A satellite network routing planning method based on bidirectional A-star algorithm

Abstract

The application discloses a satellite network routing planning method based on a bidirectional A-star algorithm, and comprises the following steps: establishing a network topology dynamic model considering regular motion of satellites, and determining model constraint conditions; establishing evaluation indexes based on an acceptable degree; establishing optimization targets according to the evaluation indexes; calculating a reference route by using a unidirectional A-star algorithm; and using a bidirectional A-star algorithm to obtain a route strategy with the highest overall acceptable degree as a current optimal route strategy. The satellite network routing planning strategy obtained by the application considers link time effectiveness and hop number influence, can adapt to dynamic changes of a future network topology, and can improve solving speed and efficiency of route calculation.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of satellite network routing planning control, and particularly relates to a satellite network routing planning method based on bidirectional A-star algorithm. BACKGROUND

[0002] With the development of mobile communication technology, satellite communication has become one of the research hotspots. Current ground mobile communication mainly relies on base stations for wireless communication, but its signal coverage range is limited and will be affected by various factors including terrain, weather, external environment. Satellite communication effectively overcomes the above defects and is widely used in disaster rescue, weather detection, terrain exploration and other tasks where ground communication network applicability is poor. Some existing researches apply software defined network (SDN) to satellite network and propose a network architecture based on SDN to achieve efficient control of satellites.

[0003] Existing researches on software defined satellite network are still in the initial stage, and how to fully utilize the characteristics of satellite movement and the advantages of software defined network is a key factor to solve the large-scale satellite network routing problem. The software defined satellite network routing planning method mainly faces the following difficulties. First, the routing performance requirements are diverse. In order to ensure network performance, routing strategy needs to meet the requirements of small routing overhead, high reliability, high packet transmission success rate and rationality. Second, the information is time-varying. The three steps of node discovering topology changes, controller responding changes and calculating new paths, and controller issuing new routing rules all need time, especially in space, the distance between nodes and controllers is very far, which will cause certain communication delay, resulting in that the information relied on by the controller is not timely enough to reflect the network situation, thus reducing the routing performance. Finally, the node mobility and regularity are strong. Satellites in satellite network move regularly according to the orbit, and the link fails frequently, making it difficult to efficiently solve the routing problem. SUMMARY

[0004] In view of the above difficulties in satellite network routing planning, the present application proposes a satellite network routing planning method based on bidirectional A-star algorithm, which overcomes the difficulties of complex performance requirements, poor information timeliness, node time-varying and strong regularity by acceptable degree and bidirectional A-star algorithm combined with the structural characteristics of the problem.

[0005] The technical scheme for achieving the object of the present application is as follows: a satellite network routing planning method based on bidirectional A-star algorithm, comprising the following steps:

[0006] Step 1, obtaining the total number of satellites, satellite positions, satellite orbits and satellite movement speeds;

[0007] Step 2, a network topology dynamic model considering the timeliness and hop number of links is established, and model constraint conditions are determined;

[0008] Step 3, based on the acceptable degree, a current feasible routing strategy set is evaluated, and evaluation indexes are established;

[0009] Step 4, according to the evaluation indexes, optimization targets are established;

[0010] Step 5, a one-way A-star algorithm evaluation function is designed, and a reference route is calculated;

[0011] Step 6, a two-way A-star algorithm evaluation function is designed, a routing strategy set with the highest overall acceptable degree is obtained according to the reference route, and the routing strategy set is taken as a current optimal routing strategy.

[0012] An electronic device comprises a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor implements the satellite network routing planning method based on the two-way A-star algorithm when executing the program.

[0013] A computer readable storage medium has a computer program stored thereon, and the program is executed by a processor to implement the satellite network routing planning method based on the two-way A-star algorithm.

[0014] The above technical scheme is adopted, and the following technical effects can be achieved:

[0015] The established satellite network topology model can effectively represent the future motion of the satellite, and simultaneously considers various routing performance requirements, so that the adaptability of the routing planning strategy to frequent changes of the satellite network link can be improved. Based on the above satellite network topology model, the evaluation indexes combined with the acceptable degree are adopted to describe the link cost changes caused by the satellite motion, and the timeliness of the information is improved. Meanwhile, the traditional A-star algorithm is improved in combination with the characteristics of the acceptable degree, the solving speed and efficiency are improved on the basis of ensuring that the routing acceptable degree is high. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 It is a principle block diagram of the satellite network routing planning method of the application.

[0017] Figure 2 It is a schematic diagram of a routing strategy after routing planning. DETAILED DESCRIPTION

[0018] The technical scheme of the application will be further described in combination with the drawings and embodiments.

[0019] As Figure 1As shown, the present application mainly comprises: an initialization module, a satellite network topology dynamic model construction module, a constraint condition construction module, an evaluation index construction module, an optimization objective function construction module, a reference route generation module, and an optimal routing strategy extraction output module. The specific implementation steps are described in combination with the flow chart as follows:

[0020] Step 1: initialization. The current total number of satellites V, satellite position p i , satellite orbit o i , and satellite movement speed v i are input by an operator.

[0021] Step 2: satellite network topology dynamic model construction. The satellite network model is defined as G(N,C,L,T), where N represents the set of satellite nodes, |N|=V, V is the total number of satellites; C=(c ij ) V×V is a matrix describing the link cost, and the element c ij in the matrix C is used to describe the link cost between satellite i and satellite j; L=(l ij ) V×V is a matrix describing the satellite connection, and l ij =1 indicates that there is a connection between satellite i and satellite j, and l ij =0 indicates that there is no connection between satellite i and satellite j; is a matrix describing the expected maintenance time of the communication link, which represents the expected maintenance time of the communication link between satellite i and satellite j. The link cost in the system can be represented as:

[0022]

[0023]

[0024]

[0025] where t1 represents the time when c ij is the smallest in the future movement of the satellite after the routing rule is successfully issued to the satellite, and t2 represents the time when c ij is the largest. represents the element in the matrix T with the largest reciprocal of the expected maintenance time of the communication link at time t, represents the element in the matrix with the smallest reciprocal of the expected maintenance time of the communication link at time t, hop(i,j)=1 indicates that the number of hops between satellite i and satellite j is 1 when they communicate directly, c ij represents the value corresponding to c ij at time t1, which is the lower limit, and represents the value corresponding to c ij at time t2, which is the upper limit.

[0026] Step 3: Constraint construction. Since the satellite routing planning needs to be satisfied while considering the existence of satellite network physical characteristics, the satellite network needs to satisfy the following constraint conditions when performing routing planning:

[0027]

[0028]

[0029]

[0030] wherein, formula (4) represents that only when the distance d ij between satellite i and satellite j is less than the maximum communication distance D ij between the two, a communication link exists, formula (5) represents the value of c ij under different conditions, and formula (6) represents the value of under different conditions.

[0031] Step 4: Evaluation index construction. The link cost is expressed in the form of average value and interval radius, as shown in the following equation:

[0032]

[0033]

[0034] For any two links (i, j) and (u, v), according to formula (7) and formula (8), A = [m ij , w ij ], B = [m uv , w uv ]. When m ij ≤ m uv , it is recorded as A < B, meaning that A has a smaller link cost than B. Define ζ(A < B) as the acceptable degree, as shown in the following equation:

[0035]

[0036] On this basis, the following acceptable degree classification is established:

[0037]

[0038] wherein, when ζ(A < B) = 0, the assumption that “A has a smaller link cost than B” is not accepted; when 0 < ζ(A < B) < 1, the decision maker accepts the assumption that “A has a smaller link cost than B”, with a satisfaction degree of 0 to 1; and when ζ(A < B) ≥ 1, the decision maker is very satisfied with the assumption that “A has a smaller link cost than B”.

[0039] For any two satellite nodes s and t, let P st denote all s-t routes, and for any s-t route, define as follows:

[0040] p = {e1, e2,..., en} (10) n ,...,e Q} (11)

[0041] where e n = (i, j) denotes a link between node i and node j, n = 1, 2,..., Q, and Q denotes the number of links in route strategy p.

[0042] The following route strategy cost equation is established:

[0043]

[0044]

[0045]

[0046] where cost(p) denotes the route strategy cost, m p is the average of the route strategy cost, and w p is the interval radius of the route strategy cost.

[0047] The following reference route selection equation is established:

[0048]

[0049] where p rf represents the reference route, and p any represents any route strategy in P st . c uv , is the link cost c uv of link (u, v) using the upper and lower bounds expressed by equation (3).

[0050] The route cost of p rf is recorded in [m rf , w rf ], where m rf and w rf are defined in equations (13) and (14), respectively, and the following route acceptability equation is established:

[0051]

[0052] where p denotes the route waiting for the acceptability evaluation.

[0053] Step 5: Optimization target establishment. The following optimization target is established:

[0054] Maxζ(p eval <p rf ) (17)

[0055] Step 6: Reference route generation. The following reference route cumulative cost equation is established:

[0056]

[0057] where e1,...,e n ∈p rf .

[0058] The following reference route target point estimated cost equation is established:

[0059]

[0060] where d jt represents the Euclidean distance between node j in the current link e n =(i,j) and the target node t, and represents the upward rounding, and D represents the maximum communication distance between satellite nodes.

[0061] The following reference route evaluation equation is established:

[0062] f rf (n) = g rf (n) + h rf (n) (20)

[0063] Step 7: Optimal routing strategy extraction output. The following routing cumulative cost equation is established:

[0064]

[0065] where

[0066] The following routing evaluation equation is established:

[0067]

[0068] where the expression of h o (n) is the same as equation (19).

[0069] For the reference route p rf , whose routing is expressed as equation (11), select as the dividing point, and divide the route p rf into p rff , p rfs , as shown in the following equation:

[0070]

[0071]

[0072] where p rff is the first half of the reference route, p rfs is the second half of the reference route, and p rf = p rff ∪ p rfs .

[0073] The following optimization objectives are established:

[0074] Maxζ(p bf <p rff ) (25)

[0075] Maxζ(p bs <p rfs ) (26)

[0076] where p bf is the optimal route obtained from the first half of the reference route with s as the starting point and j as the target point, and p bs is the optimal route obtained from the second half of the reference route with t as the starting point and j as the target point. According to p bf , p bs , the following optimal route is obtained:

[0077] p b = p bf ∪ p bs (27)

[0078] The experimental results of the present application are shown in Figure 2 . The hardware environment is: CPU i7-12700H@2.30GHz, memory 24G, and a total of 50 satellite nodes are routed. Figure 2 The satellite network routing planning results are shown. The green origin in the figure represents the starting point, and the pentagram represents the target point. It can be seen from Figure 2 that the optimal route obtained by planning intersects with the reference route at a certain node in the middle, which is because when calculating the optimal routing strategy, one point in the reference route is used as a temporary end point. It can be noted that the reference route and the optimal strategy do not completely coincide, and according to the evaluation index based on the acceptable degree, the optimal strategy has a smaller link cost than the reference route.

[0079] The satellite network topology model can effectively represent the future motion of satellites, and can improve the adaptability of the routing planning strategy to the frequent changes of satellite network links by considering various routing performance requirements. Based on the satellite network topology model, an evaluation index of acceptance degree is used to describe the link cost changes caused by satellite motion, and the timeliness of information is improved. Meanwhile, the traditional A-star algorithm is improved based on the characteristics of the acceptance degree, so as to improve the solving speed and efficiency on the basis of ensuring a high acceptance degree of the route.

[0080] The satellite network topology model can effectively represent the future motion of satellites, and can improve the adaptability of the routing planning strategy to the frequent changes of satellite network links by considering various routing performance requirements. Based on the satellite network topology model, an evaluation index of acceptance degree is used to describe the link cost changes caused by satellite motion, and the timeliness of information is improved. Meanwhile, the traditional A-star algorithm is improved based on the characteristics of the acceptance degree, so as to improve the solving speed and efficiency on the basis of ensuring a high acceptance degree of the route.

[0081] The application provides a satellite network routing planning method based on a bidirectional A-star algorithm, and there are many methods and approaches for specifically implementing the technical scheme, and the above description is only a preferred embodiment of the application. It should be pointed out that, for ordinary technical personnel in the technical field, some improvements and refinements can be made without departing from the principle of the application, and these improvements and refinements should also be regarded as the protection scope of the application. The components not explicitly described in the embodiment can be implemented by using the prior art.

Claims

1. A satellite network routing planning method based on bidirectional A* algorithm, characterized in that, Comprising the following steps: Step 1, obtaining the total number of satellites, satellite positions, satellite orbits, and satellite movement speeds; Step 2, establishing a network topology dynamic model considering link timeliness and hop count, determining model constraints, specifically including: Step 2-1, the satellite network model is defined as G(N, C, L, T), wherein N represents a set of satellite nodes, |N| = V, V is the total number of satellites; C = (c ij ) V×V is a matrix describing link cost, an element c ij in the matrix C is used to describe the link cost between satellite i and satellite j; L = (l ij ) V×V is a matrix describing the connection of satellites, l ij = 1 indicates that there is a connection between satellite i and satellite j, and l ij = 0 indicates that there is no connection between satellite i and satellite j; is a matrix describing the expected maintenance time of a communication link, which represents the expected maintenance time of a communication link between satellite i and satellite j; Step 2-2, establishing the following link cost equation: Where t1 represents the satellite's movement over a period of time after the routing rules are successfully sent to the satellite, c ij The minimum time, t2 represents c. ij The biggest moment; This represents the element in matrix T that has the largest reciprocal of the expected duration of the communication link at time t. This represents the element in matrix T with the smallest reciprocal of the expected duration of the communication link at time t. hop(i,j) = 1 indicates that when satellite i and satellite j communicate directly, the number of hops between them is 1. c ij c ij The value corresponding to time t1 is its lower bound. It is C ij The value corresponding to time t2 is its upper limit; Step 2-3, establishing the following constraint equation: wherein formula (4) represents that there is a communication link only when the distance d ij between satellite i and satellite j is less than the maximum communication distance D ij of both, formula (5) represents the value of c ij under different conditions, and formula (6) represents the value of under different conditions; Step 3, based on the acceptable degree, evaluating the current feasible routing strategy set, establishing evaluation indexes, specifically including: Step 3-1, using the average value and interval radius to represent the link cost, as shown in the following equation: Step 3-2, for any two links (i,j) and (u,v), is expressed as A = [m ij ,w ij ] and B = [m uv ,w uv ] according to equations (7) and (8); when m ij ≤ m uv , record A < B, indicating that A has a smaller link cost than B; define ζ(A < B) as the acceptable degree, as shown in the following equation: Step 3-3, establishing the following acceptable degree classification: When ζ(A < B) = 0, the assumption that "A has a smaller link cost than B" is not accepted; When 0 < ζ(A < B) < 1, the decision maker accepts the assumption that "A has a smaller link cost than B", with a satisfaction degree of 0 to 1; when ζ(A < B) ≥ 1, the decision maker is very satisfied with the assumption that "A has a smaller link cost than B"; Step 3-4, for any two satellite nodes s and t, let P st denote all s-t routes, and for any s-t route, define as follows: p = {e1, e2,..., e n ,...,e Q}(11) where e n = (i, j) denotes the link between node i and node j, n = 1, 2, …, Q, Q denotes the number of links in the routing policy p; Step 3-5, establishing the following routing strategy cost equation: wherein cost(p) represents the routing policy cost, m p is the average of the routing policy cost, w p is the interval radius of the routing policy cost; Step 3-6, establishing the following reference routing equation: where p rf represents a reference route, p any represents an arbitrary P st route strategy in p c uv , is the link cost c uv of link (u, v) using the upper and lower bounds expressed by equation (3); Step 3-7, to record p rf , the route cost of m rf , w rf , where m rf , w rf , the definition of formula (13), (14), respectively, as follows to establish the route acceptable degree equation: wherein p eval represents waiting for the route to be accepted for evaluation of the degree of acceptability; Step 4, according to the evaluation index, establishing the following optimization objective: Maxζ(p eval < p rf )(17) Step 5, designing a one-way A-star algorithm evaluation function to calculate the reference route; Step 6, designing a two-way A-star algorithm evaluation function to obtain the routing strategy set with the highest overall acceptable degree based on the reference route as the current optimal routing strategy.

2. The satellite network routing planning method based on bidirectional A* algorithm according to claim 1, characterized in that, Step 5 includes: Step 5-1, establishing the following reference route cumulative cost function: where e1,..., e n ∈p rf ; Step 5-2, establishing the following reference route target point prediction cost function: where d jt represents the current link e n = (i,j) between node j and the target node t, and represents the ceiling function, and D represents the maximum communication distance between satellite nodes. Step 5-3, establishing the following reference route evaluation function: frf(n) = grf(n) + hrf(n) (20). 3.The satellite network routing planning method based on bidirectional A* algorithm of claim 2, wherein, Step 6 includes: Step 6-1, establishing the following routing cumulative cost function: wherein e1,…,e n ∈p o ; Step 6-2, establishing the following routing evaluation function: wherein h o The expression of (n) is the same as formula (19); Step 6-3, for reference route p rf whose route representation is given by equation (11), select as a demarcation point, route p rf is divided into p rff , p rfs as shown in the following equation: where p rff is the first half of the reference route, p rfs is the second half of the reference route, satisfying p rf = p rff ∪ p rfs ; Step 6-4, establishing the following optimization sub-objective: Maxζ(p bf rff )(25)​ Maxζ(p bs < p rfs )(26) wherein p bf is the optimal route under the first half of the reference route, with s as the starting point and j as the target point, and p bs is the optimal route under the second half of the reference route, with t as the starting point and j as the target point. Step 6-5, according to p bf ,p bs , the following optimal route is obtained: p b = p bf ∪ p bs (27).

4. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to realize the satellite network routing planning method based on the two-way A-star algorithm as claimed in any one of claims 1-3.

5. A computer-readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to realize the satellite network routing planning method based on the two-way A-star algorithm as claimed in any one of claims 1-3.

Images