Beam position scheduling method based on hopping beam

Abstract

The invention belongs to the technical field of satellite communication, and particularly relates to a beam position scheduling method based on a hopping beam, which systematically solves the core bottleneck in hopping beam resource scheduling through a double-stage collaborative optimization mechanism: breaking through the limitation of single-slot scheduling by adopting cross-slot global beam position planning, and realizing the space-time balanced allocation of resources in a hopping beam period; non-uniform user distribution is dynamically matched based on density self-adaptive beam position clustering and iterative optimization, so that the high-density area coverage capability and the system load balance are remarkably improved; and internal and external interference of the system is effectively suppressed by combining minimum beam spacing constraint and a hybrid optimization strategy (greedy algorithm + simulated annealing). According to the method, the resource utilization rate is improved, the communication quality robustness is synchronously enhanced, and a high-reliability and self-adaptive beam scheduling solution is provided for a heterogeneous satellite network.

Description

Technical Field

[0001] This invention belongs to the field of satellite communication technology, specifically relating to a beam hopping scheduling method. Background Technology

[0002] With the large-scale deployment of low-Earth orbit (LEO) satellite constellations, heterogeneous satellite networks integrating LEO and GEO satellites have become a new architecture for space-air communication. LEO satellites provide high-throughput, low-latency services, while GEO satellites achieve wide-area stable coverage. However, the co-operation of the two in the same frequency band can cause serious inter-system interference, making spectrum sharing a key challenge.

[0003] Traditional multi-beam satellites use fixed beam coverage, which cannot dynamically adjust resources according to user needs, resulting in resource shortages in high-traffic areas and idle resources in low-traffic areas. Beam skipping (BH) technology divides time into time slots and dynamically switches beam direction, realizing two-dimensional temporal and spatial resource allocation, which can adapt to non-uniform user distribution and improve resource utilization efficiency.

[0004] However, existing hopping beam technology has three major drawbacks: Single-slot limitations: Most algorithms only optimize single-slot beam scheduling and lack cross-slot global collaborative planning, resulting in resource imbalance between time slots - some time slot beams are under-activated while other time slot resources are idle, and the overall system utilization rate is low. Rigid beam allocation: Regular beam allocation or simple clustering (such as K-means) is difficult to adapt to highly non-uniform user distribution, resulting in wasted beams in low-density areas, insufficient coverage in high-density areas, and poor load balancing. Insufficient interference suppression: Overlapping beams or excessively close proximity can easily cause intra-system interference, and LEO beams near GEO users can also generate inter-system interference. Existing greedy or heuristic scheduling lacks global optimization capabilities, is prone to getting trapped in local optima, and has poor robustness when user distribution changes abruptly. Summary of the Invention

[0005] The purpose of this invention is to provide a beam-hopping scheduling method, which introduces a minimum beam spacing constraint during the beam-hopping time slot scheduling stage and reserves a certain protection area for GEO users. By combining a greedy algorithm and a simulated annealing algorithm for global optimization, the anti-interference capability of the scheduling results can be improved, thereby solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a beam hopping-based position scheduling method, comprising the following steps: S1: Generate an initial set of wave position coordinates based on user density distribution: Calculate the density index based on the number of adjacent users within a user radius R, filter users in high-density areas and copy their coordinates to form a virtual user set, and generate a set of wave position center coordinates through weighted K-means clustering. The optimization objective is to minimize the sum of squared distances between users within a cluster and the cluster center; S2: Match wave positions with users based on the initial wave position coordinate set: Establish an integer linear programming model with the objective of maximizing the number of users served, and define binary decision variables. Indicates whether user i is served by beam position j, and applies beam coverage radius constraints, maximum number of users served by a single beam F constraints, and user single scheduling constraints; S3: Iterative optimization of the matching results: Traverse the inefficient wave positions where the number of users served is less than F, release their users and merge them with unscheduled users, and re-execute S1 and S2 until the number of users covered by all wave positions reaches F or the iteration limit is reached. S4: Allocate the optimized beams to M time slots: Initialize the allocation using a greedy algorithm, assigning beam sets to each time slot t. satisfy And the beam spacing within the same set is ≥ , The minimum safe distance between adjacent beam centers; S5: Perform simulated annealing optimization on the initial allocation results: randomly release the allocated beams, redistribute them to time slots according to the beam spacing constraint, take the total number of service beams as the optimization objective, and control the probability of accepting new solutions through annealing parameters.

[0007] Preferably, the objective function of the weighted K-means clustering is: ;in Indicates the first The coordinates of each user Indicates the first A set of user coordinates covered by wave positions.

[0008] Preferably, the constraints of the integer linear programming model are: ; in This represents the center distance between user i and wave position j. This represents the number of users waiting to be connected.

[0009] Preferably, the greedy algorithm initialization allocation satisfies the following constraints: ; in, For time slots Beam set within, The maximum number of beams that can be lit in each time slot, and the coordinates of the beam center are: .

[0010] Preferably, in the beam spacing constraint R is the beam coverage radius.

[0011] Preferably, the acceptance condition for the new solution optimized by simulated annealing is: when the total number of serving beams of the new solution... Accept immediately.

[0012] Preferably, in S4 and S5, time slot allocation prioritizes avoiding the GEO user protection area, and the wave position coordinates are located outside the GSO protection area.

[0013] Technical effects and advantages of the present invention: The beam hopping-based position scheduling method proposed in this invention has the following advantages compared with the prior art: This invention systematically addresses the core bottleneck in beam hopping resource scheduling through a two-stage collaborative optimization mechanism: it employs cross-timeslot global beam planning to overcome the limitations of single-timeslot scheduling, achieving a balanced spatiotemporal allocation of resources within the beam hopping period; based on density-adaptive beam clustering and iterative optimization, it dynamically matches non-uniform user distributions, significantly improving coverage in high-density areas and system load balancing; and it combines minimum beam spacing constraints with a hybrid optimization strategy (greedy algorithm + simulated annealing) to effectively suppress internal and external interference. This method improves resource utilization while simultaneously enhancing communication quality robustness, providing a highly reliable and adaptive beam scheduling solution for heterogeneous satellite networks. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the user uniform distribution wave position division according to the present invention; Figure 2 This is a schematic diagram of the non-uniform distribution wave position division for users according to the present invention; Figure 3 This is a schematic diagram of the time slot wave position distribution uniformly distributed among users according to the present invention; Figure 4 This is a schematic diagram of the non-uniformly distributed time slot wave position distribution for users according to the present invention; Figure 5 This is a flowchart of the multi-slot global beam position scheduling algorithm based on beam skipping of the present invention. Detailed Implementation

[0015] 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. The specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention. 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.

[0016] This invention provides a beam hopping-based position scheduling method, comprising the following steps: S1: Generate an initial set of wave position coordinates based on user density distribution: Calculate the density index based on the number of adjacent users within a user radius R, filter users in high-density areas, and copy the coordinates to form a virtual user set. The objective function of weighted K-means clustering is: ;in Indicates the first The coordinates of each user Indicates the first A set of user coordinates covered by wave positions.

[0017] S2: Match wave positions with users based on the initial wave position coordinate set: Establish an integer linear programming model with the objective of maximizing the number of users served, and define binary decision variables. Indicates whether user i is served by beam position j, and applies beam coverage radius constraints, maximum number of users served by a single beam F constraints, and user single scheduling constraints; Furthermore, the constraints of the integer linear programming model are: ; in This represents the center distance between user i and wave position j. This represents the number of users waiting to be connected.

[0018] S3: Iterative optimization of the matching results: Traverse the inefficient wave positions where the number of users served is less than F, release their users and merge them with unscheduled users, and re-execute S1 and S2 until the number of users covered by all wave positions reaches F or the iteration limit is reached. S4: Allocate the optimized beams to M time slots: Initialize the allocation using a greedy algorithm, assigning beam sets to each time slot t. satisfy And the beam spacing within the same set is ≥ , The minimum safe distance between adjacent beam centers; Furthermore, the greedy algorithm's initial allocation satisfies the following constraints: ; in, For time slots Beam set within, The maximum number of beams that can be lit in each time slot, and the coordinates of the beam center are: Beam spacing constraints R is the beam coverage radius.

[0019] S5: Perform simulated annealing optimization on the initial allocation results: randomly release the allocated beams, reallocate them to time slots according to beam spacing constraints, use the total number of serving beams as the optimization objective, and control the probability of accepting new solutions through annealing parameters. The condition for accepting new solutions optimized by simulated annealing is: when the total number of serving beams of the new solution... Accept immediately.

[0020] To more clearly illustrate the specific operation of the above-mentioned beam-hopping-based position scheduling method, the following will provide further explanation with reference to an embodiment: This invention proposes a multi-timeslot global beam position scheduling algorithm based on beam hopping. This method divides the resource scheduling process within a beam hopping period into two stages: first, a beam position partitioning algorithm optimizes the global beam position, generating a set of candidate beam positions; then, a beam position timeslot scheduling algorithm rationally allocates these beam positions to each beam hopping timeslot, forming a scheduling scheme that satisfies spatiotemporal constraints. These two sub-algorithms work together to achieve spatial distribution optimization and dynamic temporal scheduling of beam resources within a beam hopping period. Figure 5 As shown, the implementation steps of the two sub-algorithms are as follows.

[0021] 1. Wavelength partitioning algorithm based on ILP: In the first sub-algorithm, an ILP-based beamposition allocation algorithm is proposed. This algorithm pre-divides the schedulable beams of the LEO system throughout the entire beam-hopping cycle based on the distribution of LEO users located outside the GSO user protection zone, and aims to maximize the number of users served by each beam. The algorithm consists of three steps: generating initial beamposition coordinates, beamposition and user matching, and iterative optimization.

[0022] Generate initial wave position coordinates: In the initial wavefront generation phase, the radius of each user is calculated by weighting based on the local density of the users. The number of neighboring users within a given range is used as a density indicator. Users in high-density areas are selected, and their coordinates are copied to form a virtual user set. This allows the K-means clustering algorithm to generate more clusters in that area. After clustering, the weighted user set outputs a preliminary set of wavelet center coordinates. The objective function is to minimize the sum of squared distances between users within the cluster and the cluster center: ;in Indicates the first The coordinates of each user Indicates the first A set of user coordinates covered by wave positions.

[0023] Wave position and user matching: After calculating the initial wave position center coordinates, the number of served users within each wave position is maximized using the ILP model. Binary decision variables are defined. Indicates user Is it by wave position? Service. With the objective of maximizing the number of users served, and constrained by wavelet radius, wavelet service quantity, and user scheduling frequency, the following optimization problem is established: ; user With wave position The center distance is , The number of users to be connected, and the constraints. This means that users can only cover the radius of the wave position. Internal scheduling, constraints Indicates the maximum service per beam Individual users, constraints This means that each user will be scheduled at most once.

[0024] Iterative optimization: Because the algorithm for generating initial wave position coordinates and matching wave positions with users may leave fewer than the number of users served. To address the inefficient use of wave positions, an iterative compensation mechanism was introduced. After matching the initial wave positions with users, all wave positions were traversed and detected. Users served by inefficient wave positions were released and merged with unscheduled users. Clustering and ILP planning were then performed again, and the above steps were repeated. Through multiple iterations, the wave position distribution was gradually optimized until the number of users covered by all wave positions reached its maximum or the number of iterations reached its upper limit.

[0025] 2. Globally optimized wavelet time slot allocation algorithm: The second sub-algorithm proposes a globally optimized wavelength slot allocation algorithm, which distributes the wavelengths output from Algorithm 1 into various time slots. This algorithm will... Each wave position is assigned to one Within a time slot, a maximum of [number] lights can be lit per time slot. There are one beam, and the spacing between any two beams in the same time slot must meet the following requirements. To avoid inter-beam interference and thus ensure system performance.

[0026] The globally optimized beam slot allocation algorithm combines a greedy initialization algorithm with a simulated annealing strategy to maximize the total number of serving beams while satisfying beam spacing constraints and slot capacity limitations. The following optimization problem is established: ; in, For time slots Beam set within, The maximum number of beams that can be lit in each time slot, and the coordinates of the beam center are: Constraints This indicates that the beam allocated to each time slot does not exceed Constraints This means that the spacing between the beams illuminated in each time slot must be greater than [a certain value]. State variables Indicates beam The allocated time slots, In the case of no time slots initially allocated .

[0027] The pseudocode for Algorithm 2 is as follows: Phase 1: Generate initial solutions using a greedy algorithm initialization: ,

[0028] 1: for each time slot do: 2: for each beam do 3:

[0029] 4: Update the beam set

[0030] 5:Update the state variables,

[0031] 6: end if 7:

[0032] 8: break 9: end if 10: end for 11: end for Phase 2: Iterative solution using simulated annealing algorithm 12: Initialization: Annealing Algorithm Parameters , , Maximum number of iterations

[0033] 13: Initialization: Input the initial solution of the greedy algorithm

[0034] 14: while

[0035] 15: for 1:

[0036] 16: Randomly select and allocate time slots Internal beam and release 17: Update state variables and

[0037] 18: for each time slot do 19:

[0038] 20: Update slots beam set

[0039] 22: break twenty three:

[0040] 24: Accept the new interpretation immediately 25: else 26: Probability of accepting new solutions 27: end if 28: end for 29:

[0041] 30: end for 31: end while As shown in the table below, Table 1 contains simulation parameters:

[0042] Under the simulation test parameters in Table 1, the theoretical number of users scheduled by the system is 432. When the users are evenly distributed, the actual number of users called using Algorithm 1 is 430, and the resource utilization rate is 99.5%. Figure 1 As shown. When the access users are unevenly distributed with local hotspot clusters, and the number of hotspot areas is 12, the actual number of users accessing the system using Algorithm 1 is 430, and the resource utilization rate is 99.5%. The wavelet partitioning result is as follows. Figure 2 As shown.

[0043] Simulation results show that, under the same user distribution conditions, the iterative compensation mechanism can effectively release wavelets with fewer users during the wavelet allocation stage. The released users are then re-allocated to wavelets, which can effectively increase the number of users actually scheduled by the system and improve the system resource utilization rate.

[0044] Under the simulation test parameters in Table 1, the system can illuminate 96 beams in 6 time slots. When users are evenly distributed, the actual number of beams that the system can activate is 96. Figure 3As shown; when the access users are unevenly distributed with local hotspot clusters, and the number of hotspot areas is 12, the actual number of beams that the system can activate is 90, as shown. Figure 4 As shown, the number of beams lit is slightly lower than the upper limit of the system's beams that can be lit, but the number of beams lit in each time slot is relatively uniform, about 14 beams. This is because the number of beams divided in the high user density area is too large, and it is impossible to light them all while ensuring the beam spacing distance, which leads to a slight decrease in system performance.

[0045] In summary, the multi-slot global beam-position scheduling algorithm based on beam skipping proposed in this invention has the following significant advantages: First, this invention solves the problem that beam hopping resource scheduling in the prior art is limited to single-slot optimization. It adopts a user density-based beam position partitioning and integer linear programming method, combined with a globally optimized slot allocation mechanism, to uniformly schedule all beam positions within the beam hopping period. This achieves overall planning and optimization of beam resources across time slots, improves system resource utilization, and avoids problems such as beam duplication and resource waste.

[0046] Secondly, this invention enhances the system's adaptability to the non-uniformity of user spatial distribution through density-aware multi-stage wavelet clustering, iterative optimization, and a joint optimization method of greedy and simulated annealing, and effectively improves the global search capability of the scheduling algorithm, enabling better scheduling results under constraints such as multi-beam and multi-time slot.

[0047] Finally, this invention sets minimum distance constraints between beams in beam position scheduling, reasonably controls the spatial position of active beams, significantly reduces internal interference, effectively ensures communication quality, and can reserve sufficient protection space for GEO systems, thus having good engineering practical value.

[0048] The abbreviations and key terms in this embodiment are defined as follows: GEO: Geostationary Earth Orbit LEO: Low Earth Orbit BH: BeamHopping ILP: Integer Linear Programming.

[0049] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A beam-hopping-based position scheduling method, characterized in that, Includes the following steps: S1: Generate an initial set of wave position coordinates based on user density distribution: Calculate the density index based on the number of adjacent users within the user radius R, filter users in high-density areas and copy the coordinates to form a virtual user set, generate a set of wave position center coordinates through weighted K-means clustering, and optimize the goal of minimizing the sum of squared distances between users in the cluster and the center; S2: Match the initial beam position coordinate set with users: Establish an integer linear programming model with the goal of maximizing the number of users served. Define a binary decision variable to represent whether user i is served by beam position j, and apply beam coverage radius constraints, maximum number of users served by a single beam F constraints, and user single scheduling constraints. S3: Iterative optimization of the matching results: Traverse the inefficient wave positions where the number of users served is less than F, release their users and merge them with unscheduled users, and re-execute S1 and S2 until the number of users covered by all wave positions reaches F or the iteration limit is reached. S4: Assign the optimized beam positions to M time slots: Use a greedy algorithm to initialize the assignment. The beam set assigned to each time slot t satisfies that the beam spacing within the same set is ≥, which is the minimum safe spacing between adjacent beam centers. S5: Perform simulated annealing optimization on the initial allocation results: randomly release the allocated beams, redistribute them to time slots according to the beam spacing constraint, take the total number of service beams as the optimization objective, and control the probability of accepting new solutions through annealing parameters.

2. The beam hopping-based position scheduling method according to claim 1, characterized in that, The objective function for the weighted K-means clustering is: ;in Indicates the first The coordinates of each user Indicates the first A set of user coordinates covered by wave positions.

3. The beam hopping-based position scheduling method according to claim 1, characterized in that, The constraints of the integer linear programming model are: ； in This represents the center distance between user i and wave position j. This represents the number of users waiting to be connected.

4. The beam hopping-based position scheduling method according to claim 1, characterized in that, The greedy algorithm's initial allocation satisfies the following constraints: ； in, For time slots Beam set within, The maximum number of beams that can be lit in each time slot, and the coordinates of the beam center are: .

5. A beam-hopping-based position scheduling method according to claim 4, characterized in that, In the beam spacing constraint R is the beam coverage radius.

6. The beam hopping-based position scheduling method according to claim 1, characterized in that, The acceptance condition for the new solution of the simulated annealing optimization is: when the total number of serving beams of the new solution is... Accept immediately.

7. The beam hopping-based position scheduling method according to claim 1, characterized in that, In S4 and S5, time slot allocation prioritizes avoiding the GEO user protection area, and the waveform coordinates are located outside the GSO protection area.

Images