A cell-satellite bilateral matching method suitable for multi-layer low earth orbit satellite constellation

Abstract

The application discloses a cell-satellite two-sided matching method suitable for a multi-layer low-orbit satellite constellation, and belongs to the satellite application industry. In view of the matching problem of multiple satellites and multiple ground cells in the multi-layer low-orbit satellite constellation scene, the application models the associated matching problem of the cells and the satellites into a one-to-many two-sided matching problem, then designs the benefit function of the two sides according to the user time delay demand and the satellite load balancing demand, and establishes the own preference list of the cells and the satellites respectively, and the matching of the satellites and the cells is carried out by using a delayed acceptance algorithm. The application comprehensively considers the difference of the cell service characteristics and the satellite load status, realizes the dynamic matching of the cells and the satellites, can select the service satellite for the cell with multiple visible satellites, reduces the total time delay of the real-time service of the cell while balancing the inter-satellite load, is suitable for the multi-layer low-orbit satellite constellation scene, and maximizes the utilization of the on-satellite resources while meeting the Qos demand of the cell.

Description

Technical Field

[0001] This invention relates to the satellite application industry and belongs to the field of satellite communication technology. Specifically, it refers to a cell-satellite bilateral matching method applicable to multi-layer low-Earth orbit satellite constellations. Background Technology

[0002] Low Earth Orbit (LEO) satellite constellations are satellite networks composed of multiple LEO satellites with relatively stable spatial geometry. They can alleviate problems such as limited onboard resources and limited Earth coverage of single-satellite constellations, offering advantages such as low cost, low transmission loss and latency, flexible networking, and strong resilience. Currently, the development trend of LEO satellite constellations is from single-layer to multi-layer and from small-scale to large-scale. For example, the first-generation Iridium constellation used only 66 satellites to achieve Earth coverage, while the second-generation OneWeb constellation uses 720 LEO satellites, and Starlink's satellite count ranges from hundreds to thousands. In large-scale, multi-layered LEO satellite constellation scenarios, there are often overlapping coverage areas. A single ground cell often has multiple visible satellites, and a single satellite can simultaneously cover multiple ground cells. Therefore, it is necessary to further define the service cell set for each satellite and the service satellites for each cell to achieve bilateral matching between multiple cells and multiple satellites.

[0003] Bilateral matching refers to the process by which matching decision-makers, based on relevant information about the other party provided by the matching parties, obtain the optimal match between the two parties according to their own preferences. Early research on bilateral matching mainly focused on marriage matching, university admissions, and hospital-intern matching. Currently, many scholars have built upon this foundation to solve related problems in multidimensional fields such as human resource management, resource allocation, and e-commerce. Bilateral matching also has wide applications in the field of communications, such as matching satellites and missions in computational offloading and matching resource blocks with D2D users. In bilateral matching, both parties can establish preference lists based on their liking for each other. The inviting party sends matching requests according to the order of preference lists, and the receiving party chooses to agree or reject the matching request based on preferences to obtain maximum satisfaction. Bilateral matching can take into account the needs of both parties while maintaining high flexibility and low complexity, making it very suitable for scenarios with limited onboard resources. Therefore, research on the application of bilateral matching in cell-satellite dynamic association is of great significance for achieving optimal allocation of low-Earth orbit satellite network resources.

[0004] A reasonable cell-satellite matching scheme must consider the different related needs of both sides, balancing on-board load while meeting cell QoS requirements. If matching is based solely on the needs of one side, the matching result will only benefit that side, neglecting the needs of the other side, leading to unbalanced satellite load or reduced user experience in the cell, making multi-objective optimization difficult. If only a single need is considered, such as using distance as the access metric, all cells will select the nearest satellite for access, resulting in satellite overload near hotspot areas and surplus satellite resources in other areas. Therefore, the matching scheme must consider the needs of both sides and multi-dimensional access metrics, reasonably establishing utility and preferences to achieve better adaptation between satellite resources and terrestrial cell services.

[0005] Common satellite resource allocation schemes primarily focus on adapting single-satellite multi-dimensional resources to multiple ground cells. This means a single satellite provides services to multiple ground cells, allocating appropriate time-domain, frequency-domain, and power-domain on-board resources based on factors such as the service characteristics of each cell. However, this approach doesn't consider multi-satellite scenarios within low-Earth orbit (LEO) constellations. For multi-satellite scenarios, greedy algorithms are common in multi-satellite resource and ground cell adaptation schemes. In each round of matching, the optimal matching object is selected sequentially according to certain rules, such as each cell choosing the satellite with the lowest existing load / closest distance for access, until matching is complete. In summary, current satellite resource and ground cell adaptation schemes largely fail to consider the large-scale / multi-layered characteristics of LEO satellite constellations, lacking research on matching scenarios between cells and multiple / dual-layer satellites. Furthermore, when designing matching rules, it's necessary to simultaneously consider the access needs of both the satellite and the cell, maximizing the utilization of on-board resources while meeting cell QoS requirements. Summary of the Invention

[0006] To address the matching problem between multiple satellites and multiple ground cells in a multi-layer LEO satellite constellation scenario, this invention provides a cell-satellite bilateral matching method applicable to multi-layer LEO satellite constellations. Based on the differences in cell service characteristics and the current satellite load, a bilateral utility function and preference list are established to realize a dynamic cell-satellite matching strategy. This method selects the serving satellite for a cell with multiple visible satellites, thereby reducing the total latency of real-time services in the cell while balancing inter-satellite load.

[0007] The present invention provides a cell-satellite bilateral matching method applicable to multi-layer low-Earth orbit satellite constellations, comprising the following steps:

[0008] (i) First, the association and matching problem between cells and satellites is modeled as a one-to-many two-sided matching problem, that is, each satellite has multiple serving cells, while a cell can only be served by one satellite.

[0009] (ii) Subsequently, based on user latency requirements and satellite load balancing requirements, a two-sided benefit function was designed and a preference list was established; where:

[0010] The utility function of a cell for a satellite is only related to the total real-time service latency of the cell after the satellite is selected; each cell calculates the benefit value of each low-Earth orbit satellite according to the utility function, and sorts the satellites in descending order of benefit value to establish a satellite preference list;

[0011] The utility function of a satellite for a cell consists of two parts: load balancing benefits and latency benefits. Each satellite calculates the benefit value of each cell based on the utility function, sorts the cells in descending order of benefit value, and establishes a cell preference list. The cell preference list records all unmatched cells within the visible range of the satellite.

[0012] (iii) A delayed reception algorithm is used to match satellites with cells, and the matching results are output;

[0013] A delayed acceptance algorithm is used for multi-round matching. In each round of matching, the satellite sends a matching request to the cell with the highest cell in its cell preference list, and the cell receives the matching request from the satellite with the highest cell in its cell preference list. After each round of matching, the matching list, satellite load status and satellite cell preference list are updated. When the satellite is fully loaded or there are no cells that can be invited within its line of sight, the satellite will not send new invitations.

[0014] The advantages and positive effects of this invention are as follows: The method of this invention simultaneously considers the non-uniform distribution characteristics of real-time and non-real-time services in the cell, as well as the satellite's location and load status. It designs a utility function and generates a preference list based on the access requirements of both parties, and uses a delayed reception algorithm to obtain the matching result. Experimental verification shows that, compared to random access algorithms, this method can balance the number of access cells per satellite while balancing the on-board load, thus maximizing the utilization of on-board resources while meeting cell QoS requirements. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of a cell-satellite matching association model;

[0016] Figure 2 This is a schematic diagram illustrating the process of achieving cell-satellite bilateral matching in this invention;

[0017] Figure 3 This is a technical effect diagram comparing the method of the present invention with existing matching methods; a1 and a2 are comparisons of the different types of services served by each satellite under different matching methods, and b1 and b2 are comparisons of the total services of each satellite and the number of serving cells under different matching methods. Detailed Implementation

[0018] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0019] Because satellites vary in altitude and nadir position, and ground cell services are unevenly distributed, different types of cells or satellites have different access preferences. For a single satellite, selecting a serving cell requires considering factors such as channel attenuation and cell service characteristics (service type and volume). For a satellite system, multi-satellite load balancing must also be considered; that is, after matching, the total service volume of each satellite and the number of its connected cells should be similar to avoid overloading or underloading of any particular satellite, thus maximizing the utilization of onboard resources. For cells, service types include real-time and non-real-time services. Each cell can select its access satellite based on its service characteristics. Cells with higher real-time service volume can access L-LEO (low-level low-Earth orbit) satellites / satellites with high elevation angles to obtain lower propagation latency, while cells with lower real-time service volume have lower latency requirements and can choose to access H-LEO (high-level low-Earth orbit) satellites for service. Therefore, effectively designing the utility functions of both parties and establishing a preference list based on their respective needs is crucial to the performance of bilateral matching.

[0020] The cell-satellite bilateral matching method of this invention is applied to multi-layer low-Earth orbit satellite constellations, such as... Figure 1 As shown. There are several cells on the ground with different service characteristics, and multiple low-Earth orbit satellites with different altitudes and orbits in the air. The purpose of this invention is to achieve the correspondence between multiple satellites and multiple cells through a bilateral matching strategy. The implementation of this invention mainly involves: firstly, establishing the access preferences of the two sides, i.e., H-LEO prefers to access cells with closer proximity / few real-time services, while L-LEO prefers to access cells with more real-time services. Both types of satellites also hope to achieve a more balanced system load after accessing cells, and the cells prefer satellites with lower latency. Subsequently, a utility function is designed based on the above access preferences. Then, the satellites / cells rank the utility values ​​of cells / satellites within their line of sight according to their preferences, and each round of matching is completed based on the preference list. The matching output result is the satellite to which each cell belongs. The satellite can further determine which cells it provides service beams for in a certain time slot within its service cell set.

[0021] This invention models the cell-satellite matching problem as a one-to-many bilateral matching problem, where a cell can choose one serving satellite, and a satellite can simultaneously serve multiple cells. The method first establishes a preference list by rationally setting a bilateral utility function based on the optimization objectives: minimizing the total real-time service latency of the cell and balancing satellite load. Satellites select inviting ground cells according to the preference list. Cells then choose the satellite with the highest priority in their preference list from the invited satellites and reject invitations from other satellites. Each round of matching updates the matching results, the list of unserved cells, the current satellite load, and the satellite preference list. After several rounds of matching, the satellite assigned to each cell and the set of serving cells for each satellite are finally determined, thus completing the many-to-one cell-satellite matching. The following section combines... Figure 2 This invention provides a detailed explanation of one implementation process of its method.

[0022] Step 1: Determine the optimization objective.

[0023] This invention models the cell and satellite association matching problem as a one-to-many bilateral matching problem, that is, each satellite has multiple serving cells, and a cell can only be served by one satellite.

[0024] Step 2: Establish the utility function and construct a list of bilateral preference relationships.

[0025] Community C n For satellites i utility function U n (i) Only compared with the total real-time service latency of the cell after selecting this satellite. Related, can be represented as

[0026]

[0027] in, For community c n With satellites i The propagation delay during matching is determined by the satellite location relationship of the cell. c is the speed of light in a vacuum, c = 3 × 10⁻⁶ 8 m / s, For community c n With satellites i The distance between them, let They are respectively community c n Geographical coordinates and elevation, visible height of the community Satellites s i The geographical coordinates and altitude are then For community c n Real-time traffic volume.

[0028] The satellite's utility function consists of two parts: load balancing benefit and latency benefit. Regarding load, the satellite load ρ after each round of matching... i Compared with ideal average load The closer the cells are, the higher the load balancing benefit. Regarding latency, for H-LEO, the total latency of the matching cells is... The lower the delay, the higher the efficiency; for L-LEO, in terms of latency... Under approximately the same conditions, for real-time traffic volume More matching cells result in higher efficiency. Therefore, high-level satellites s i For community C n utility function U i (n) can be represented as

[0029]

[0030] Low-level satellites i For community c n utility function U i (n) can be represented as

[0031]

[0032] Here, ||·|| represents normalization, and α∈(0,1) serves as a weighting factor to adjust the weights of load balancing and latency factors. Furthermore, since the list of matchable cells is updated with each satellite-cell matching operation, and the satellite load ρ... i The results will change with each round of matching, so the satellite has a dynamic preference for cells.

[0033] Ideal average load It can be calculated using the following formula:

[0034]

[0035] in, For community c n The volume of business, The sum of traffic volume for all cells, where n is the cell number; For satellites s i The initial load, Let N be the initial payload of all satellites, where i is the satellite number; S This refers to the number of satellites.

[0036] Community C n The business volume is its real-time business volume Non-real-time business volume The sum of, i.e.

[0037] Each community cn The utility function (1) can be used to calculate the benefit value of each satellite. The satellites are arranged in descending order of their benefit values ​​to construct cell c. n The satellite preference list P(n).

[0038] Each satellite i Calculate the benefit value for each cell based on the utility function (2) or (3), arrange the cells in descending order of the calculated benefit value, and construct satellite s. i The cell preference list P(i). Based on the cell preference list P(i), satellites s i The community with the highest benefit value, i.e., the one ranked highest in the community preference list, will be prioritized. n * =argmaxU i (n) Initiate a matching invitation. The cell can accept or reject the satellite's matching request according to its preference order.

[0039] The satellite preference list for a cell contains all visible satellites for that cell, while the cell preference list for a satellite contains all unmatched cells within the visible range of that satellite.

[0040] Step 3: Use the delayed reception algorithm to match satellites with cells and output the matching results.

[0041] The delayed acceptance algorithm is shown in Table 1. In each round of matching, the satellite first sends an invitation to the cell with the highest preference. The cell can accept the satellite at the top of its own satellite preference list and reject other inviting satellites. After each round, the matching list, satellite load status, and set of unmatched cells are updated. If the satellite is fully loaded or there are no cells within its line of sight that can be invited, no new invitations will be sent until the number of matching attempts ends.

[0042] Table 1 shows satellite-cell association matching using the delayed reception algorithm.

[0043]

[0044]

[0045] In the table above, the maximum number of matches N is a large number that is set in advance to ensure that a matching result can be output after N rounds of matching.

[0046] As described in the table above, in each round of matching, each satellite, based on its own cell preference list, submits a matching request to the cell with the highest benefit value, and adds the invited cell to the array `temp_cell`. Upon receiving a matching request, a cell, based on its own satellite preference list, accepts the request from the satellite with the highest preference among the inviting satellites, and rejects matching requests from other satellites. If only one satellite submits a matching request to a cell, the cell is accepted directly. After each round of matching, the successfully matched cell-satellite pair is added to the overall matching list, and then the matched cell is removed from each satellite's cell preference list. The satellite's cell preference list is then updated according to the utility function. During each round of matching, if there are no available free cells within the satellite's line of sight or the satellite is fully loaded, no new cells will be matched for the satellite.

[0047] After N rounds of matching, the satellite-cell matching vectors Γ(s1), Γ(s2), ..., Γ(s) are finally output. i ), Γ(s i Recorded as satellite s i All matched cells.

[0048] The effectiveness of the method of this invention is compared with that of a scheme using traditional random matching, such as... Figure 3 As shown in the figure. This experiment uses the association of 6 LEO satellites with 143 cells as an example. Five of the 6 LEO satellites are L-LEOs, numbered 1 to 5, and one is an H-LEO, numbered 6. The matching results using the cell-satellite bilateral matching method and the random matching method of this invention are plotted respectively. Figure 3 In this context, a1 and b1 represent the cases where the method of this invention is used, while a2 and b2 represent the cases where a random method is used for matching. Figure 3 In the four graphs, the horizontal axis represents the satellite number, the vertical axis in a1 and a2 represents the service volume served by the satellite, and the vertical axis in b1 and b2 compares the total satellite service volume with the number of access cells. The graphs compare the service volume of each satellite under the two methods, including real-time service volume, non-real-time service volume, and total service volume, as well as the number of access cells. Figure 3As can be seen, using the random matching method, a cell randomly selects one satellite from the visible satellite set for access. Therefore, the number of matched cells and service volume of the high-level satellite H-LEO (satellite 6), which has a wide coverage area, are much higher than other satellites. However, for the low-level satellite L-LEO (satellite 5), which is farther away from the cell, the number of matched cells and service volume are lower than other satellites due to the limited number of visible cells. The method of this invention provides a more balanced service volume across satellites, and the cell will select a more suitable satellite based on the total real-time service latency. Experimental results show that the cell-satellite bilateral matching method of this invention can effectively balance the service load and the number of access cells for each satellite at different altitudes / locations, solving the matching problem between multi-level low-Earth orbit satellites and multiple ground cells, maximizing the utilization of onboard resources while meeting cell QoS requirements.

[0049] Except for the technical features described in the specification, all other technologies are known to those skilled in the art. Descriptions of well-known components and technologies are omitted in this invention to avoid redundancy and unnecessary limitation. The embodiments described above do not represent all embodiments consistent with this application. Various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of this invention are still within the protection scope of this invention.

Claims

1. A cell-satellite bilateral matching method applicable to multi-layer low-Earth orbit satellite constellations, characterized in that, include: Step 1: Model the association and matching problem between cells and satellites as a one-to-many two-sided matching problem, where each satellite has multiple serving cells, and a cell can only be served by one satellite. Step 2: Design a two-sided utility function and establish a preference list based on user latency requirements and satellite load balancing requirements; where: The utility function of a cell for a satellite is only related to the total real-time service latency of the cell after the satellite is selected; each cell calculates the benefit value of each low-Earth orbit satellite according to the utility function, and sorts the satellites in descending order of benefit value to establish a satellite preference list; The utility function of establishing a cell for the satellite is as follows: Set up a community For satellites The utility function is ,as follows: (1) in, For the community Real-time business volume; For the community With satellite The propagation delay during matching is obtained based on the location relationship between the cell and the satellite. c is the speed of light in a vacuum. For the community With satellite The distance between; The utility function of a satellite for a cell consists of two parts: load balancing benefits and latency benefits. Each satellite calculates the benefit value of each cell based on the utility function, sorts the cells in descending order of benefit value, and establishes a cell preference list. The cell preference list records all unmatched cells within the visible range of the satellite. Specifically, based on whether the satellite is a high-level low-Earth orbit satellite or a low-level low-Earth orbit satellite, the corresponding utility function is established as follows: If satellite Establish satellites for high-altitude and low-Earth orbit satellites. For the community utility function as follows: (2) If satellite To establish satellites for low-altitude, low-Earth orbit. For the community utility function as follows: (3) in, This is a weighting factor used to adjust the weights of load balancing and latency factors. For satellite The load; Ideal average load; For the community Real-time business volume; For the community With satellite Propagation delay during matching; Indicates normalization; Step 3: Use a delayed reception algorithm to match satellites with cells and output the matching results; A delayed acceptance algorithm is used for multi-round matching. In each round of matching, the satellite sends a matching request to the cell with the highest cell in its cell preference list, and the cell receives the matching request from the satellite with the highest cell in its cell preference list. After each round of matching, the matching list, satellite load status and satellite cell preference list are updated. When the satellite is fully loaded or there are no cells that can be invited within its line of sight, the satellite will not send new invitations.

2. The method according to claim 1, characterized in that, In step 2, the ideal average load is calculated as follows: (4) in, For the number of satellites, For satellite The initial load, For the community The traffic volume of a cell consists of real-time traffic volume and non-real-time traffic volume. This represents the total transaction volume of all residential communities. The sum of the initial loads of all satellites.

3. The method according to claim 1, characterized in that, In step 3, a maximum number of matching times N is preset. In each round of matching, each satellite submits a matching request to the cell with the highest benefit value according to its own cell preference list. After receiving the matching request, the cell accepts the satellite with the highest benefit value among the invited satellites according to its own satellite preference list and rejects the matching requests of other satellites. If only one satellite submits a matching request to the cell, the matching request is accepted directly. After each round of matching, the cells-satellites that were successfully matched in this round are added to the matching list. Then, the matched cells are removed from the cell preference list of each satellite, and the cell preference list and load status of the satellite are updated.

Images