A QoS-oriented on-board communication resource allocation method based on joint clustering and intra-cluster beam hopping

Abstract

The present invention discloses a QoS-oriented on-board communication resource allocation method for joint clustering and intra-cluster hopping, establishes a satellite-ground joint communication resource allocation architecture, and the ground station shares the burden of the on-board decision-making clustering scheme. In addition, the ground station can adaptively adjust the duration of the clustering pattern according to the existing information, so that the business demand between each cluster is in a dynamic equilibrium state for a long time with a lower update frequency. In the intra-cluster hopping strategy of this patent, the multi-beam satellite dispatches communication resources on time and on demand according to the business demand and delay requirements of the wave position cell, improves the throughput of the multi-beam satellite communication system, reduces the waste of communication resources, and ensures the QoS performance of the wave position cell within the coverage area. The intra-cluster hopping strategy dispatches communication resources on time and on demand according to the business demand and delay requirements of the wave position cell, improves the communication resource utilization rate of the multi-beam satellite, and ensures the QoS performance of the wave position cell within the coverage area.

Description

Technical Field

[0001] The present invention relates to the field of multi-beam satellite communications, and more specifically to a multi-beam satellite communication resource allocation scheme that coordinates clustering and intra-cluster beam hopping in a multi-beam satellite communication system, and belongs to the category of communication resource allocation technology in microwave communications. Background Art

[0002] As a technology for resource allocation in multi-beam satellite communication systems, beam hopping technology realizes the redistribution of on-board power and bandwidth resources from a time domain perspective, thereby improving the resource utilization of multi-beam satellite communication systems. Clustering divides the wave cells under the coverage of multi-beam satellites into different clusters according to certain rules. Intra-cluster beam hopping uses beam hopping technology to schedule the communication resources allocated to each cluster by multi-beam satellites. Joint clustering and intra-cluster beam hopping are beneficial to the management of beams by multi-beam satellite communication systems and improve the utilization of on-board communication resources. However, the existing clustering and intra-cluster beam hopping lack consideration for the uneven service demand between clusters, which may lead to "uneven work and rest" of the working beams between clusters, resulting in the inability to fully utilize the idle communication resources within a cluster.

[0003] The patent "Service-Adaptive Multi-Beam Satellite Beam Hopping System Resource Allocation Method" applied by Ding Xiang et al. proposes to allocate idle communication resources of each cluster based on the intra-cluster beam hopping results and the remaining service demand. However, in this scheme, intra-cluster beam hopping only considers the service demand of the wave-position cell, and lacks consideration of the service demand delay, resulting in the inability to meet the service demand of the wave-position cell in a timely manner.

[0004] The patent "High-throughput beam-hopping scheduling method for differentiated services" applied by Bai Weigang et al. proposes to divide clusters according to the principle of balancing business demands between clusters and the principle of geographical proximity of cells within clusters, and consider the delay of business demands in the process of resource allocation within clusters. However, in this patent, the multi-beam satellite communication system needs to frequently update the clustering scheme, which will only increase the computational burden on the satellite.

[0005] Therefore, the present invention designs an on-board communication resource allocation scheme for joint clustering and intra-cluster hopping for QoS. The scheme establishes a joint satellite-ground communication resource allocation architecture, and the ground station shares the burden of the on-board decision-making clustering scheme. In addition, the ground station can adaptively adjust the duration of the clustering pattern based on the existing information, so that the service demand between clusters is in a dynamic equilibrium state for a long time with a lower update frequency. In the intra-cluster hopping strategy of this patent, the multi-beam satellite dispatches communication resources on time and on demand according to the service demand and delay requirements of the wave position cell, improves the throughput of the multi-beam satellite communication system, reduces the waste of communication resources, and ensures the QoS performance of the wave position cell within the coverage area. Summary of the invention

[0006] The purpose of the present invention is to overcome the shortcomings of the above-mentioned prior art, and propose a QoS-oriented on-board communication resource allocation method for joint clustering and intra-cluster hopping, establish a satellite-ground joint communication resource allocation architecture, alleviate the on-board computing burden, and reduce the update frequency of clustering patterns. In addition, the intra-cluster hopping strategy proposed in this scheme can schedule communication resources on time and on demand according to the service needs and delay requirements of the wave position cell, improve the communication resource utilization rate of the multi-beam satellite, and ensure the QoS performance of the wave position cell within the coverage area.

[0007] The key point of the present invention is to establish a satellite-ground joint communication resource allocation architecture, provide a service-adaptive clustering scheme and a cluster-hopping strategy for timely and on-demand allocation. In the satellite-ground joint communication resource allocation architecture, the ground station shares part of the on-board computing burden. The service-adaptive clustering scheme maintains a long-term balance of inter-cluster service demand with a small number of cluster pattern updates. The cluster-hopping strategy for timely and on-demand allocation schedules communication resources in a timely manner, improves the throughput of multi-beam satellites, and ensures the QoS performance of the wave position cells under the coverage of multi-beam satellites.

[0008] The technical solution adopted by the present invention is a satellite communication resource allocation method for joint clustering and intra-cluster hopping for QoS. The satellite-ground joint communication resource allocation architecture for implementing the method is divided into two parts: (1) the ground gateway determines the clustering scheme and (2) the multi-beam satellite decides the intra-cluster beam hopping strategy. The clustering scheme includes a clustering pattern and the duration of the clustering pattern. In the first part, the ground gateway observes and counts the service demand of the wave position cells within the coverage area of ​​the multi-beam satellite, and determines the clustering pattern based on the principle of balancing the service demand between clusters and the geographical proximity of the wave position cells within the cluster. Afterwards, the ground station adaptively adjusts the duration of the clustering pattern according to the prediction information and broadcasts the scheme to the multi-beam satellite. For the second part, the multi-beam satellite executes the clustering scheme from the ground station, distributes the power and bandwidth equally to the clusters, and the bandwidth between clusters adopts four-color frequency division multiplexing. Then, the on-board beam hopping control center schedules the time slots and bandwidth resources of each cluster on time according to the service demand and delay requirements of each cluster.

[0009] (1) Business-adaptive clustering solution;

[0010] Assume that at time t, the clustering scheme needs to be updated for the kth time. The ground station divides N cells into M clusters according to the principle of balanced service demand between clusters and the principle of geographical proximity of cells within a cluster. The number of cells in each cluster is At time t, the service demand of cell j is The business demand balance function E between clusters t It is expressed as:

[0011]

[0012] is the average value of the business demand among clusters. Indicates whether the cell j belongs to cluster i at time t. If the cell j belongs to cluster i, otherwise,

[0013] The duration of the clustering pattern is defined as the interval between two broadcasts of the clustering scheme by the ground station to the multi-beam satellite. The duration of the kth clustering pattern is denoted as B k . B k ≤I F , I F is the prediction duration threshold, I F The internal prediction algorithm accurately predicts the service demand of each cluster. F Taking time step b as the unit, the mean value of service demand of each cluster within b remains unchanged. F Can be decomposed into is the period from (f-1)b to fb. The service demand forecast information of cluster i is Calculated based on the ground station's prediction information for each cluster Indicates that under the current clustering pattern The predicted inter-cluster business demand equilibrium function value within the time period. F and E t Get the change of the balance function value of each cluster's business demand Indicates that under the current clustering pattern Balanced changes in business demand during the period. The value of is expressed as:

[0014]

[0015] ε determines the difference range of the maximum business demand balance, when With E t When the business demand difference is within this range, it is considered It is feasible to execute the clustering pattern at time t in the time period. The value is 1, otherwise, The value is 0.

[0016] The clustering pattern is broadcasted by the ground station to the multi-beam satellite. Therefore, the continuous update of the clustering pattern will cause a large signaling overhead between the satellite and the ground. In order to reduce the update frequency of the clustering pattern while maintaining a dynamic balance of service requirements between clusters, we Adaptively determine the duration of the clustering pattern B k for:

[0017]

[0018] Indicated in arrive It is feasible to execute the clustering pattern at time t within this continuous time period. for The maximum f value at B k The value is arrive This continuous period of time.

[0019] (II) Intra-cluster beam hopping strategy for on-time and on-demand allocation

[0020] B k At time t in time, whether the multi-beam satellite provides communication resources to the beam position cell n in cluster i under the coverage range is expressed as It indicates that at time t, the multi-beam satellite does not provide communication resources to the cell n; It indicates that at time t, the multi-beam satellite provides communication resources to the wave position cell n. The bandwidth allocated by the multi-beam satellite to the wave position cell n is In order to reduce the service failure rate and the waste of communication resources, the throughput provided by the multi-beam satellite to the wave position cell n at time t is The range is:

[0021]

[0022] At time t, the service failure rate of wave position cell n It can be defined as:

[0023]

[0024] Represents the service demand of cell n at time t. Can be expanded to For business needs that are about to become obsolete. Indicates time t, the delay tolerance is lT slot The business demand, l is the number of time slots, T slot It is the duration of each time slot. As each time slot passes, the corresponding delay tolerance will decrease by one time slot. If the delay of the service demand exceeds its tolerance, the service demand is not met and the service fails.

[0025] In order to realize the on-time and on-demand allocation of beam hopping strategy within the cluster, according to the principle of maximizing the throughput of each cluster of multi-beam satellites and minimizing the service failure rate, B k The hopping beam model for on-time and on-demand allocation within cluster i can be expressed as:

[0026]

[0027] α is the preset weight to achieve a trade-off between intra-cluster throughput and service failure rate, Φ max and r max are the normalization constants of intra-cluster throughput and service failure rate, respectively. k is the duration of the kth clustering pattern. C1 indicates that in each time slot, the multi-beam satellite only provides communication resources to the K cells in cluster i. C2 gives the throughput provided by the multi-beam satellite to cell n. The value range of ensure Greater than the service demand that is about to fail in cell n ensure Less than the total service demand of wave cell n Reduces the waste of on-board communication resources.

[0028] The effects of the present invention are as follows: compared with the existing fixed clustering scheme and dynamic clustering scheme, the present scheme can adaptively adjust the clustering pattern and the duration of the clustering pattern, so that the inter-cluster demand is in a balanced state for a long time, and at the same time reduce the update frequency of the clustering scheme. Compared with the existing multi-beam satellite intra-cluster beam hopping resource allocation scheme, the intra-cluster beam hopping strategy of on-time and on-demand allocation proposed in the present scheme can flexibly allocate time slots and bandwidth resources according to the service requirements and delay requirements of the wave position cells within the coverage area, improve the multi-beam satellite throughput, reduce the service failure rate of the wave position cells, and ensure that communication resources can be allocated to the wave position cells on time and on-demand. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 Diagram of the satellite-ground joint communication resource allocation architecture.

[0030] Figure 2 Block diagram of the resource allocation process for satellite-ground joint communications.

[0031] Figure 3 A graph showing how inter-cluster business demand changes over time within an hour.

[0032] Figure 4 A graph showing how multi-beam satellite throughput varies with business demand.

[0033] Figure 5 A graph showing how the failure rate of multi-beam satellite services varies with service demand. DETAILED DESCRIPTION

[0034] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0035] like Figure 1As shown in the figure, the joint satellite-ground communication resource allocation architecture is divided into two parts: (1) the ground gateway determines the clustering scheme and (2) the multi-beam satellite decides the intra-cluster beam hopping strategy. The clustering scheme includes the clustering pattern and the duration of the clustering pattern. In the first part, the ground gateway observes and counts the service demand of the wave position cells within the coverage area of ​​the multi-beam satellite, and determines the clustering pattern based on the principle of balancing the service demand between clusters and the geographical proximity of the wave position cells within the cluster. Afterwards, the ground station adaptively adjusts the duration of the clustering pattern based on the predicted information and broadcasts the scheme to the multi-beam satellite. For the second part, the multi-beam satellite executes the clustering scheme from the ground station, distributes the power and bandwidth equally to the clusters, and uses four-color frequency division multiplexing for the bandwidth between clusters. Subsequently, the on-board beam hopping control center schedules the time slots and bandwidth resources of each cluster on time according to the service demand and delay requirements of each cluster.

[0036] In the experiment of this scheme, the altitude of the multi-beam satellite is 36000km. The total bandwidth of the multi-beam satellite is 500MHz, and the total power is 33dBW. The number of cells within the coverage of the multi-beam satellite is 49. The number of clusters is 7, and the number of working beams in each cluster in each time slot is 2. The time step b is △T = 5min, and the duration of each time slot is T slot The maximum number of service retention time slots is 10T. slot The population size is 10, the number of iterations is 200, the crossover probability is 0.6, and the mutation probability is 0.02.

[0037] like Figure 5 As shown in the figure, a satellite-ground joint communication resource allocation architecture is established, providing a service-adaptive clustering scheme and a cluster-hopping strategy for on-time and on-demand allocation. In the satellite-ground joint communication resource allocation architecture, the ground station shares part of the on-board computing burden. The service-adaptive clustering scheme maintains a long-term balance of inter-cluster service demand with a small number of cluster pattern updates. The on-time and on-demand allocation of intra-cluster hopping strategy timely schedules communication resources, improves the throughput of multi-beam satellites, and ensures the QoS performance of the wave position cells under the coverage of multi-beam satellites.

[0038] Under the above simulation conditions, this example simulates the multi-beam satellite beam strategy, obtains the number of updates of the clustering scheme within one hour, and the change of the inter-cluster service demand balance function over time, and compares it with the fixed clustering scheme, the short-duration clustering scheme, and the long-duration clustering scheme, as shown in Tables 1 and 2. Figure 3 In addition, in the example, the throughput and service failure rate of the multi-beam satellite are obtained, and compared with the existing intra-cluster beam hopping strategy, Figure 4 and Figure 5 From Table 1 and Figure 3 It can be seen from the figure that the service adaptive clustering solution of this patent achieves long-term service demand balance between clusters with a low number of updates. Figure 4 and Figure 5 It can be seen that, compared with other inter-cluster beam hopping strategies, the solution proposed in the present invention can further improve the throughput of multi-beam satellites, reduce the service failure rate of multi-beam satellites, and ensure the QoS of the beam cell.

[0039] Table 1 Comparison of the number of updates in one hour for each clustering scheme

[0040]

Claims

1. A QoS-oriented on-board communication resource allocation method for joint clustering and intra-cluster hopping. The satellite-ground joint communication resource allocation architecture for implementing this method is divided into two parts: (1) the clustering scheme determined by the ground gateway station and (2) the intra-cluster hopping strategy determined by the multi-beam satellite. The clustering scheme includes the clustering pattern and the duration of the clustering pattern; The ground gateway observes and counts the service demand of the cells within the coverage of the multi-beam satellite, and determines the clustering pattern based on the principle of balanced service demand between clusters and the geographical proximity of the cells within the cluster; Afterwards, the ground station adaptively adjusts the duration of the clustering pattern according to the predicted information and broadcasts the plan to the multi-beam satellite. The multi-beam satellite executes the clustering plan from the ground station, distributes power and bandwidth equally to the clusters, and uses four-color frequency division multiplexing for the bandwidth between clusters. Then the on-board beam hopping control center schedules the time slots and bandwidth resources of each cluster on time according to the service needs and delay requirements of each cluster. The method is characterized in that the steps of implementing the method are as follows: The adaptive clustering scheme for S1 service is as follows: Assume that at time t, the clustering scheme needs to be updated for the kth time; the ground station divides N wave position cells into M clusters according to the principle of balanced service demand between clusters and the principle of geographical proximity of wave position cells within a cluster. The number of wave position cells in each cluster is At time t, the service demand of cell j is The business demand balance function E between clusters t It is expressed as: is the average value of the business demand among clusters; Indicates whether the cell j belongs to cluster i at time t. If the cell j belongs to cluster i, otherwise, The duration of the clustering pattern is defined as the interval between two broadcasts of the clustering scheme by the ground station to the multi-beam satellite; the duration of the kth clustering pattern is recorded as B k ; B k ≤I F , I F is the prediction duration threshold, I F The internal prediction algorithm accurately predicts the service demand of each cluster; F Taking time step b as the unit, the mean value of service demand of each cluster within b remains unchanged; F Decompose into is the period from (f-1)b to fb; the service demand forecast information of cluster i is Calculated based on the ground station's prediction information for each cluster Indicates that under the current clustering pattern The predicted inter-cluster business demand equilibrium function value within the time period; F and E t Get the change of the balance function value of each cluster's business demand Indicates that under the current clustering pattern Balanced changes in business demand during time periods; The value of is expressed as: ε determines the difference range of the maximum business demand balance, when With E t When the business demand difference is within this range, it is considered It is feasible to execute the clustering pattern at time t in the time period. The value is 1, otherwise, The value is 0; The clustering pattern is broadcasted by the ground station to the multi-beam satellite. In order to reduce the update frequency of the clustering pattern while maintaining the dynamic balance of the service demand between clusters, Adaptively determine the duration of the clustering pattern B k for: Indicated in arrive It is feasible to execute the clustering pattern at time t within this continuous time period. for The maximum f value at B k The value is arrive This continuous period of time; S2 Intra-cluster Bundle Hopping Strategy Based on Time and Demand Allocation B k At time t in time, whether the multi-beam satellite provides communication resources to the beam position cell n in cluster i under the coverage range is expressed as It indicates that at time t, the multi-beam satellite does not provide communication resources to the cell n; It indicates that at time t, the multi-beam satellite provides communication resources to the wave position cell n; the bandwidth allocated by the multi-beam satellite to the wave position cell n is In order to reduce the service failure rate and the waste of communication resources, the throughput provided by the multi-beam satellite to the wave position cell n at time t is The range is: At time t, the service failure rate of wave position cell n Defined as: It represents the service demand of cell n at time t; Expand to For business needs that are about to become obsolete; Indicates time t, the delay tolerance is lT slot The business demand, l is the number of time slots, T slot It is the duration of each time slot. As each time slot elapses, the corresponding delay tolerance will be reduced by one time slot. If the delay of the service demand exceeds its tolerance, the service demand is not met and the service fails.

2. According to the QoS-oriented joint clustering and intra-cluster beam hopping on-board communication resource allocation method of claim 1, in order to realize the on-time and on-demand allocation of intra-cluster beam hopping strategy, according to the principle of maximizing the intra-cluster throughput of each multi-beam satellite and minimizing the service failure rate, B k The hopping beam model for on-time and on-demand allocation within cluster i is expressed as: α is the preset weight to achieve a trade-off between intra-cluster throughput and service failure rate, Φ max and r max are the normalized constants of intra-cluster throughput and service failure rate respectively; B k is the duration of the kth clustering pattern; C1 indicates that in each time slot, the multi-beam satellite only provides communication resources to the K cells in cluster i; C2 gives the throughput provided by the multi-beam satellite to cell n The value range of ensure Greater than the service demand that is about to fail in the wave cell n ensure Less than the total service demand of wave cell n Reduce the waste of onboard communication resources.

Images