsequent SCA re-evaluation and beam shut-off processing based on weighted AoI.

[0134] Furthermore, in step S7 Candidate complex evaluation and beam shut-off algorithm based on weighted AoI specifically include:

[0135] Perform SCA complex evaluation on each candidate beam pattern output in step S6 to obtain the corresponding average AoI, and use the complex evaluation result as the basis for judging the merits of the candidate pattern.

[0136] Under the current candidate beam pattern, the weights defined in step S6 are used to sort the shutdown priority of the lit cells; the larger the weight, the more valuable the cell is to continue to provide services, and the smaller the weight, the lower the current service priority. Therefore, the beam corresponding to the cell with the smaller weight is selected as the candidate to be shut down.

[0137] Each time a candidate beam is temporarily disabled, the beam pattern is reconstructed and SCA is called to solve for power allocation and common rate matching. The average AoI after disabling is then calculated according to the AoI update rule. If disabling is infeasible or the average AoI deteriorates, the beam is rejected and rolled back. If the average AoI is not worse than the result before disabling, the disabling action is accepted and attempts continue. After the candidate is processed, the final beam pattern is selected according to the principle of minimum average AoI. If the average AoI is the same, the scheme with more successfully updated cells and fewer lit beams is selected first.

[0138] Furthermore, in step S8, a time-slot-by-time hierarchical solution strategy is adopted, sequentially performing power and common rate matching solutions under a fixed beam pattern, candidate pattern search based on AoI sensing, and so on, according to the time slot order. Candidate re-evaluation and beam shut-off processing based on weighted AoI are performed until all time slots are solved. The beam hopping patterns, common power allocation results, private power allocation results and common rate matching results obtained in each time slot are sent by the gateway station to the relevant low-Earth orbit satellites for execution.

[0139] The advantages and beneficial effects of this invention are as follows:

[0140] This invention addresses the challenges of uneven distribution of low-Earth orbit (LEO) satellite IoT services, strong multi-satellite interference, and the difficulty in balancing resource utilization and information freshness. It proposes a beam management method based on information age perception. Combining claims 1 to 4, this invention unifies the modeling of multi-satellite dynamic visibility, beam hopping, RSMA transmission, and information age evolution, jointly optimizing beam patterns, power allocation, and rate matching. This overcomes the limitations of existing technologies that focus on single satellites, static beams, and throughput optimization. The deep coupling of multiple dynamic elements and the complex relationships between variables and constraints are unprecedented in the industry and represent a non-conventional approach. Furthermore, this invention prioritizes minimizing the average information age across the entire network, aligning with the real-time business needs of IoT. Unlike traditional throughput-oriented optimization methods, it uses information age as a global core optimization indicator embedded in the entire constraint system, a departure from conventional technological choices. To address the challenge of solving mixed-integer non-convex models, this invention, based on claims 1, 5 to 8, employs a hierarchical solution framework combining successive convex approximation, an improved genetic algorithm, and a weighted beam-closing algorithm. These three algorithms are specifically adapted to different sub-problems, rather than simply applying general algorithms. Instead, they are customized and improved according to the characteristics of the problem, with layered linkage. The algorithm combination and adaptation logic have no mature, universally applicable solutions and are not easily conceived using conventional techniques. The overall solution balances information freshness with on-board resource utilization, demonstrating innovation and practical value in scene modeling, optimization objectives, and solution methods. Attached Figure Description

[0141] Figure 1 This is a low-orbit satellite Internet of Things scenario diagram constructed according to a preferred embodiment of the present invention;

[0142] Figure 2 This is a flowchart of a preferred embodiment of the low-orbit satellite Internet of Things beam management scheme based on AoI sensing provided by the present invention. Detailed Implementation

[0143] The technical solutions of the embodiments of the present invention will be clearly and thoroughly described below with reference to the accompanying drawings. The described embodiments are merely some embodiments of the present invention.

[0144] The technical solution of the present invention to solve the above-mentioned technical problems is:

[0145] The technical solution of this invention to solve the above-mentioned technical problems is: a low-Earth orbit satellite Internet of Things (IoT) beam management method based on AoI (Aspect-Oriented Intelligence) perception. It collaboratively considers the dynamic visibility of multiple low-Earth orbit satellites, multi-beam full-frequency reuse interference, RSMA common and private flow transmission, and the AoI evolution process. By jointly designing beam hopping patterns, common power allocation, private power allocation, and common rate matching, it minimizes the average AoI of all fixed Earth cells within the scheduling period. The specific steps are as follows:

[0146] Step 1: Set scene parameters, including: number of LEO satellites, orbital parameters, number and location coordinates of fixed Earth cells, number of IoT users, maximum number of beams per satellite, number of time slots for beam transition cycles, duration of each time slot, carrier wavelength, total power budget, and single beam power budget; calculate the satellite visibility variables to the cells based on the elevation angle criterion, and capture scene parameters within any time slot; import IoT service information, including data packet length and rate thresholds for each cell, etc.

[0147] Step 2: Model all variables to be optimized and their constraints during the scheduling process, including beam hopping patterns, common power allocation, private power allocation, and common rate matching; calculate the channel gain of each beam-cell link based on the shadowed Rice channel; calculate the common and private message rates of each user based on RSMA multicast transmission, and then obtain the total beam rate and cell reachable rate; construct an AoI evolution model to obtain the average AoI of all cells within the scheduling period, and construct a cooperative beam management model P0.

[0148] Step 3: Analyze the cooperative beam management model P0 and determine that it is a mixed-integer nonlinear nonconvex programming problem. This is because the objective function is nonlinear, constraint C8 is nonconvex, and the binary beam pattern variables and constraints (C1)-(C4) are coupled, making the original model unsolvable directly. Therefore, the original problem is decomposed into subproblems corresponding to each time slot.

[0149] Step 4: For each time slot, a time slot-by-time hierarchical solution framework is adopted: First, under a given beam pattern, the power allocation and common rate matching are solved using the SCA method; then, a discrete beam pattern is searched using an AoI-aware genetic algorithm, and the output is... Candidate patterns; final selection Candidate patterns undergo SCA re-evaluation and beam shut-off post-processing based on weighted AoI, with the average AoI after re-evaluation used as the final selection criterion.

[0150] Step 5: For the subproblem with a fixed beam pattern, a continuous relaxation and successive convex approximation method is used for solution. First, the piecewise update AoI evolution model is rewritten as a single linear expression; then, the binary variables of successful transmission are continuously relaxed; auxiliary variables are introduced to convert the non-convex constraints into DC form; a first-order Taylor expansion is performed on the convex functions in the DC form constraints to obtain its convex approximation; finally, the convex optimization problem P1 is obtained, which is solved using the CVX tool.

[0151] Step 6: For the beam pattern design sub-problem, an improved genetic algorithm based on priority is used. First, the local beam service state is considered. As a gene, the service status of all satellites and all local beams within the same time slot is concatenated in a fixed order to form a chromosome / individual. ;in Indicates the service community , This indicates beamout is off, and one chromosome corresponds to a complete single-slot beam pattern. During initialization, cells whose AoI (Aspect-Oriented Identification) in the previous slot reach the threshold are prioritized for allocation, and idle beams are supplemented from the set of satellite-visible cells not served by other beams. In the genetic algorithm evaluation phase, achievable rates are estimated using a fixed common power, private power, and common rate average allocation, and then the average relaxed AoI is calculated as the fitness based on a relaxed AoI evolution model. Elite individuals are retained in each generation, and satellite-level mild / severe mutations are performed after sampling from a pool of high-quality parents. During mutation, weights are constructed based on the previous slot AoI, channel gain, and rate threshold for roulette wheel selection. After iteration, duplicate chromosomes are removed and the result is output. Candidate beam patterns.

[0152] Step 7: Analyze the output of the genetic algorithm. For each candidate beam pattern, a SCA re-evaluation is performed to obtain cell rate, successful update variables, and average AoI. Then, for each candidate pattern, a weighted AoI-based beam shutdown post-processing is performed: the lit cells are prioritized for shutdown using the aforementioned weights, and the beams corresponding to cells with lower weights are selected as candidate shutdown targets. After temporary shutdown, the SCA solution is re-executed. If shutdown is not feasible or the average AoI deteriorates, the pattern is rejected and rolled back. If the average AoI is not worse than the result before shutdown, the pattern is accepted and attempts continue. After all candidates have been processed, the final beam pattern is selected in the order of minimum average AoI, more successfully updated cells, and fewer lit beams.

[0153] Step 8: Within the time-slot-by-time hierarchical framework, sequentially execute the power allocation solution under the fixed beam pattern and the AoI-based sensing-based genetic algorithm according to the time slot order. Candidate pattern search, SCA re-evaluation, and beam shut-off post-processing based on weighted AoI are performed until all time slots are solved. The beam hopping pattern, common power allocation results, private power allocation results, and common rate matching results for each time slot are output. Finally, the gateway station distributes the calculated beam scheduling results to the relevant low-Earth orbit satellites for execution.

[0154] The model involved in this invention is as follows:

[0155] 1. Network Model

[0156] The primary application of this invention is multi-low-Earth orbit satellite multi-beam satellite Internet of Things, such as... Figure 1 As shown in the diagram, in this scenario, multiple LEO satellites are deployed collaboratively to provide communication services to multiple fixed Earth cells distributed on the ground. Each satellite is equipped with a limited number of spot beams, and all beams share the same frequency band using full-frequency reuse. A large number of IoT terminal devices are deployed in the fixed Earth cells, and these devices need to periodically report status update information. Ground-based gateway stations are deployed, responsible for calculating and distributing beam hopping patterns, power allocation, and common rate matching schemes based on channel state information, satellite visibility relationships, and service requirements. Because LEO satellites exhibit significant relative motion with respect to the ground, the visibility of satellites to ground cells changes dynamically over time. Therefore, a beam hopping mechanism is introduced to dynamically allocate beam resources to different cells in different time slots.

[0157] 2. RSMA transmission model

[0158] This invention employs a single-layer RSMA downlink multicast transmission mechanism. All IoT users in each cell are divided into user groups, and each beam provides services to its corresponding user group. The messages transmitted by each beam are split into a common message portion and a private message portion: the common message portions of each beam are jointly encoded into a common stream shared by all users; the private message portions of each beam are independently encoded into a private stream specific to that beam. The receiving end first decodes the common stream (treating the private stream as interference), subtracts the common stream through continuous interference cancellation (serial interference cancellation), and then decodes the corresponding private stream (treating the private streams of other beams as interference). This mechanism enables flexible interference management in multi-beam interference scenarios, improving system spectrum efficiency and user fairness.

[0159] 3. AoI Evolution Model

[0160] Within each time slot, if cell c is served and the reachable rate is not less than the decoding rate threshold, the AoI of that cell is refreshed to 1; otherwise, the AoI is incremented by 1. The average AoI of all cells over the entire scheduling period is the optimization objective of this invention.

[0161] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions.

[0162] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0163] The above embodiments should be understood as illustrative only and not as limiting the scope of protection of the present invention. After reading the description of the present invention, those skilled in the art can make various alterations or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.

Claims

1. A low-Earth orbit satellite Internet of Things (IoT) beam management method based on information age perception, characterized in that, Includes the following steps: S1: Construct a multi-low-orbit satellite multi-beam satellite IoT scenario that includes multiple low-orbit satellites, fixed Earth cells, IoT terminals, and gateway stations, and establish satellite visibility relationships, beam-cell service relationships, and beam transition cycle models. S2: Construct a satellite-to-cell user channel model based on the shadowed Rice channel, and construct an information age (AoI) evolution model and a cell rate threshold model; S3: Single-layer rate segmentation multiple access (RSMA) downlink multicast transmission is adopted, and each beam message is divided into public messages and private messages. A model of public flow and private flow transmission power, public rate matching and cell reachability rate is established. S4: With the goal of minimizing the average AoI of all fixed Earth cells within a scheduling cycle, construct a collaborative beam management model that jointly optimizes beam hopping patterns, common power allocation, private power allocation, and common rate matching, and set constraints on the maximum number of satellite beams, satellite visibility, total power budget, single beam power budget, and maximum AoI of the cell. S5: For the power allocation and common rate matching subproblems under single time slot fixed beam pattern, the successful transmission variables in AoI evolution are continuously relaxed, and the common power, private power and common rate matching are solved by the successive convex approximation method. S6: For the beam pattern design subproblem, an improved priority-based genetic algorithm is used for candidate beam pattern search. Local beam state coding is employed. A fitness function and mutation weights are constructed based on the previous time slot AoI, rate satisfaction, channel gain, and rate threshold. Through forced serving cell priority allocation, initial feasible population generation, elite retention, parent pool sampling, satellite-level light / heavy mutation, and deduplication sorting, the output beam pattern is determined. Candidate beam patterns; S7: Forward Each candidate beam pattern undergoes a successive convex approximation SCA complex evaluation, and a beam shut-off algorithm based on weighted AoI is executed on the candidate patterns. The low-weight beams are shut off first, and the low-yield beams are removed if the average AoI does not deteriorate after shutting off. Otherwise, a rollback mechanism is used to remove low-yield beams, thereby obtaining the complex evaluation results of the candidate patterns. S8: Under the time-slot-by-time hierarchical solution framework, execute steps S5 to S7 sequentially according to the time slot order. After completing the solution of the current time slot, update the AoI state and enter the next time slot. Continue until all time slots are solved, and output the beam transition pattern, common power allocation result, private power allocation result and common rate matching result for the entire scheduling cycle.

2. The method for low-Earth orbit satellite IoT beam management based on information age perception according to claim 1, characterized in that, The multi-low-Earth orbit satellite multi-beam satellite Internet of Things system model in step S1 specifically includes the following: Let the set of LEO satellites and the set of fixed Earth cells be denoted as follows: and ,definition Fixed Earth Community The collection of IoT users, the sum of the number of IoT users in all communities is ;satellite The available beam set is ,definition The maximum number of beams for the satellite, a binary variable. Used to identify beams Whether by satellite generate; The beam transition period contains one Time slot, time slot index is binary variables Used to indicate satellites In the time slot Is it visible in the community? This variable is determined by the elevation angle of the cell center; if it is less than the set minimum elevation angle, the cell will not have visible satellites; binary variable. and Indicates in time slot Beam Earth-fixed communities Service relationships and satellites Earth-fixed communities Service relationship; Indicates the community In the The formula for determining whether a time slot is served is as follows: 。 3. The method for low-Earth orbit satellite IoT beam management based on information age perception according to claim 1, characterized in that, The channel model and AoI evolution model in step S2 are specifically as follows: (1) Channel model: Earth Fixed Community The first in IoT users and generated beams In the The channel gain for each time slot is: ； in, For receiving antenna gain, The angle between the beam axis and the direction of the receiving antenna, i.e., the offset angle, is a random shaded Rice variable. Obeying the parameter is the direct component Shadow fading component Multipath components The distribution of shadow Rice, For users in the first Each time slot and generated beam satellite The distance between them For carrier wavelength, The specific formula for calculating the transmit antenna gain is as follows: ； in, The angle between the beam axis and the direction of the receiving antenna, i.e., the offset angle. It is a first-order Bessel function of the first kind. Where is the antenna radius. For carrier wavelength, The peak gain of the antenna is calculated using the following formula: ； Define noise power as The specific expression is: ； in, Boltzmann's constant, Noise temperature of the receiving system It is the bandwidth allocated to the user's link; (2) AoI evolution model The condition for a successful state update is equivalent to a rate threshold constraint. Successfully transferred binary variables. To characterize the result of signal transmission; if the rate is achievable Exceeding the rate threshold Then the first The first of the IoT communities The signal of the IoT user in the first Each time slot can be successfully transmitted, using Indicate; otherwise, ; Definition of the first The first of the IoT communities The number of IoT users in the first AoI expression for each time slot: ； in, ; In the The average AoI of all cells in each time slot is: ； Furthermore, the first The average AoI of each IoT cell over the entire scheduling period is: 。 4. The method for low-Earth orbit satellite Internet of Things beam management based on information age perception according to claim 1, characterized in that, In step S3, single-layer RSMA downlink multicast transmission is used, and the relevant signal and rate models are as follows: The signal model employs single-layer RSMA downlink multicast transmission, dividing all IoT users in each cell into a group. Each time slot A multicast message was sent to In the group of cells illuminated by beams, by The index divides each message into a public part and a private part, i.e. The common parts of each group are merged and encoded into a common stream shared by all groups, i.e. The private portion is independently encoded into a private stream for each group, i.e. ; Beam The transmitted signal is ,in and This indicates the beamwidth allocated to the transmission power of public and private messages, respectively, let... Let the transmission power vector of the public message be... Let the transmission power vector of the private message be represented. Then, the transmission power matrices of the public and private messages during the entire scheduling period are defined as follows: and When each user receives a signal, the public message is first decoded, and the private message is treated as interference. In the Each time slot, collection users in The signal-to-interference-plus-noise ratio (SINR) of the public message is: ； user The achievable rate for public messages is: ； Due to common flow Shared by all beams, to ensure successful decoding for every user. The achievable rate of the system's common flow is: ； make ,in Indicates in time slot The contribution rate of each beam to the reachable rate of the system's common flow; the rate matching matrix of the common message during the entire hopping beam period is expressed as follows: Subsequently, the system uses serial interference cancellation to subtract the common stream, each user decodes its corresponding private stream, and treats the private streams of other users as interference; Therefore, in the Each time slot, user The signal-to-interference-plus-noise ratio of the private message is: ； in Indicates the first Each time slot service cell The corresponding beam, For users with corresponding channel gain of this beam The achievable rate for private messages is: ； To ensure that all users within a multicast group can successfully decode their required private signals, beamforming The private message rate is: ； No. Each time slot, beam The total rate is: ； Earth Fixed Community In the The achievable rate for each time slot is: ； Users in each cell have the same actual data rate. Therefore, the AoI evolution model for a cell can be expressed as follows: ； in Successful transmission of binary variables in the cell is represented as: ； in This is because users within the same cell have the same service, so the cell rate threshold is the same. Throughout the scheduling cycle The average AoI for all cells is: 。 5. A low-Earth orbit satellite Internet of Things beam management method based on information age perception according to claim 4, characterized in that, The cooperative beam management model constructed in step S4 aims to minimize the average AoI of all cells throughout the entire scheduling period, and is expressed as follows: ； ； C1 represents satellite In the time slot The number of communities served does not exceed C2 indicates that any cell is served by at most one beam in the same time slot; C3 indicates the visibility variable. Service relationship variables , C4 represents binary variables; C4 indicates that the common power, private power, and common rate matching are all non-negative; C5 indicates that the cell AoI does not exceed the preset upper limit. C6 indicates that the total transmission power of a single satellite does not exceed the total power of the satellite. C7 indicates that the sum of common rate matching does not exceed the system's achievable common flow rate; C8 indicates that the transmit power of a single beam does not exceed the beam power limit. .

6. The method for low-Earth orbit satellite Internet of Things beam management based on information age perception according to claim 1, characterized in that, In step S5, the power allocation and common rate matching subproblem under a single-slot fixed beam pattern is solved using a successive convex approximation method, specifically including: (1) Rewrite the objective function To express the objective function as a linear summation, we first rewrite the piecewise update of equation (17) into a single expression: ； Secondly, for The relaxation process was performed, and the specific relaxation results are as follows: ； At the same time, a new constraint is added here to ensure the effective utilization of beam resources: ； At this point, the original conditions are relaxed to This facilitates subsequent solutions, especially for the relaxed portion. It can be written as the following expression: ； (2) SCA solution First, we introduce an auxiliary variable. ,in Indicates beam The lower bound of the total rate leads to the following inequality: ； according to By definition, the above constraints can be equivalent to: ； because Given the existence of , the above constraints remain non-convex; therefore, we introduce an auxiliary variable here. and ,in This represents the lower bound of the SINR of a user's private messages, while Therefore, the above constraints can be rewritten as: ； ； Then, the SCA method is used to make And order ,but It is the difference DC function, which is a convex function. The first-order Taylor approximation is as follows: ； in yes In the The value in the next iteration; therefore, the above constraint can be approximated as: ； Next, taking the natural logarithm of constraint (30) yields its equivalent convex form: ； Similarly, for constraint C7, an auxiliary variable is introduced. ,make Indicates user The lower bound of SINR for public messages, C7 can be rewritten as: ； ； Similarly, constraint (32) can also be converted into DC form, let and ,but The first-order Taylor approximation is shown below: ； Therefore, constraint (34) can be approximated as: ； Then, perform the same operation as above on constraint (33): ； Finally, constraints C4, C6, C7, (23) can be rewritten as: ； ； ； ； Therefore, for a single time slot It can be equivalently converted to : ； ； Through the above steps Transformation of convex functions The optimal solution for this time slot can be obtained by using the CVX tool to solve for the optimal values ​​of each parameter.

7. A low-Earth orbit satellite Internet of Things beam management method based on information age perception according to claim 1, characterized in that, The priority-based improved genetic algorithm in step S6 includes: (1) Encoding and initialization: Let Indicates the first Satellites in each time slot The Local beam service status, among which This indicates that the beam serves a fixed cell on Earth. , This indicates that the beam is off; based on this, according to the fixed order of satellite number and local beam number, the service status of all local beams within a time slot is concatenated into a single entity. , No. Each chromosome corresponds to a complete single-slot beam pattern; the current slot population is denoted as... During generation and mutation, constraints on satellite visibility, single-cell single-service, and maximum number of beams per satellite are met; during initialization, cells whose AoI in the previous time slot has reached the threshold are allocated first, and idle beams are supplemented from the set of visible but unserved cells; unallocated beams remain closed. (2) Fitness assessment: For each chromosome in the population The corresponding beam pattern, during the genetic algorithm evaluation phase, uses a fixed common power, private power, and common rate average allocation to estimate the achievable rate of each cell. Based on the aforementioned continuously relaxed AoI evolution model, the single-slot average relaxed AoI of this pattern is calculated and directly used as the fitness function. ； in Candidate chromosomes Corresponding residential area Single-slot relaxation AoI; (3) Selection and mutation: Each generation is sorted in ascending order of fitness, elite individuals are retained, and new individuals are generated by sampling from the pool of high-quality parents; mutation is triggered on a satellite-by-satellite basis, mild mutation reselects one local beam, and severe mutation reselects multiple local beams; new serving cells are selected by roulette based on weights constructed from the previous time slot AoI, channel gain and rate threshold, and the genetic algorithm does not use crossover operation; (4) Candidate output: After iterating to the maximum number of generations or satisfying the convergence condition, sort the population by fitness and remove duplicate chromosomes, retaining A candidate beam pattern is used as input for subsequent SCA re-evaluation and beam shut-off processing based on weighted AoI.

8. A low-Earth orbit satellite Internet of Things beam management method based on information age perception according to claim 1, characterized in that, In step S7 Candidate complex evaluation and beam shut-off algorithm based on weighted AoI specifically include: Perform SCA complex evaluation on each candidate beam pattern output in step S6 to obtain the corresponding average AoI, and use the complex evaluation result as the basis for judging the merits of the candidate pattern. Under the current candidate beam pattern, the weights defined in step S6 are used to sort the shutdown priority of the lit cells; the larger the weight, the more valuable the cell is to continue to provide services, and the smaller the weight, the lower the current service priority. Therefore, the beam corresponding to the cell with the smaller weight is selected as the candidate to be shut down. After each temporary shutdown of a candidate beam, the beam pattern is reconstructed, and SCA is called to solve for power allocation and common rate matching. The average AoI after shutdown is then calculated according to the AoI update rule. If shutdown is infeasible or the average AoI deteriorates, the beam is rejected and rolled back. If the average AoI is not worse than the result before shutdown, the shutdown action is accepted, and attempts continue. After the candidate is processed, the final beam pattern is selected according to the principle of minimum average AoI. If the average AoI is the same, the scheme with more successfully updated cells and fewer lit beams is selected first.

9. A low-Earth orbit satellite Internet of Things beam management method based on information age perception according to claim 1, characterized in that, In step S8, a time-slot-by-time hierarchical solution strategy is adopted, sequentially performing power and common rate matching solution under fixed beam patterns, candidate pattern search based on AoI sensing, and other steps according to the time slot order. Candidate re-evaluation and beam shut-off processing based on weighted AoI are performed until all time slots are solved. The beam hopping patterns, common power allocation results, private power allocation results and common rate matching results obtained in each time slot are sent by the gateway station to the relevant low-Earth orbit satellites for execution.

Images