Joint Planning Method for Satellite Network Topology and Routing Based on Time-Expanded Graph

Abstract

The present invention discloses a method for jointly planning satellite network topology and routing based on a time-expanded graph, which mainly solves the problems of high transmission delay of the planned path and low utilization rate of link resources in the prior art. The solution is as follows: construct a time-expanded graph according to the topology change time and link information of the satellite network; calculate its routing in the time-expanded graph according to the service size, plan the shortest transmission path from the source node to the destination node, and reverse-search for the longest sub-path in it. For the services that have found the conditions satisfied, according to the service size that needs to plan the transmission path and the bandwidth and link switching time of each link above; perform topology planning by reducing the single transmission size of the service on each link; replace the longest sub-path in the original shortest transmission path with the planned path. The planned path of the present invention has low transmission delay and high utilization rate of link resources, and can be used in satellite networks with limited resources.

Description

Technical Field

[0001] The present invention belongs to the field of communication technologies, and particularly relates to a method for jointly planning resource topology and routing, which can be used for satellite networks with limited resources. Background Art

[0002] Due to the low orbital altitude, the satellite-ground link transmission is less interfered, and the hardware configuration cost is relatively low. Therefore, low-earth orbit (LEO) satellite networks are favored by commercial spaceflight. At the same time, due to the relatively low orbit, the LEO satellites move at a relatively high speed, resulting in a relatively fast change in the topology of LEO satellite networks. In addition, for some low-cost satellites, there are usually limitations in power or the number of transceivers. For LEO satellites with a single transceiver limitation and adjustable beam directions, they can establish communication links with other satellites that meet the link establishment conditions by adjusting the beam directions. However, at most only one communication link with other satellites can be established at the same time. All these make the topology and resources of LEO satellite networks have strong time-varying characteristics. In such a situation, the traditional topology and routing planning algorithms and corresponding routing protocols applied to terrestrial networks cannot provide stable, efficient, and reliable end-to-end transmission services for space services.

[0003] In satellite networks with limited resources, topology planning is usually carried out according to the movement rules of satellite nodes and different service metrics. Most of the existing topology planning schemes take the maximum network capacity or the fairness of link establishment and transmission among nodes in the network as the planning goal. Fraire J et al. proposed in the literature On the Design and Analysis of Fair Contact Plans in Predictable Delay-Tolerant Networks that due to factors such as satellite power budget limitations, hardware design, and signal interference, the situation of only carrying one transceiver is considered, and the fairness of node transmission opportunities and maximizing network capacity are taken as the goals, transforming the topology planning problem into a matching problem for solution. This method uses the connection time of links in the network as the judgment criteria for fairness and network capacity, without considering the routing problem. Moreover, the topology planning scheme only considers parameters such as the overall capacity or fairness of the network. Therefore, in the transmission scenario of a single service, especially an emergency service, the routing planning results obtained are often low link utilization rate and large end-to-end transmission delay. Summary of the Invention

[0004] The purpose of the present invention is to propose a method for jointly planning satellite network routing and topology based on a time-expanded graph to efficiently and quickly plan low-delay transmission paths for services and improve the utilization rate of link resources in view of the above deficiencies of the prior art.

[0005] The technical idea of the present invention is as follows: By utilizing the known parameters such as the orbits and speeds of satellites in the satellite network, accurate prediction of network information is carried out, thereby improving the routing efficiency; according to the predictability of satellite network parameters, the time-varying satellite network is modeled into a time-expanded graph, and by converting the joint planning problem of routing and topology in the network into a routing search and link planning problem in the time-varying graph, a low-complexity approximate solution method for the joint planning NP problem of routing and topology is given to improve the utilization rate of link resources. The implementation scheme includes the following:

[0006] (1) Construct a time-expanded graph according to the topological change moments and link information of the satellite network

[0007] (1a) Divide a number of connection time slots τ according to the connectivity status of the network n , to obtain a set T of N time slots = {τ1,..., τ n ,..., τ N}, and keep the connectivity relationship of the satellite network unchanged within one time slot, where τ n = [t n-1 , t n ) represents the size of the nth time slot,

[0008] (1b) Let V = {V τ1 ,..., V τn ,..., V τN} represent the set of nodes in each time slot of the satellite network, represents the potential connected edges in each time slot network, represents the link bandwidth between node i and node j in τ n , and δ represents the link switching time, where V τn represents the set of nodes in time slot τ n ;

[0009] (1c) According to the results of (1a) and (1b), construct a time-expanded graph:

[0010] (2) According to the service size Size and service start time t s of the transmission path to be planned, calculate the route for this service in the time-expanded graph and plan the shortest transmission path from the source node S to the destination node D

[0011] (2a) Initialize the set R of reached nodes = {S}, find the set N p = {S} of the currently neighboring nodes, let C be the set of all nodes, and the set of neighbor nodes of the nodes in N p ​Set the current time slot τ as the service start time t s The time slot where it is located;

[0012] (2b) Starting from time slot τ, traverse the set of nodes that have not arrived (C-R) and N p , according to the link bandwidth Service size Size, N p The service arrival time of the nodes in, find the previous node with the minimum service arrival time for the nodes in (C-R) in N p , and when finding the previous node, ensure that the service arrival time of the nodes in (C-R) does not exceed this time slot, then add these previous nodes and the corresponding next hops to N pn , and let R = R ∪ N pn , N p = N pn ,

[0013] (2c) Determine whether the destination node D is in the node set R:

[0014] If D ∈ R, then backtrack the path from the destination node D Execute step (3);

[0015] If D is not in R, and then return to (2b);

[0016] If D is not in R, and then let N p = R, return to (2b), and execute from the start time of the next time slot;

[0017] (3) Starting from the destination node D, find the longest sub-path in the reverse direction in the shortest transmission path and satisfy that Any link in can be connected at any moment within the connection time of , and the number of nodes on is greater than 3;

[0018] (4) For the sub-path that has been found to meet the conditions , according to the service size Size required to plan the transmission path, this sub-path the bandwidth of each link on link switching time δ, perform topology planning by reducing the single transmission size of the service on each link:

[0019] (4a) Obtain the number of nodes (k + 1) on the sub-path , according to the bandwidth of each link on the sub-path find this sub-path on The minimum link bandwidth B of all links above min ;

[0020] (4b) Transmit the service with size Size in m times, starting from m = 1 and increasing sequentially, and find the m that minimizes the total service transmission time T t =(2m + k - 2)T+(2m + k - 3)δ, where T = Size / (m*B min ) represents the longest time required for the link on the path to transmit the service with size (Size / m);

[0021] (4c) According to the found transmission times m, judge whether the planning is successful:

[0022] If m = 1, the planning fails,

[0023] If m > 1, the planning is successful, and the slot size T = Size / (m*B min ) is obtained;

[0024] (4d) Calculate the minimum link bandwidth B' of the links in this sub-path on each link according to the bandwidth in the time (t s , t s +T t ), and judge its size compared with the minimum link bandwidth of the sub-path : min If B' min

[0025] < B min , then let B' min = B min , and return to (4b); min Otherwise, the planning ends, and the planned path

[0026] is output

[0027] (5) Replace the longest sub-path in the original shortest transmission path with the planned path as the final service transmission path and topology planning scheme .

[0028] Compared with the prior art, the present invention has the following advantages:

[0029] First, because the present invention utilizes the predictability of the satellite network and constructs a time-expanded graph according to the satellite network topology change moment When performing routing calculation and topology planning, the single transceiver limitation, link switching time, and different link rates are fully considered to perform reliable routing and topology planning for service transmission, which can be applied to routing calculation and topology planning in the scenario of single transceiver limited satellite networks.

[0030] Second, since the present invention utilizes a time-expanded graph On the basis of ensuring complete service transmission, the shortest end-to-end transmission path of hop-by-hop transmission is first planned, and then topology planning is performed on sub-paths that meet the set conditions among them. By simultaneously using the links on the path for service transmission, the end-to-end delay of service transmission is further reduced.

[0031] Third, since the present invention simultaneously uses the links on the path for service transmission when performing topology planning on the service transmission path, the utilization rate of link resources and the average throughput of the transmission path are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 is the network scenario diagram of the present invention;

[0033] Figure 2 is the implementation flowchart of the present invention;

[0034] Figure 3 is the time-expanded graph constructed in the present invention DETAILED DESCRIPTION OF THE EMBODIMENTS

[0035] The following further describes the embodiments and effects of the present invention in detail with reference to the accompanying drawings and specific embodiments:

[0036] Referring to Figure 1 , the usage scenario of this example includes seven nodes A, B, C, D, E, F, and G. The nodes A, B, C, D, E, and F are six low-earth orbit satellites with single transceiver limitations on the same polar orbit, that is, one satellite can establish a communication link with at most one node at the same time; the G node is a ground station. Figure 1 identifies the potential connection links between each node, that is, the links that meet the link establishment conditions between nodes. Among them, Figure 1 (a) is a schematic diagram of the potential connection relationship between each node in the time period from t0 to t1. There are potential connection links between node A and node B, node B and node C, node D and node E, node E and node F, and node F and node G in the figure. At time t0, satellite A completes data collection over the observation target. Due to satellite movement, the potential connection relationship between node B and node C, node C and node D, node E and node G, and node F and node G changes at time t1. The schematic diagram of the potential connection relationship between each node after the change is as shown in Figure 1As shown in (b), there are potential communication links between node A and node B, node C and node D, node D and node E, node E and node F, and node E and node G, and Figure 1 The potential communication relationships between the nodes in (b) remain unchanged from t1 to t2.

[0037] Figure 1 The topology change times in are t0, t1, and t2 respectively. In this example, it is assumed but not limited to t0 = 0, t1 = 25s, and t2 = 70s. After the satellite node A in the scenario collects the data of the observation target, it needs to start transmitting the observation data back to the ground station G at t = 1s. It is assumed but not limited to that the size of the observation data service to be sent by the satellite node A is 100Mb, the link bandwidth in the scenario is 10Mbps, and the link switching delay δ = 1s.

[0038] Referring to Figure 2 , the implementation steps for the joint planning of the satellite network topology and routing based on the above scenario are as follows:

[0039] Step 1: Construct a time-expanded graph according to the topology change moments and link information of the satellite network.

[0040] Referring to Figure 3 , the specific implementation of this step is as follows:

[0041] 1.1) According to the topology change moments t0, t1, and t2 of the network, divide the network into two time slots T = {τ1, τ2}, where τ1 = [t0, t1) and τ2 = [t1, t2);

[0042] 1.2) Let V = {V τ1 ,..., V τn ,..., V τN} represent the set of nodes in each time slot of the satellite network, represent the potential connected edges in each time slot network, represent the link bandwidth between node i and node in τ n , and δ represents the link switching time, where V τn represents the set of nodes in time slot τ n ;

[0043] 1.3) According to the results of 1.1) and 1.2), construct a time-expanded graph: As Figure 3 shown, Figure 3 identifies the time slots τ1, τ2, all the nodes in V, the edges in E including potential connected edges and storage edges, and the bandwidth of each potential connected edge A1, B1, C1, D1, E1, F1, G1 are the identifiers of nodes A, B, C, D, E, F, G in time slot τ1 respectively, and A2, B2, C2, D2, E2, F2, G2 are the identifiers of nodes A, B, C, D, E, F, G in time slot τ2 respectively. t0, t1, t2 are the topological change times of this instance. The link bandwidth of the potential connection link between different nodes in the same time slot is marked on the potential connection edge. The same node in adjacent time slots, such as A1 and A2, is connected by a storage edge, indicating the storage capacity of the node.

[0044] In this instance:

[0045] Let V = {V τ1 , V τ2}, where V τ1 = {A1, B1, C1, D1, E1, F1, G1}, V τ2 = {A2, B2, C2, D2, E2, F2, G2}.

[0046] V τ1 and V τ2 represent the sets of nodes in time slots τ1 and τ2 respectively. A1, B1, C1, D1, E1, F1, G1 are the identifiers of nodes A, B, C, D, E, F, G in time slot τ1 respectively, and A2, B2, C2, D2, E2, F2, G2 are the identifiers of nodes A, B, C, D, E, F, G in time slot τ2 respectively;

[0047] As Figure 3 shown, after dividing the network scenario into 2 time slots, in time slot τ1, it includes edges Let the link bandwidths of these edges be In time slot τ2, it includes edges Let the link bandwidths of these edges be The same node in two time slots is connected by a storage edge The link bandwidth of the potential connection link between nodes i and j in time slot τ is not considered for the storage edge, where n indicates that there is a potential connection relationship between node i and node j in time slot τ represents the storage edge of node i, indicating that the node has the ability to cache services, represents the link bandwidth of the potential connection link between node i and node j in time slot τ n , i ≠ j.

[0048] Let δ = 1, indicating that the link switching delay in this instance is 1 s.

[0049] Step 2: Calculate the service route in the time-expanded graph and plan the shortest transmission path from the source node A to the destination node G

[0050] 2.1) Initialize the set of reached nodes R = {A}, the set of nodes N p = {A} that need to find neighbors, and the set of all nodes C = {A, B, C, D, E, F, G}. Let N p be the set of neighbor nodes of the nodes in the current time slot τ = τ1;

[0051] 2.2) Starting from the time slot τ in 2.1), traverse the sets (C - R) and N p = {A}, and find the predecessor node in N p that minimizes the service arrival time for the nodes in (C - R), where (C - R) represents the set of unreached nodes;

[0052] 2.2.1) In the time slot τ = τ1, find the nodes in the set of unreached nodes (C - R) = {B, C, D, E, F, G} that have potential connected edges with the nodes in N p = {A}. The node is B. Add node B to N pn and N pn = {B};

[0053] 2.2.2) For node B in N pn , the predecessor node in N p that can make the service reach B earliest is A. Update the service arrival time of node B to not exceeding this time slot.

[0054] 2.2.3) Update the set of reached nodes: R = R ∪ N pn = {A, B}; Update the set of nodes that need to find neighbors: N p = N pn = {B}; Clear the set of neighbor nodes of the nodes in N p :

[0055] 2.3) Determine whether the destination node G is in the set of reached nodes R, and judge the next operation to be performed according to whether the set N p is empty:

[0056] If G is not in R, and then return to 2.2) to find the predecessor node in N p that minimizes the service arrival time for the nodes in (C - R);

[0057] If G is not in R, and then let N p = R, and return 2.2) to find the predecessor node in N p for the node in (C - R) that minimizes the service arrival time;

[0058] If G is in R, then backtrack the shortest path from the source node A to the destination node G.

[0059] In this example, when returning to 2.2) for iterative search in (C - R), the process of finding the predecessor node in N p that minimizes the service arrival time for the node is as follows:

[0060] First iteration: R = {A, B}, the destination node G is not in the set R, and return 2.2) to continue finding the predecessor node in N p for the node in (C - R) that minimizes the service arrival time;

[0061] In time slot τ = τ1, for the node C in (C - R) = {C, D, E, F, G}, find the node in the set N p = {B} that enables the earliest arrival of the service for node C, which is node B.

[0062] Update the service arrival time of node C:

[0063] Update the reached node set R = {A, B, C};

[0064] Update the set of nodes for which neighbors need to be found: N p = {C};

[0065] Empty the set of neighbor nodes of the nodes in N p :

[0066] Second iteration: R = {A, B, C}, the destination node G is not in the set R, and return 2.2) to continue finding the predecessor node in N p for the node in (C - R) that minimizes the service arrival time:

[0067] In time slot τ = 1, for the nodes in (C - R) = {D, E, F, G}, no node can be found in the set N p = {C} that enables the earliest arrival of the service.

[0068] Update the reached node set R = {A, B, C};

[0069] Update the set of nodes for which neighbors need to be found:

[0070] Empty N p Set of neighbor nodes of the middle node:

[0071] Third iteration: R = {A, B, C}, the destination node G is not in the set R, and Let N p = R, τ = τ2, return to 2.2) and continue to find the predecessor node in N for the nodes in (C - R) that minimizes the service arrival time: p

[0072] In time slot τ = τ2, for node D in (C - R) = {D, E, F, G}, find the node in the set N p = {A, B, C} that enables the service of node D to arrive earliest, which is node C,

[0073] Update the service arrival time of node D:

[0074] Update the reached nodes: R = {A, B, C, D}

[0075] Update the set of nodes for which neighbors need to be found: N p = {D};

[0076] Empty N p Set of neighbor nodes of the nodes in

[0077] Fourth iteration: R = {A, B, C, D}, the destination node G is not in the set R, and Return to 2.2) and continue to find the predecessor node in N for the nodes in (C - R) that minimizes the service arrival time: p

[0078] In time slot τ = 2, for node E in (C - R) = {E, F, G}, find the node in the set N p = {D} that enables the service of node E to arrive earliest, which is node D,

[0079] Update the service arrival time of node E:

[0080] Update the reached nodes R = {A, B, C, D, E};

[0081] Update the set of nodes for which neighbors need to be found: N p = {E};

[0082] Empty N p Set of neighbor nodes of the nodes in

[0083] ​​The fifth iteration: R = {A, B, C, D, E}, the destination node G is not in the set R, and Return 2.2) Continue to find the predecessor node in N for the nodes in (C - R) p such that the service arrival time is minimized:

[0084] In time slot τ = τ2, for node G in (C - R) = {F, G}, find in the set N p = {E} the node that enables the earliest arrival of the service for node G, which is node E;

[0085] Update the service arrival time of node G:

[0086] Update the set of nodes that have arrived: R = {A, B, C, D, E, G};

[0087] Update the set of nodes for which neighbors need to be found: N p = {G};

[0088] Empty the set of neighbor nodes of the nodes in N p :

[0089] The sixth iteration: R = {A, B, C, D, E, G}, the destination node G is in the set R, and the shortest path A → G can be obtained by backtracking the path

[0090] Step 3: Reverse-search for the longest sub-path in the shortest transmission path :

[0091] 3.1) Determine whether the number of nodes on the shortest transmission path is greater than 3:

[0092] If the number of nodes on it is greater than 3, initialize the current link as the last link on it, and execute 3.2);

[0093] Otherwise, not found

[0094] In this example, the number of nodes on the path is greater than 3, and the current link is initialized as {E2, G2};

[0095] 3.2) Add the current link to the sub-path and, based on E in the time-expanded graph G and the start and end times of the sub-path judge whether each link on this sub-path satisfies the condition within There are potential connected edges in each time slot from the start time to the end time:

[0096] If satisfied, update the current link to the previous link of the current link, and repeat 3.2);

[0097] If not satisfied, then from this sub-path delete the last added link and execute 3.3);

[0098] In this example, the links {E2, G2}, {D2, E2}, and {C2, D2} are added in sequence to i.e., At this time the start time of is t' s = t1 = 25s, and the end time is are all within the time slot τ2, and at this time the three links in are all connected within τ2.

[0099] Update the current path to {B1, C1}, and add {B1, C1} to According to the time-expanded graph, it can be seen that the link {B1, C1} is only connected within the time slot τ1, and the link {B1, C1} does not meet the conditions. Delete the link {B1, C1} from At this time

[0100] 3.3) Judge the shortest transmission path of the sub-path whether the number of nodes on is greater than 3:

[0101] If the number of nodes on is greater than 3, then this sub-path is the longest sub-path found and output;

[0102] Otherwise, the longest sub-path is not found.

[0103] In this example, the path has 4 nodes, which is greater than 3, and output

[0104] Step 4: Perform topological planning on the path .

[0105] 4.1) Obtain the number of nodes on the path and the minimum link bandwidth of all links on the path :

[0106] In this example, the number of nodes on the path is 4, i.e., (k + 1) = 4, k = 3;

[0107] The planned service is on the sub-path The start transmission time t' s = t1 = 25s, and the service is transmitted to The arrival time at the last node on t' s ~t' a The time slot where it is located is τ2. For all links on the sub-path The minimum link bandwidth B min = 10Mbps;

[0108] 4.2) Transmit the service with size Size in m times, starting from m = 1 and increasing sequentially, and find the m that makes the total service transmission time T t = (2m + k - 2)T + (2m + k - 3)δ the smallest, where T = Size / (m * B min ):

[0109] It can be seen that T t = (2m + k - 2)T + (2m + k - 3)δ is a quadratic function of m, and the maximum point is the maximum value point.

[0110] In this example, when m = 1, T t = 32s, when m = 2, T t = 29s, when m = 3, T t = 29.33s. So when m = 2, the time required for the service with size Size to complete transmission on the path is the least;

[0111] 4.3) According to the found number of transmissions m, judge whether the plan is successful:

[0112] If m = 1, the plan fails,

[0113] If m > 1, the plan is successful, and the time slot size T = Size / (m * B min );

[0114] In this example, m = 2 > 1, the plan is successful, and the time slot size T = Size / (m * B min ) = 100 / (2 * 10) = 5s;

[0115] 4.4) According to the minimum link bandwidth of the link in the time (t' s , t' s + T t ) to judge whether to end the plan:

[0116] If B' min < B min , then let B' min = Bmin , return (4b);

[0117] Otherwise, the planning ends, and the planned path is output

[0118] In this example, t' s = t1 = 25s, t' a = t' s + T t = 54s, t' s ~ t' a The time slot where it is located is τ2, The minimum link bandwidth B' of the link in the time slot τ2 min = 10Mbps = B min , the planning ends.

[0119] The planned path is: The link {C2, D2} is connected from 25 to 30s, the link {D2, E2} is connected from 31 to 36s, the link {C2, D2} and the link {E2, G2} are connected from 37 to 42s, the link {D2, E2} is connected from 43 to 48s, and the link {E2, G2} is connected from 49 to 54s.

[0120] Step 5: Use the planned to replace the original in as the final service transmission path and topology planning scheme.

[0121] The above description is only a specific example of the present invention. Obviously, for professionals in the field, after understanding the content and principle of the present invention, various corrections and changes in form and details may be made without departing from the principle and structure of the present invention, such as graph model transformation, selection of routing methods, etc. However, these corrections and changes based on the idea of the present invention are still within the protection scope of the claims of the present invention.

Claims

1. A method for jointly planning satellite network topology and routing based on time-expanded graph, characterized in that It includes the following steps: (1)Construct a time-expanded graph based on the topological change moments and link information of the satellite network (1a) Divide a number of connection time slots τ according to the connection state of the network n , and obtain a set T of N time slots = {τ1,..., τ n ,..., τ N}, and keep the connection relationship of the satellite network unchanged within one time slot, where τ n = [t n - 1, t n ) represents the size of the nth time slot. (1b) Let \(V = \{V τ1 ,\ldots,V τn ,\ldots,V τN \}\) represent the set of each time-slot node in the satellite network, represent the potential connected edges in each time-slot network, represent the link bandwidth between node \(i\) and node \(j\) at \(\tau n , \delta\) represents the link switching time, where \(V τn represents the set of nodes in time-slot \(\tau n \); (1c) Construct a time-expanded graph based on the results of (1a) and (1b): (2) Plan the service size Size and service start time t of the transmission path as needed s , and calculate the route for this service in the time expansion diagram , and plan the shortest transmission path from the source node S to the destination node D (2a) Initialize the set R of nodes that have been reached as R = {S}, and the set N of nodes that need to find neighbors p = {S}. Let C be the set of all nodes, and N p be the set of neighbor nodes of the nodes in Let the current time slot τ be the time slot t when the service starts s where it is located; (2b) Traverse the set of nodes that have not arrived (C-R) and N starting from time slot τ p , according to the link bandwidth service size Size, N p the service arrival time of the nodes in, find the predecessor node in N that minimizes the service arrival time for the nodes in (C-R), and when finding the predecessor node, ensure that the service arrival time of the nodes in (C-R) does not exceed this time slot, then add these predecessor nodes and the corresponding next hops into N p and let R = R ∪ N pn , and let N pn = N p = N pn , (2c) Determine whether the destination node D is in the node set R: If D ∈ R, then backtrack the path from the destination node D Execute step (3); If and then return (2b); If and then set N p = R, return to (2b), and execute starting from the start time of the next time slot; (3) Starting from the destination node D, find the longest sub-path in reverse in the shortest transmission path and satisfy that any one of the links in can be connected at any moment during the connection time, and the number of nodes on is greater than 3; (4) For the sub-path that has found to meet the conditions According to the service size Size of the transmission path to be planned as needed, and this sub-path The bandwidth of each link on it Link switching time δ, perform topology planning by reducing the single transmission size of the service on each link: (4a) Obtain the sub-path The number of nodes (k + 1) on it, according to the sub-path The bandwidth of each link on it Search for this sub-path The minimum link bandwidth B of all links on it min ; (4b) Transmit the service of size Size in m times, starting from m = 1 and increasing sequentially, and find the m that minimizes the total transmission time T of the service t =(2m + k - 2)T+(2m + k - 3)δ, where T = Size / (m*B min ) represents the longest time required for the link on the path to transmit a service of size (Size / m); (4c) According to the found number of transmissions m, determine whether the planning is successful: If m = 1, the planning fails, If m > 1, the planning is successful, and the slot size T = Size / (m*B min ) (4d) According to the bandwidth of each link on the sub-path calculate the minimum link bandwidth B' of the links on this sub-path in the time period (t ', t s ', t s '+ T t ) and determine its magnitude compared to the minimum link bandwidth of the sub-path min : ​ If B' min <B min , then let B' min =B min , and return (4b); Otherwise, the planning ends and the planned path is output (5) Use the planned path to replace the original shortest transmission path with the longest sub-path as the final service transmission path and topology planning scheme.

2. The method according to claim 1, characterized in that, In step (1a), a number of connection time slots τ are divided according to the connection state of the network n , which is divided according to the network topology change time, that is, assuming that the linkable relationship between each node in the network changes at times t0, t1,..., t N , and based on this, the network topology is divided into time slots to obtain τ1,..., τ n ,..., τ N , a total of N time slots, where τ n = [t n-1 , t n ), represents the size of the nth time slot.

3. The method according to claim 1, wherein In step (2b), find the predecessor node in N for the node in (C-R) that minimizes the service arrival time, as follows: p ​ (2b1) Traverse N p and the nodes in (C-R), and according to the in (C-R) that are related to N p the nodes in which have potential connection edges with the nodes in N in the current time slot are saved in the neighbor set N p of N pn ; (2b2) Traverse N pn , for each N pn traverse the nodes A in p , according to N p calculate the time for each node in p to reach node A based on the service arrival time, service size and link bandwidth of the nodes in N p and select the node in that can make the service reach node A earliest, denoted as the predecessor node of A, and save the time when the service reaches node A.

4. The method according to claim 1, characterized in that In step (3), find the longest sub-path in reverse along the shortest transmission path as follows: The implementation is as follows: (3a) Determine whether the number of nodes on the shortest transmission path is greater than 3: If the number of nodes on is greater than 3, initialize the current link as the last link on Otherwise, not found (3b) Add the current link to the sub-path and, according to the time-expanded graph E in and the start and end times of the sub-path judge whether each link on has potential connected edges in each time slot from the start to the end time of If satisfied, update the current link to the previous link of the current link, and repeat (3b); If not satisfied, then from this sub-path Delete the last added link and execute (3c); (3c) Determine the sub-path whether the number of nodes on it is greater than 3: If the number of nodes on is greater than 3, then this sub-path is the longest sub-path found and output it; otherwise, the longest sub-path is not found.

5. The method according to claim 1, wherein In step (4a), according to the sub-path bandwidths of each link find the minimum link bandwidth B of all links on this sub-path as follows: min , the implementation is as follows: (4a1) According to the start transmission time t ' of the planned service on the sub-path s ' and the arrival time t' at the last node transmitted to a , calculate the time slot where t ' to t' s is located according to the time slot division T in the time-expanded graph a during this period; (4a2) Traverse the time slots and sub-paths calculated in (4a1) on the links, and according to the link bandwidths of the links on these time slots find the minimum link bandwidth among all the links in these time slots on and denote it as B min .

6. The method according to claim 1, wherein In step (4d), according to the sub-path and the bandwidth of each link calculate the minimum link bandwidth B' of the links in this sub-path s during the time period (t s ', t t '+T min ) as follows: (4d1) Let t' a = t s '+ T t , calculate t ' to t' s within the time slot where this period of time is located according to the time slot division T in the time expansion diagram a ; (4d2) Traverse the time slots and sub-paths calculated in (4d1) on the link, and according to the link bandwidth of the link on these time slots find the minimum link bandwidth among all links on these time slots and denote it as B'. min .

Images