A graph representation method supporting joint scheduling of observation transmission and computing resources and application thereof

Abstract

The application discloses a kind of chart representation method and its application supporting observation transmission and computing resource joint scheduling, mainly solve the problem that the prior art uses observation and transmission independent scheduling, transmission and computing independent scheduling, leading to the long time delay of ground observation task completion.Its implementation scheme includes: given the satellite network directed graph including multiple satellite nodes and the connection edge between node.According to its construction observation subgraph and two communication subgraphs;Observation subgraph is connected through observation satellite node between first communication subgraph;First communication subgraph is connected through computing satellite node between second communication subgraph, constitute graph representation model, so that observation, transmission and computing can be characterized in the same graph, support multi-dimensional resource joint scheduling.The application can map observation, transmission and computing joint scheduling problem from complex mathematical programming problem to routing problem in graph, reduce the complexity of problem solving, guarantee low latency demand, can be used for satellite network ground observation task.

Description

Technical Field

[0001] This invention belongs to the field of communication technology, and specifically relates to a graph representation method that can be used for Earth observation missions in satellite networks to perform efficient observation scheduling, transmission scheduling, and computation scheduling. Background Technology

[0002] Earth observation involves the selection of observation satellites, transmission links, and computing nodes. Different strategies can lead to excessively long completion times for satellite network Earth observation tasks, failing to meet the needs of emergency observations. However, existing Earth observation methods often employ independent scheduling of observation and transmission, and independent scheduling of transmission and computation, resulting in excessively long application times and failing to meet the low-latency requirements of emergency observations and other mission scenarios. There is an urgent need to construct a joint representation model of observation, transmission, and computation resources to support the joint scheduling of multi-dimensional resources, thereby reducing the completion time of Earth observation applications.

[0003] Patent document CN1168837048 discloses a method, device, and storage medium for collaborative Earth observation mission planning and optimization of a mega-constellation. It achieves functions such as dimensionality reduction of the imaging mission planning solution space, rapid generation of collaborative relationships, multiple coverage, and missing network replacement through core methods including target clustering, rapid routing coordination, and lightweight redundancy design. It also enables parallel processing of complex calculations, facilitating the rapid generation and iterative optimization of large-scale satellite mission planning schemes. However, because this method does not consider on-board computation and post-computation transmission, it leads to excessively long completion times for satellite network Earth observation missions, failing to meet the needs of emergency observations and impacting the service quality of the satellite network.

[0004] Patent document CN1061007198 discloses an efficient resource scheduling method for small satellite networks based on Earth observation missions. This method addresses the problems of low end-to-end throughput and low network resource utilization efficiency in resource-constrained small satellite network Earth observation scenarios, caused by insufficient consideration of network resources during the scheduling process. It improves resource utilization, increases the efficiency of data transmission for important tasks, and enhances the performance of small satellite Earth observation networks by allocating onboard resources without conflict and rationally planning data transmission methods. However, this resource scheduling method does not solve the problem of joint scheduling of observation transmission and computing resources, resulting in limited performance improvement for the satellite network and making it difficult to meet the needs of emergency observations. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of the prior art by providing a graph representation method that supports the joint scheduling of observation transmission and computing resources, thereby improving the performance of satellite networks, reducing the completion delay and solution complexity of Earth observation tasks, and meeting the needs of Earth observation tasks such as emergency observation.

[0006] To achieve the above objectives, the technical solution of the present invention includes: a graph representation method supporting joint scheduling of observation, transmission, and computing resources, and the application of the representation graph thereon, wherein:

[0007] 1. A graph representation method supporting joint scheduling of observation transmission and computing resources, characterized in that it includes:

[0008] (1) Given a snapshot of a satellite network G = (V, E) over a given time period, where V is the set of nodes and E is the set of edges;

[0009] (2) Divide the nodes of different types in the node set V into different node sets, namely the observation target node set S, the observation satellite node set V1, the calculation satellite node set V2, and the ground station node set D;

[0010] (3) Divide the connecting edges between different types of nodes in edge set E into different edge sets, namely, the connecting edge set E1 between the observation target node and the observation satellite node, and the connecting edge set E2 between any satellite node and any satellite or ground station node. The edges in edge set E1 have one time delay attribute, and the edges in E2 have two time delay attributes.

[0011] (4) Combine all nodes in the target node set E and the satellite node set V1 and all edges in the edge set E1 into an observation subgraph to characterize the observation process of the satellite on the target node.

[0012] (5) Combine all nodes in node set V except for the observation target node set S and all edges in edge set E2 to form the first communication subgraph for transmitting uncomputed observation data;

[0013] (6) Select the corresponding observation satellite node between the observation sub-map and the first communication sub-map, and connect the observation sub-map and the first communication sub-map through the observation satellite node;

[0014] (7) Combine all nodes in node set V except for the observation target node set S and all edges in edge set E2 to form a second communication subgraph for transmitting the calculated observation data;

[0015] (8) Select the corresponding computing satellite node between the first communication subgraph and the second communication subgraph, and then connect the two communication subgraphs together through the computing satellite node to form a graph model that can jointly schedule the observation, transmission and computing resources of the satellite network.

[0016] Furthermore, dividing the different types of nodes in the node set V into different node sets means dividing the multiple observation target nodes, multiple satellite nodes, and multiple ground station nodes included in the node set V into different node sets according to their different functions:

[0017] The observation target nodes in the observation task are used to form the observation target node set S.

[0018] The observation satellite node set V1 is composed of satellite nodes with observation capabilities.

[0019] A set of computing satellite nodes, V2, is composed of satellite nodes with computing capabilities.

[0020] The set of ground station nodes D is composed of the destination nodes located at the ground stations.

[0021] Furthermore, dividing the connecting edges between different types of nodes in edge set E into different edge sets means dividing the edges in edge set E into observation edges during the observation process and edges for transmitting observation data after observation, namely, the set of connecting edges between the observation target node and the observation satellite node E1, and the set of connecting edges between any satellite node and any satellite or ground station node E2.

[0022] 2. A method for applying representation graphs that supports joint scheduling of observation transmission and computational resources, characterized in that it includes:

[0023] When scheduling observation, transmission and computation resources simultaneously, the joint scheduling problem of observation, transmission and computation is first transformed into a routing problem from the observation target node in the observation subgraph to the destination node in the second communication subgraph through a representation graph.

[0024] Then, the end-to-end optimal path for the Earth observation mission is obtained through existing routing algorithms to support the low latency requirements in the Earth observation scenario, reduce the completion latency of the Earth observation mission, and reduce the complexity of solving the joint scheduling problem of network resources.

[0025] Compared with the prior art, the present invention has the following advantages:

[0026] Firstly, this invention integrates observation, communication, and computation within the same graph, enabling joint scheduling of multi-dimensional resources. Compared to independent scheduling methods for observation and communication, and independent scheduling methods for communication and computation, this reduces the completion delay of Earth observation tasks.

[0027] Secondly, by utilizing the graph model obtained in this invention, the joint scheduling problem of observation, communication and computing can be mapped from a complex mathematical programming problem to a routing problem in the graph, reducing the complexity of solving the joint scheduling problem of network resources and meeting the needs of Earth observation tasks such as emergency observation. Attached Figure Description

[0028] Figure 1 This is a flowchart illustrating the implementation of the present invention;

[0029] Figure 2 This is the input graph model in this invention;

[0030] Figure 3 This is the observation subgraph constructed in this invention;

[0031] Figure 4 This is the first communication subgraph constructed in this invention;

[0032] Figure 5 It is a connected graph of the observation subgraph and the first communication subgraph in this invention;

[0033] Figure 6 This is the second communication subgraph constructed in this invention;

[0034] Figure 7 It is a connected graph of the first communication subgraph and the second communication subgraph in this invention. Detailed Implementation

[0035] The embodiments of the present invention will now be described in further detail with reference to the accompanying drawings.

[0036] This embodiment addresses the issue that independent scheduling of observation, communication, and computation in Earth observation missions can all lead to excessive delays in mission completion. It proposes a graph representation method that supports joint scheduling of observation transmission and computational resources. This method transforms the joint scheduling problem of Earth observation missions into a routing problem by connecting observation scheduling, communication scheduling, and computational node scheduling using subgraphs, thereby supporting efficient problem-solving. The implementation flowchart is as follows: Figure 1 As shown.

[0037] Example 1: A graph representation method that supports joint scheduling of observation, transmission, and computing resources.

[0038] Reference Figure 1 The implementation steps of this example include the following:

[0039] Step 1: Given a snapshot of the satellite network over a given time period.

[0040] The snapshot diagram refers to the network topology diagram of the satellite network in an Earth observation scenario, where the satellite network topology remains unchanged within a certain period. It describes the nodes included in the current network and the connections between them. Subsequent processing of this input graph model yields a graph model capable of jointly scheduling the observation, transmission, and computing resources of the satellite network, such as... Figure 2 As shown, it is represented as: G=(V,E), where V is the set of nodes and E is the set of edges.

[0041] In this embodiment, V = {u,v,s,a,b,d}, which includes multiple observation target nodes, multiple satellite nodes, and multiple ground station nodes.

[0042] E = {(u,s),(v,s),(s,a),(s,b),(a,b),(a,d),(b,d)}, which contains the edges between all nodes.

[0043] The weights of each link in the diagram are l(u,s)=1, l(v,s)=2, l′(s,a)=5, l″(s,a)=1, l′(s,b)=6, l″(s,b)=2, l′(a,b)=6, l″(a,b)=1, l′(a,d)=4, l″(a,d)=2, l′(b,d)=5, l″(b,d)=1. Calculate the weight l of the satellite node. a =3,l b =2.

[0044] Step 2: Divide the multiple observation target nodes, multiple satellite nodes, and multiple ground station nodes included in the node set V into different node sets according to their different functions.

[0045] The observation target nodes u and v in the observation task are used to form the observation target node set S = {u, v}.

[0046] A set of observation satellite nodes V1 = {s} is formed by s satellite nodes with observation capabilities.

[0047] A set of computing satellite nodes V2 = {a, b} is formed using satellite nodes a and b that have computing capabilities.

[0048] The ground station node set D = {d} is composed of the destination node d located at the ground station.

[0049] The amount of observation data obtained by any observation satellite from observing all observation targets is the same, and the amount of calculated data obtained by any calculation satellite after processing the observation data is the same.

[0050] Step 3: Divide the edges connecting different types of nodes in edge set E into different edge sets.

[0051] 3.1) The first edge set is formed by connecting the observation target node and the observation satellite node.

[0052] E1 = {(u,s),(v,s)}

[0053] The edges in edge set E1 have a time delay property, which is used to characterize the observation time delay, wherein:

[0054] The edge (u,s) has a time delay attribute l(u,s) = 1, which is used to characterize the observation time delay of the observation satellite node s to the observation target node u;

[0055] The edge (v,s) has a time delay attribute l(v,s) = 2, which is used to characterize the observation time delay of the observation satellite node s to the observation target node v;

[0056] 3.2) Form a second edge set using the connecting edges between any satellite node and any satellite or ground station node:

[0057] E2={(s,a),(s,b),(a,b),(a,d),(b,d)},

[0058] The edges in edge set E2 have two delay properties: the communication delay required to transmit uncomputed observation data and the communication delay required to transmit computed observation data, wherein:

[0059] The edge (s, a) has a communication delay l required to characterize the transmission of uncomputed observation data between s and a. ′ (s,a)=5, and the communication delay l″(s,a)=1 required to transmit the calculated observation data between s and a;

[0060] The edge (s,b) has a communication delay l required to characterize the transmission of uncomputed observation data between s and b. ′ (s,b)=6, and the communication delay l″(s,b)=2 required to transmit the calculated observation data between s and b;

[0061] Edge (a, b) has a communication delay l required to characterize the transmission of uncomputed observation data between a and b. ′ (a,b) = 6, and the communication delay l″(a,b) = 1 required to transmit the calculated observation data between a and b;

[0062] The edge (a,d) has a communication delay l required to characterize the transmission of uncomputed observation data between a and d. ′ (a,d)=4, and the communication delay l″(a,d)=2 required to transmit the calculated observation data between a and d;

[0063] Edge (b,d) has a communication delay l required to characterize the transmission of uncomputed observation data between b and d. ′ (b,d) = 5, and the communication delay l″(b,d) = 1 required to transmit the calculated observation data between b and d.

[0064] Step 4: Construct the observation subgraph.

[0065] All nodes {u,v,s} in the target node set S and the satellite node set V1, and all edges {(u,s),(v,s)} in the first edge set E1 are placed in the same graph model to form an observation subgraph, such as... Figure 3 As shown;

[0066] This observation subgraph is used to characterize the observation process of the observation satellite on the target node.

[0067] Step 5: Combine all nodes in node set V except for the observation target node set S and all edges in edge set E2 to form the first communication subgraph.

[0068] 5.1) Select all nodes {s,a,b,d} in the node set V except for the observation target node set S, and rename the selected nodes to {s1,a1,b1,d1}.

[0069] 5.2) Select all edges {(s,a),(s,b),(a,b),(a,d),(b,d)} in edge set E2, and rename the selected edges to {(s1,a1),(s1,b1),(a1,b1),(a1,d1),(b1,d1)}, where the delay of each link is the communication delay required to transmit uncalculated observation data, i.e., l(s1,a1)=l′(s,a)=5, l(s1,b1)=l′(s,b)=6, l(a1,b1)=l′(a,b)=6, l(a1,d1)=l′(a,d)=4, l(b1,d1)=l′(b,d)=5;

[0070] 5.3) Combine all the nodes selected and renamed in step 5.1) and all the edges obtained in step 5.2) into the same graph model to form the first communication subgraph, used for transmitting uncomputed observation data, such as... Figure 4 As shown.

[0071] Step 6: Connect the observation subgraph to the first communication subgraph.

[0072] 6.1) Select node s1 corresponding to the observed satellite node s in the observation subgraph in the first communication subgraph;

[0073] 6.2) Add an observation satellite link (s,s1) between the two nodes s and s1 to achieve connectivity between the observation subgraph and the first communication subgraph, such as... Figure 5 As shown;

[0074] The link delay is l(s,s1)=0, which is used to characterize the waiting time required for the observation satellite u to be able to observe the observation target from the start of the observation mission.

[0075] Step 7: Combine all nodes in node set V except for the observation target node set S and all edges in edge set E2 to form the second communication subgraph.

[0076] 7.1) Select all nodes {s,a,b,d} in the node set V except for the observation target node set S, and rename the selected nodes to {s2,a2,b2,d2}.

[0077] 7.2) Select all edges {(s,a),(s,b),(a,b),(a,d),(b,d)} in edge set E2, and rename the selected edges to {(s2,a2),(s2,b2),(a2,b2),(a2,d2),(b2,d2)}, where the delay of each link is the communication delay required to transmit the calculated observation data, i.e., l(s2,a2)=l″(s,a)=1, l(s2,d2)=l″(s,b)=2, l(a2,b2)=l″(a,b)=1, l(a2,d2)=l″(a,d)=2, l(b2,d2)=l″(b,d)=1;

[0078] 7.3) Combine all the nodes selected and renamed in step 7.1) and all the edges obtained in step 7.2) into the same graph model to form a second communication subgraph, used for transmitting the calculated observation data, such as... Figure 6 As shown.

[0079] Step 8: Connect the first communication subgraph and the second communication subgraph to form a graph model that can jointly schedule the observation, transmission and computing resources of the satellite network.

[0080] 8.1) In the first communication subgraph, find the node {a1,b1} that corresponds to all computational satellite nodes {a,b} in the satellite node set V2;

[0081] 8.2) In the second communication subgraph, find the node {a2,b2} that corresponds to all computational satellite nodes {a,b} in the satellite node set V2;

[0082] 8.3) Add a computation link (a1, a2) between the corresponding nodes a1 and a2, with a link delay of l(a1, a2) = l. a =3, used to characterize the computation delay l of the observed data at satellite node a. a ;

[0083] 8.4) Add a computation link (b1, b2) between the corresponding nodes b1 and b2, with a link delay of l(b1, b2) = l. b =2, used to characterize the computation delay l of the observed data at satellite node b. b ;

[0084] By adding computational links (a1,a2) and (b1,b2), the first and second communication subgraphs are connected, forming a graph model capable of jointly scheduling observation, transmission, and computational resources of the satellite network, such as... Figure 7 As shown.

[0085] This model diagram can be used to transform the joint scheduling problem of observation, transmission, and computation into... Figure 7 The path finding problem between target nodes u and v and destination node d in the observation system significantly reduces the solution complexity of the network resource joint scheduling problem and the completion delay of the Earth observation mission.

[0086] Example 2 supports the application of representation graphs that support joint scheduling of observation transmission and computing resources.

[0087] In Earth observation missions, it is typically necessary to schedule observation, transmission, and computing resources simultaneously. However, existing Earth observation methods often employ independent scheduling for observation and transmission, and independent scheduling for transmission and computing, resulting in excessively long latency in Earth observation applications and failing to meet the low-latency requirements of missions such as emergency observation. This invention addresses these problems by employing a graph model. Specifically, it achieves joint scheduling of observation, transmission, and computing resources through an observation subgraph, a first communication subgraph, and a second communication subgraph within this representation graph. The specific implementation is as follows:

[0088] S1) The joint scheduling problem of observation, transmission and computation is transformed into a routing problem from the observation target node in the observation subgraph to the destination node in the second communication subgraph;

[0089] For example, the joint scheduling problem of observation, transmission, and computation of a satellite network can be transformed using the graphical model constructed at the end of Example 1. Figure 7 The problem of finding the path between target nodes u and v and destination node d in the observation.

[0090] S2) Calculate the optimal path from each observed target node in the observation subgraph through the observation subgraph, the first communication subgraph, and the second communication subgraph, and finally to the destination node in the second communication subgraph using the existing routing algorithm;

[0091] For example, using the shortest path first algorithm, the node with the lowest cost is selected from the candidate node list as the current node each time. All neighbors of the current node are traversed, the cost and next hop information of the neighbor nodes are updated, and the neighbor nodes are added to the candidate list. This process is repeated until the candidate node list is empty, thus obtaining the optimal path between the observation target node and the destination node of the Earth observation task. This supports the low latency requirements in the Earth observation scenario and reduces the completion latency of the Earth observation task and the complexity of solving the joint scheduling problem of network resources.

[0092] The above descriptions are merely a few specific examples of the present invention and do not constitute any limitation on the present invention. Obviously, those skilled in the art, after understanding the content and principles of the present invention, may make various modifications and changes in form and details without departing from the principles and structure of the present invention. However, these modifications and changes based on the ideas of the present invention are still within the scope of protection of the claims of the present invention.

[0093] It should be noted that the step numbers in the specification and claims of this invention are only for the purpose of clearly describing the embodiments of this invention and facilitating understanding, and their order is not limited.

Claims

1. A graph representation method supporting joint scheduling of observation transmission and computing resources, characterized in that, include: (1) A snapshot of the satellite network within a given time period. ,in It is a set of nodes. It is an edge set; (2) Set the nodes Different types of nodes are divided into different node sets, namely the observation target node set. Observation satellite node set Calculate the set of satellite nodes Ground station node set ; (3) Set the edge set The edges connecting different types of nodes are divided into different edge sets, namely the set of edges connecting the observation target node and the observation satellite node. The set of connecting edges between any satellite node and any satellite or ground station node. edge set The edges in the equation have a time delay property. The edges in the equation have two time delay properties; (4) Set the observation target nodes and the set of observation satellite nodes All nodes and edge sets All edges in the graph are combined to form an observation subgraph, which is used to characterize the observation process of the observation satellite on the observation target node; (5) Set the nodes In addition to the set of observed target nodes All external nodes and edge sets All edges are combined to form the first communication subgraph, which is used to transmit uncomputed observation data; (6) Select the corresponding observation satellite node between the observation sub-map and the first communication sub-map, and connect the observation sub-map and the first communication sub-map through the observation satellite node; (7) Set the nodes In addition to the set of observed target nodes All external nodes and edge sets All edges are combined to form a second communication subgraph, which is used to transmit the calculated observation data; (8) Select the corresponding computing satellite node between the first communication subgraph and the second communication subgraph, and then connect the two communication subgraphs together through the computing satellite node to form a graph model that can jointly schedule the observation, transmission and computing resources of the satellite network.

2. The method according to claim 1, characterized in that, In step (2), the node set The different types of nodes in the data are divided into different node sets, which is to divide the node sets... The multiple observation target nodes, multiple satellite nodes, and multiple ground station nodes included are divided into different node sets according to their different functions: The observation target nodes are composed of the observation target nodes in the observation task. , A set of observation satellite nodes is composed of satellite nodes with observation capabilities. , A set of computing satellite nodes is composed of satellite nodes with computing capabilities. , The ground station node set is composed of destination nodes located at the ground station. .

3. The method according to claim 1, characterized in that, Step (3) set the edge set The edges connecting different types of nodes in the graph are divided into different edge sets, which is to divide the edge sets into different edge sets. The edges in the data are divided into observation edges during the observation process and edges that transmit observation data after observation, i.e., the set of connecting edges between the observation target node and the observation satellite node. The set of connecting edges between any satellite node and any satellite or ground station node. .

4. The method according to claim 1, characterized in that, The set of connecting edges between the observation target node and the observation satellite node mentioned in step (3) An edge in a time delay property refers to the time delay of any observed target node. With any observation satellite node The edge between Its time delay attribute , used to characterize observation delay.

5. The method according to claim 1, characterized in that, The set of connecting edges between any satellite node and any satellite or ground station node mentioned in step (3) The edges in the equation have two time delay properties, which are used to characterize the time delay in time delay. and Communication delay required to transmit uncalculated observation data between them and in and The communication delay required to transmit calculated observation data between them. ,in It is any satellite node With any satellite or ground station node The edges between them.

6. The method according to claim 1, characterized in that, In step (5), the node set In addition to the set of observed target nodes All external nodes and edge sets All edges are combined to form the first communication subgraph, the implementation of which includes: (5a) Selecting a set of nodes In addition to the set of observed target nodes Exclude all nodes, and select any node from the selected nodes. Rename ; (5b) Selecting the edge set All edges in the selected edges, and any edge from the selected edges. Rename and set Link latency Communication latency required to transmit uncomputed observation data ,Right now ; (5c) All nodes selected and renamed in step (5a) and all edges obtained in step (5b) are placed in the same graph model to form the first communication subgraph.

7. The method according to claim 1, characterized in that, Step (6) involves selecting the corresponding observation satellite node between the observation subgraph and the first communication subgraph, and connecting the observation subgraph and the first communication subgraph through this observation satellite node. This process includes: (6a) For any observed satellite node in the observation subgraph Find its corresponding node in the first communication subgraph ; (6b) At two nodes and Add an observation satellite link between them This establishes connectivity between the observation subgraph and the first communication subgraph, with a link latency of [value missing]. Used to characterize observation satellites The waiting time required from the start of the observation mission to the point where the observation target can be observed.

8. The method according to claim 1, characterized in that, In step (7), the node set In addition to the set of observed target nodes All external nodes and edge sets All edges are combined to form a second communication subgraph, the implementation of which includes: (7a) Selecting a set of nodes In addition to the set of observed target nodes Exclude all nodes, and select any node from the selected nodes. Rename ; (7b) Selecting the edge set All edges in the selected edges, and any edge from the selected edges. Rename and set Link latency Communication delay required to transmit calculated observation data ,Right now ; (7c) Place all the nodes selected and renamed in step (7a) and all the edges obtained in step (7b) into the same graph model to form the second communication subgraph.

9. The method according to claim 1, characterized in that, Step (8) involves selecting the corresponding computational satellite node between the first and second communication subgraphs, and then connecting these two communication subgraphs together through this computational satellite node. This process includes: (8a) For calculating the set of satellite nodes Any computational satellite node in Find its corresponding node in the first communication subgraph Find the corresponding node in the second communication subgraph. ; (8b) At two nodes and Add a computing link between them This establishes connectivity between the first and second communication subgraphs, with a link latency of [value missing]. Used to characterize observation data in the computation of satellite nodes computation delay at the location .

10. An application of the graph representation method according to claim 1, characterized in that, include: When scheduling observation, transmission and computation resources simultaneously, the joint scheduling problem of observation, transmission and computation is first transformed into a routing problem from the observation target node in the observation subgraph to the destination node in the second communication subgraph through a representation graph. Then, the end-to-end optimal path for the Earth observation mission is obtained through existing routing algorithms to support the low latency requirements in the Earth observation scenario, reduce the completion latency of the Earth observation mission, and reduce the complexity of solving the joint scheduling problem of network resources.

Images