A terminal grouping switching method in a low earth orbit satellite internet of things scenario
By employing a spectral clustering algorithm and a utility function-optimized group handover method, the problems of frequent handovers and uneven resource allocation in low-Earth orbit satellite networks were solved, achieving efficient terminal group handover and improving communication reliability and the continuity of emergency communications.
Patent Information
- Application Number
- CN202411846425.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-16
- Publication Date
- 2026-02-17
- Estimated Expiration
- 2044-12-16
AI Technical Summary
In low Earth orbit satellite networks, traditional terminal handover mechanisms result in frequent handovers, high signaling overhead, low handover success rates, and unbalanced loads. Furthermore, existing grouping methods are highly complex and struggle to achieve effective emergency communication continuity and stability.
The initial grouping is performed using a spectral clustering algorithm. A similarity matrix is calculated based on bandwidth requirements, location coordinates, and service priorities. A utility function is constructed to optimize intra-group and inter-group switching. A game theory framework is used to optimize resource allocation, reduce algorithm complexity, and improve switching success rate.
It improves the accuracy and stability of grouping, reduces algorithm complexity, optimizes signaling overhead and resource allocation, ensures the continuity and stability of emergency communication, and adapts to various satellite communication scenarios.
Smart Images
Figure CN119743808B_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application belongs to the technical field of low-orbit satellite wireless communication, and particularly relates to a terminal grouping switching method in a low-orbit satellite Internet of Things scenario. BACKGROUND
[0002] With the rapid development of communication technology, satellite communication systems, as a bridge connecting the ground and space, play a crucial role in providing seamless coverage services worldwide. Under the background that the Internet of Things technology has become the core support in many fields, satellite networks occupy an indispensable position due to their unique wide-area coverage capability. Specifically, the coverage radius of a single satellite can extend to hundreds of thousands or even millions of square kilometers, making it an optimal solution for providing efficient and extensive network connection services to large-scale user groups.
[0003] However, in a low-earth orbit satellite network, due to the much higher running speed of low-earth orbit satellites than mobile nodes, the satellite-terminal switching management problem needs to be solved. In scenarios with a large number of users and complex communication requirements, the traditional switching mechanism based on a single user not only may cause frequent switching and large signaling overhead problems, but also may lead to low switching success rate and load imbalance, threatening the reliability and efficiency of communication, so a group switching method is more appropriate.
[0004] Group switching is closely related to the group grouping method, but the current traditional grouping method may have problems such as uneven resource allocation and unreasonable group division. In addition, individual grouping methods are relatively complex, and as the number of users increases, the algorithm complexity is too large, making it difficult to implement. In addition, in the current grouping method, the situation of emergency communication demand is rarely considered, which cannot effectively ensure the continuity and stability of emergency communication. SUMMARY
[0005] To solve the above problems, the application provides a terminal grouping switching method in a low-orbit satellite Internet of Things scenario, which includes the following steps:
[0006] S1. Constructing a multi-satellite coverage communication system model including a plurality of user terminals and a plurality of low-orbit satellites, wherein a plurality of fixed wave positions are also included;
[0007] S2. For the user terminals in each wave position, calculating a similarity matrix based on bandwidth demand, service priority, and location coordinates;
[0008] S3. According to the similarity matrix, the user terminals are initially divided to obtain a plurality of initial groups;
[0009] S4. According to the predicted switching success rate, the predicted satellite load, the predicted switching delay, and the actual service priority, constructing a utility function for each user terminal;
[0010] S5. Based on the initial grouping, in-group optimization and inter-group optimization switching are performed according to the utility function to obtain a final grouping switching result.
[0011] The present application has the following advantages:
[0012] The present application is a user grouping switching method based on potential game in a low-orbit satellite Internet of Things scenario. The method focuses on a ground fixed cell, wherein satellite beams hop according to ground-specified beam positions to ensure consistency of switching time. In order to make user terminals grouped into the same group close in terms of elevation angle and signal strength, one of the grouping attributes selects the position coordinates of the user terminal relative to the beam center point of the beam position in which the user terminal is located. In addition, in the method of single user switching, users tend to select the candidate satellite that best meets their own needs for switching, and user terminals with similar bandwidth requirements may select the same satellite. Therefore, in order to divide these users into a group, the second grouping attribute selects the bandwidth requirement of the user. When an emergency communication requirement occurs, in order to improve the continuity and stability of communication services, the last grouping attribute selects the service priority of the user. The position attribute, bandwidth requirement and service priority are three very important factors in satellite communication. Compared with other grouping switching methods, the grouping attribute selection of the present application is more concise and reasonable, can better adapt to various satellite communication scenarios, and improve the accuracy and practicality of grouping.
[0013] In the present application, the user is initially grouped using a spectral clustering algorithm, which can fully consider the similarity of users in terms of position attribute, bandwidth requirement and service priority, so that the grouping result is more accurate, and users similar in each attribute are grouped into a group, which can improve the stability and reliability of grouping. At the same time, as the number of users increases, the spectral clustering algorithm can also maintain good performance, thereby effectively reducing the computational load and complexity of the algorithm.
[0014] In the present application, under the framework of game theory, the switching success rate, service priority, satellite load and switching delay are comprehensively considered and a utility function of user switching to a satellite is established. Then, grouping optimization and scheduling switching are performed according to the utility function, and finally user grouping switching is completed. According to the utility function value of the user, the overall utility of the system can be maximized, and the signaling overhead, resource allocation and performance of the overall system are optimized. BRIEF DESCRIPTION OF DRAWINGS
[0015] Figure 1 A flowchart of a user grouping switching method based on potential game in a low-orbit satellite Internet of Things scenario of the present application;
[0016] Figure 2 A ground fixed beam position division and multi-satellite coverage schematic diagram of the present application;
[0017] Figure 3 The switching signaling interaction flow chart for the user of the application;
[0018] Figure 4 The switching flow chart for the user in the group of the application. DETAILED DESCRIPTION
[0019] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative work are within the protection scope of the application.
[0020] The application provides a terminal grouping switching method in a low-orbit satellite Internet of Things scenario, as shown in the following formula (1) : Figure 1 The method comprises the following steps:
[0021] S1. Constructing a multi-satellite coverage communication system model, which comprises a plurality of user terminals and a plurality of low-orbit satellites.
[0022] Specifically, the multi-satellite coverage communication system model based on low-orbit satellites and a large number of ground user terminals is as shown in the following formula (2) : Figure 2 Since the user terminals in the satellite Internet of Things scenario have low mobility, the motion speed thereof can be ignored compared with the low-orbit satellites, a ground fixed wave position division mode is adopted, which can greatly reduce the overhead of position area update compared with a dynamic position area. The ground fixed wave position division mode comprises a plurality of wave positions in the coverage area of the low-orbit satellites, each wave position is fixed relative to the position of the earth, as shown in the following formula (3) : Figure 2 The dashed box represents the coverage area of the satellite, and the solid box represents the divided wave position; the beams of the satellites S1, S2, S3 and S4 serve the users in a wave position through system scheduling in the service time, since the ground is a fixed wave position, when the users in the current wave position need to be switched, the current service beam will be scheduled to other wave positions to continue to serve, for the users in the wave position, the change and scheduling of the beam occur at the same time. In addition, as shown in the following formula (4) : Figure 2 Since it is multi-satellite coverage, the beams of other satellites may also cover the wave position covered by the current satellite through scheduling.
[0023] S2. For each user terminal in a wave position, a similarity matrix is calculated based on bandwidth demand, service priority and position coordinates.
[0024] Specifically, as shown in the following formula (5) : Figure 3As shown, during the switching process, the user terminal, the service satellite and the target satellite will generate signaling interaction. If it is the switching process of n single user terminals, n signaling will be generated in the signaling interaction between any two of the user terminal, the service satellite and the target satellite, which will cause a large amount of signaling overhead. If the n user terminals are regarded as a group, except the signaling related to the information operation of the user terminal, the signaling of other information will be greatly reduced. Therefore, the grouping after switching can significantly reduce the repeated signaling transmission, thereby reducing the signaling overhead of the system.
[0025] Users with similar attributes tend to submit switching requests to the same satellite in a similar or same time period. Therefore, for the grouping of user terminals, how to select the relevant attributes as the basis for grouping user terminals is very important. In order to adapt to the satellite communication scene and improve the grouping accuracy, the present application considers the following aspects:
[0026] Considering the ground fixed wave position, when the satellite occurs beam switching, the beam jumps according to the wave position divided by the ground, so the switching time is considered to be consistent. In order to make the user terminals grouped into the same group close in terms of attributes such as elevation angle and signal strength, the position coordinates of the wave position center point of the user terminal relative to the wave position in which the user terminal is located are selected as one of the grouping attributes.
[0027] In the single user switching method, the user terminal tends to select the best candidate satellite that meets its own needs for switching, and user terminals with close bandwidth requirements may initiate switching to the same target satellite. Therefore, in order to divide these user terminals into the same group, the bandwidth requirement of the user terminal is selected as one of the grouping attributes.
[0028] Considering the emergency communication demand, in order to be able to preferentially process those user terminals with high priority, improve the continuity and stability of communication service, and minimize the impact on user communication caused by possible delay or interruption in the switching process; therefore, the service priority of the user terminal is selected as one of the grouping attributes.
[0029] Based on the above three grouping attributes, the comprehensive similarity between user terminals is calculated, which provides a basis for subsequent grouping.
[0030] The specific process of step S2 for calculating the similarity matrix of the user terminals in a wave position includes:
[0031] S21. Obtain the position coordinates of each user terminal relative to the wave position center point of the wave position in which the user terminal is located, calculate the position similarity between the user terminals, and represent it as
[0032]
[0033] Wherein, sim(i,j) represents the location similarity between user terminal i and user terminal j, ε d represents a constant, d(i,j) represents the Euclidean distance between user terminal i and user terminal j, (x i ,y i ) represents the location coordinates of user terminal i, and (x j ,y j ) represents the location coordinates of user terminal j.
[0034] S22. Obtain the bandwidth requirement of each user terminal, calculate the bandwidth similarity between each pair of user terminals, and represent it as
[0035]
[0036] wherein, sim(i,j) represents the bandwidth similarity between user terminal i and user terminal j, ε b represents a constant, b i represents the bandwidth requirement of user terminal i.
[0037] S23. Obtain the service priority of each user terminal, calculate the service priority similarity between each pair of user terminals, and represent it as
[0038]
[0039] wherein, sim(i,j) represents the service priority similarity between user terminal i and user terminal j, ε e represents a constant, e i represents the service priority of user terminal i.
[0040] S24. According to the location similarity, the bandwidth similarity and the service priority similarity, calculate the comprehensive similarity between each pair of user terminals, and represent it as
[0041]
[0042] wherein, sim(i,j) represents the comprehensive similarity between user terminal i and user terminal j, and α, β, γ represent weight coefficients, wherein α+β+γ=1; in general scenarios, α=β=γ=1 / 3, that is, the three grouping attributes are equally important. If there is an urgent communication demand (such as earthquake, emergency rescue, etc.), the weight coefficient of γ can be increased to ensure the priority, continuity and stability of the emergency communication service.
[0043] S25. Construct a similarity matrix according to the comprehensive similarity, which can be represented as
[0044]
[0045] The element of the i-th row and the j-th column in the similarity matrix represents the comprehensive similarity between the user terminal i and the user terminal j.
[0046] Specifically, the application divides the service priority of the user terminal into four levels: emergency, high, medium and low.
[0047] S3. The user terminals are initially divided according to the similarity matrix to obtain a plurality of initial groups.
[0048] Specifically, step S3 specifically comprises:
[0049] S31. A corresponding degree matrix D is constructed according to the similarity matrix S, wherein D is a diagonal matrix, and the diagonal elements are the row sum (or column sum) of the similarity matrix S; then a Laplacian matrix L is calculated according to the formula L=D-S;
[0050] S32. The Laplacian matrix L is subjected to eigenvalue decomposition to obtain eigenvalues and eigenvectors, and the eigenvectors are normalized to form a feature matrix V;
[0051] S33. The rows (i.e. users) of the feature matrix V are clustered by using a K-means algorithm to obtain a plurality of initial groups.
[0052] The application mainly performs grouping switching on all user terminals in each wave position, each wave position can be covered by different low-orbit satellites, and multiple low-orbit satellites exist in the same wave position to simultaneously perform coverage. Therefore, when the user terminals in a wave position are initially grouped and divided, the application embodiment obtains the same number of initial groups as the number of low-orbit satellites covering the wave position.
[0053] S4. The utility function of each user terminal is constructed according to the predicted switching success rate, the predicted satellite load, the predicted switching delay and the actual service priority.
[0054] Specifically, the switching after the group grouping can be regarded as a process in which multiple groups compete for satellite resources, which is a potential game, and each user terminal has a utility function and always selects a strategy that optimizes its own utility. In order to further optimize the initial groups, the utility function is set up for intra-group and inter-group optimization and scheduling, and the utility function is composed of a gain function and a loss function. The reason for selecting the utility function is to integrate the switching criteria after grouping, so as to not only guarantee the switching success rate and the needs of emergency communication, but also balance the load and reduce the delay.
[0055] The utility function of each user terminal is expressed as
[0056] u i,k =g i,k -l i,k
[0057]
[0058] wherein, u i,k represents the utility function of user terminal i switching to low-orbit satellite k, g i,k represents the gain function of user terminal i switching to low-orbit satellite k, l i,k represents the loss function of user terminal i switching to low-orbit satellite k; a1, a2, a3, a4 are weight coefficients, and a1+a2+a3+a4=1; represents the predicted switching success rate of user terminal i switching to low-orbit satellite k; e i represents the actual service priority of user terminal i; represents the predicted satellite load of low-orbit satellite k after a plurality of user terminals i switch to low-orbit satellite k; represents the predicted switching delay of user terminal i switching to low-orbit satellite k; in addition, it is assumed that the predicted data can correctly reflect the results of user terminals in the current environment.
[0059] The application is based on historical data, and a machine learning model is used to predict the switching success rate, satellite load, and switching delay.
[0060] Specifically, the purpose of grouping is mainly to improve the success rate of user terminal switching. The historical switching success rate SuccessRate i,k can be expressed by the following formula
[0061]
[0062] The predicted switching success rate can be predicted by a classification model:
[0063]
[0064] wherein, h i,k represents the case that user terminal i initiates a switching attempt from the serving satellite to target satellite k, if the switching is successful, h i,k =1, otherwise h i,k =0; N k represents the number of user terminals initiating a switching attempt to target satellite k; represents the number of user terminals successfully switching to target satellite k; f represents a classification model.
[0065] Specifically, after completing the switching, in order to ensure that the bandwidth utilization rate of each satellite is as balanced as possible, the satellite load is defined as the ratio of the used bandwidth to the maximum bandwidth after the user terminal switches to the target satellite, and the historical satellite load LoadBalance i,k can be expressed by the following formula
[0066]
[0067] Predicting satellite load The prediction can be made by a regression model:
[0068]
[0069] where B max represents the maximum bandwidth of the low earth orbit satellite, B k represents the available bandwidth of the low earth orbit satellite k, b i represents the bandwidth requirement of the user terminal i. The smaller the satellite load, the smaller the bandwidth usage of the satellite. h represents the regression model.
[0070] Specifically, the switching delay problem caused by the grouping switching is also worth attention, considering the transmission delay and processing delay in the grouping switching process. It is assumed that the entire process from the initiation of the switching request to the completion of the switching action involves k links, and the transmission delay of each link is d. In addition, if the satellite can process x switching requests per second, and the actual number of arriving switching requests is n, then the historical switching delay Delay i,k The switching delay can be expressed by the following formula:
[0071]
[0072] Predicting switching delay The prediction can be made by a regression model:
[0073]
[0074] where z represents the regression model.
[0075] S5. Based on the initial grouping, intra-group optimization and inter-group optimization switching are performed according to the utility function, to obtain the final grouping switching result.
[0076] Specifically, as shown in Figure 4 , step S5 specifically includes:
[0077] S51. Obtain the initial grouping, and initialize the iteration number l = 1.
[0078] S52. Each user terminal can have multiple alternative low earth orbit satellites at each switching, so in the lth iteration, the utility function value of each user terminal switching to different low earth orbit satellites (i.e. each alternative low earth orbit satellite corresponding to the user terminal) is calculated, the maximum utility function value is selected as the final utility function value (which also shows the low earth orbit satellite that the user terminal finally selects to switch), the sum of the utility function values of each grouping is calculated according to the final utility function value, and the grouping with the highest sum of utility function values is determined; wherein in the first iteration, the initial grouping is directly calculated.
[0079] Suppose we have N group groups, each group containing multiple user terminals, then the sum of utility function values U j of each group G group,j can be expressed as:
[0080]
[0081] Determine the group with the highest sum of utility function values:
[0082]
[0083] S53. For each user terminal in the group with the highest sum of utility function values, switch it to the low earth orbit satellite corresponding to its own final utility function value, and record this low earth orbit satellite as the switching satellite.
[0084] S54. Update the historical satellite load of each low earth orbit satellite (at this time, only for the low earth orbit satellites covering the beam currently undergoing group switching), determine whether the historical satellite load of the switching satellite is greater than 1.5 times the average historical satellite load, if so, perform step S55, otherwise perform step S56;
[0085] S55. For each switching satellite, arrange all user terminals served by it in descending order of final utility function value, remove the user terminals switched in this iteration, and select the user terminal corresponding to the largest final utility function value for switching, and return to step S54;
[0086] S56. Determine whether the system has reached a local Nash equilibrium, if so, obtain the final group switching result for the current beam; otherwise, let l = l + 1, and group all user terminals served by each low earth orbit satellite in the current beam, and return to step S52.
[0087] In the present application, unless otherwise explicitly specified and limited, the terms "mount", "set", "connect", "fix", "rotate" and the like should be understood broadly, for example, can be fixedly connected, or can be detachably connected, or can be integrated; can be mechanically connected, or can be electrically connected; can be directly connected, or can be indirectly connected through an intermediate medium, can be the internal communication of two elements or the interaction relationship between two elements, unless otherwise explicitly limited, the above-mentioned terms in the present application can be understood according to the specific meaning according to the specific circumstances by those skilled in the art.
[0088] Although embodiments of the present application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present application, the scope of the present application being defined by the appended claims and their equivalents.
Claims
1. A method for terminal group handover in a low earth orbit satellite Internet of Things scenario, characterized in that, Comprising the following steps: S1. Constructing a multi-satellite coverage communication system model comprising a plurality of user terminals and a plurality of low earth orbit satellites, wherein a plurality of fixed wave positions are also included; S2. For each user terminal in a wave position, calculating a similarity matrix based on bandwidth demand, service priority and location coordinates; the service priority of the user terminal is divided into four levels: emergency, high, medium and low; Step S2 specifically comprises: S21. Obtaining the location coordinates of each user terminal relative to the center point of the wave position in which it is located, calculating the location similarity between each pair of user terminals, denoted as wherein, represents a position similarity between user terminal i and user terminal j, represents a constant, d(i,j) represents the Euclidean distance between user terminal i and user terminal j, (x i ,y i ) represents the position coordinates of user terminal i, and (x j ,y j ) represents the position coordinates of user terminal j; S22. Obtaining the bandwidth demand of each user terminal, calculating the bandwidth similarity between each pair of user terminals, denoted as wherein, denotes the bandwidth similarity between user terminal i and user terminal j, denotes a constant, b i denotes the bandwidth demand of user terminal i; S23. Obtaining the service priority of each user terminal, calculating the service priority similarity between each pair of user terminals, denoted as wherein, represents the service priority similarity between user terminal i and user terminal j, represents a constant, e i represents the service priority of user terminal i; S24. According to the location similarity, bandwidth similarity and service priority similarity, calculating the comprehensive similarity between each pair of user terminals, denoted as Wherein, sim(i, j) represents the comprehensive similarity between the user terminal i and the user terminal j, and a, b, g represent weight coefficients, wherein a+b+g=1; in a general scenario α = β = S25. According to the comprehensive similarity, constructing a similarity matrix, the element in the i-th row and j-th column of the similarity matrix represents the comprehensive similarity between user terminal i and user terminal j; =1 / 3, if there is an emergency communication demand, the weight coefficient of S3. According to the similarity matrix, performing initial division on the user terminals to obtain a plurality of initial groups; is increased. S4. According to the predicted handover success rate, the predicted satellite load, the predicted handover delay and the actual service priority, constructing the utility function of each user terminal; The utility function of each user terminal in step S4 is denoted as S5. Based on the initial groups, performing intra-group optimization and inter-group optimization handover according to the utility function to obtain the final group handover result. In the multi-satellite coverage communication system model, for each low earth orbit satellite, there are a plurality of wave positions in its coverage area, and each wave position is fixed relative to the position of the earth. wherein, u i,k represents the utility function of the user terminal i switching to the low earth orbit satellite k, g i,k represents the gain function of the user terminal i switching to the low earth orbit satellite k, l i,k represents the loss function of the user terminal i switching to the low earth orbit satellite k; a1, a2, a3, a4 are weight coefficients, and a1+a2+a3+a4=1; represents the predicted switching success rate of the user terminal i switching to the low earth orbit satellite k; e i represents the actual service priority of the user terminal i; represents the predicted satellite load of the low earth orbit satellite k after the multiple user terminals i switch to the low earth orbit satellite k; represents the predicted switching delay of the user terminal i switching to the low earth orbit satellite k; Step S3 specifically comprises: 2.The terminal grouping handover method in a low earth orbit satellite Internet of Things scenario of claim 1, wherein S31. Constructing a corresponding degree matrix according to the similarity matrix, and then calculating a Laplacian matrix; 3.The method of claim 1, wherein, S32. Performing eigenvalue decomposition on the Laplacian matrix to obtain eigenvalues and eigenvectors, and then normalizing the eigenvectors to form a feature matrix; S33. Using the K-means algorithm to cluster the rows of the feature matrix to obtain a plurality of initial groups. Step S5 specifically comprises: S51. Obtaining the initial groups, initializing the iteration number l=1; 4. The terminal grouping switching method in a low-orbit satellite Internet of Things scene according to claim 1, characterized in that, Predicting handover success rate is represented as: where SuccessRate i,k denotes the historical handover success rate of user terminal i to low earth orbit satellite k, h i,k denotes the case that user terminal i initiates a handover attempt from a serving satellite to target satellite k, h i,k = 1 if the handover is successful, otherwise h i,k = 0; N k denotes the number of user terminals that initiate a handover attempt to target satellite k; denotes the number of user terminals that successfully handover to target satellite k; f denotes a classification model; LoadBalance i,k denotes the historical satellite load of low earth orbit satellite k, which is calculated as Predicting satellite load is represented as where B max denotes the maximum bandwidth of a low earth orbit satellite, B k denotes the available bandwidth of a low earth orbit satellite k, b i denotes the bandwidth requirement of a user terminal i, h denotes a regression model; Predicted handover delay is represented as: where Dealy i,k where Dealy i,k represents the history handover delay, k represents the number of link segments; d represents the average transmission delay of each link segment; x represents the number of handover requests that the satellite can process per second; n represents the actual number of handover requests arrived; and z represents the regression model.
5. The terminal grouping handover method in a low earth orbit satellite Internet of Things scenario according to claim 1, characterized in that, S52. In the l-th iteration, calculating the utility function value of each user terminal switching to different low earth orbit satellites, selecting the maximum utility function value as the final utility function value of itself, calculating the utility function value sum of each group according to the final utility function value, and determining the group with the highest utility function value sum; S53. For each user terminal in the group with the highest utility function value sum, switching it to the low earth orbit satellite corresponding to its final utility function value, and recording this low earth orbit satellite as the switching satellite; S54. Updating the historical satellite load of each low earth orbit satellite, and judging whether the historical satellite load of the switching satellite is greater than 1.5 times the average historical satellite load, if yes, executing step S55, otherwise executing step S56; S55. For each switching satellite, arrange all user terminals served by it in descending order of the final utility function value, select the user terminal corresponding to the maximum final utility function value after eliminating the user terminals switched in this iteration, and return to step S54; S56. Determine whether the system reaches a local Nash equilibrium. If yes, the final grouping switching result is obtained, otherwise, let l = l + 1, and form a group by all user terminals served by each low earth orbit satellite in the current wave position; and return to step S52.