An Optimization Deployment Method and System for Communication Relay Nodes in a Cooperative Action System

Abstract

The present invention discloses a method and system for optimizing the layout of communication relay nodes in a cooperative action system. Based on the quantitative description of the attribute information of the cooperative action system, from the perspective of communication connection in the physical layer and information interaction in the logical layer, decision variables are defined, constraint conditions are analyzed, and an objective function is designed to construct a two-layer communication relay node optimization layout model including the physical layer and the logical layer. According to the hierarchical characteristics of the model, based on the artificial bee colony algorithm with high fitness and fast convergence speed, combined with the non-feasible solution forced conversion mechanism, a hierarchical artificial bee colony system for upper and lower layer iterative games is constructed. Through the iterative solution of the hierarchical artificial bee colony system, an optimized layout scheme of communication relay nodes that meets the quality requirements of communication information services is generated, solving the problems of insufficient quantity, lack of optimality, and low adaptability in the existing field of optimizing the layout modeling of communication relay nodes, and being unable to generate an optimized layout of communication relay nodes with relatively high quality and efficiency well.

Description

Technical Field

[0001] The present invention relates to the technical field of aerial communication relay, and particularly to a method and system for optimizing the layout of communication relay nodes in a collaborative operation system. Background Art

[0002] In a collaborative operation system, the operation planning department usually assigns tasks with similar geographical locations and resource requirements to a task group composed of multiple operation nodes according to the geographical distribution and resource requirements of all tasks in the mission, so as to effectively reduce the voyage cost during the task execution of the operation nodes. In the actual battlefield environment, due to the limited communication capabilities of the nodes within the task group and the vulnerability of communication to terrain and landforms, direct communication may not be possible between different task groups. Therefore, it is mainly considered to use unmanned aerial vehicles (UAVs), airships, balloons, etc. as communication relay nodes, and through the reasonable layout of these relay nodes, to achieve full communication connectivity between different task groups. Regarding the problem of the layout of high-altitude communication relay nodes, researchers have conducted in-depth studies, mainly focusing on aspects such as the optimal layout of the spatial positions of communication relay nodes, communication topology design, and asynchronous communication mechanisms.

[0003] The invention patent with the publication number CN108718454B discloses a method for collaborative autonomous layout of a multi-UAV communication relay platform. This method first designs a relay node collaboration framework and mechanism, then models the communication geographical environment and electromagnetic environment, and makes full use of the autonomy and mobility characteristics of the unmanned relay platform to plan the positions of UAV nodes using a heuristic algorithm. However, although the heuristic algorithm used in this method can generate a node position planning scheme relatively quickly, it is difficult to guarantee the optimality of the solution.

[0004] The patent application document with the publication number CN108156613A discloses a method for laying out relay nodes in a UAV relay multi-hop communication system. It assumes that the wireless communication channel is a Rayleigh fading channel, and constructs an optimization model with the outage probability of the mobile station as the optimization objective and the distance between the base station and the mobile station as the constraint condition. According to the monotonic characteristics of the optimization problem function, the objective function is simplified and the traditional optimization theory is used to solve the model, and then a closed-form solution for the optimal position of the unmanned relay platform is generated. This method can effectively reduce the complexity, but has a convexity requirement for the constructed model and is more applicable to the civilian field.

[0005] The patent application document with the publication number CN110380772A discloses a method for resource allocation and flight route optimization of a UAV relay system. Using a single UAV as the communication relay platform, the weights of the communication bandwidth resource usage rights of different users are designed with the weighted sum of user rates as the decision objective, so as to scientifically and reasonably determine the transmission power flight trajectory of the UAV. This method is targeted at a single-UAV communication relay system and does not consider the situation of multi-UAV collaborative communication relay.

[0006] In summary, most of the existing methods for optimizing the layout of communication relay nodes have certain application limitations, mainly concentrated in aspects such as the single quantity and type of communication relay nodes, relatively idealized models and algorithms, and insufficient scene applicability. It is necessary to further combine the special requirements of the military cooperation operation system for communication relays to generate an optimized layout plan for communication relay nodes that can best meet the communication information quality requirements of the task group for task completion. Summary of the Invention

[0007] Aiming at the above problems, the present invention aims to provide a method and system for optimizing the layout of communication relay nodes in a cooperation operation system to solve the problems of insufficient quantity, sub-optimality, and low adaptability in the field of optimizing the layout of existing communication relay nodes, and being unable to generate an optimized layout of communication relay nodes with relatively high quality and efficiency.

[0008] To achieve the above object, the technical solutions adopted by the present invention are as follows:

[0009] A method for optimizing the layout of communication relay nodes in a cooperation operation system, characterized by including the following steps:

[0010] S1: Obtain the attribute information of the cooperation operation system;

[0011] S2: According to the attribute information of the cooperation operation system obtained in step S1, establish an optimized layout model for communication relay nodes;

[0012] S3: Establish a hierarchical artificial bee colony system corresponding to the optimized layout model of communication relay nodes, and use the iterative method for solution;

[0013] S4: Perform feasibility processing on the infeasible solutions obtained in step S3 to generate an optimized layout plan for communication relay nodes.

[0014] Further, the cooperation operation system described in step S1 includes communication relay nodes and task groups, and the task groups are formed by group leader nodes and group member nodes;

[0015] The task group set is expressed as where tcc i represents the i-th task group, i = 1, 2,..., N tcc , N tcc is the number of task groups;

[0016] The group leader node set is expressed as where N m is the number of group leader nodes, and the position of the j-th (j = 1, 2,..., N m ) group leader node m j in the X-Y coordinate system is

[0017] The set of communication relay nodes is represented as where N zj is the number of communication relay nodes, and the k-th (k = 1, 2,..., N zj ) communication relay node zj k has the projected coordinates in X-Y as

[0018] Furthermore, the specific operations of step S2 include the following steps

[0019] S201: Determine the decision-level variables of the communication relay nodes, where the decision-level variables include physical-layer decision variables and logical-layer decision variables;

[0020] S202: Determine the constraint conditions of the communication relay node optimization layout model and establish the communication relay node optimization layout model; the constraint conditions include physical-layer constraint conditions and logical-layer constraint conditions;

[0021] S203: Determine the objective function of the communication relay node optimization layout model, where the objective function includes physical-layer objective function and logical-layer objective function.

[0022] Furthermore, the physical-layer decision variables in step S201 include: δ ij is the decision variable representing whether the representative group head node m j belongs to the i-th task group tcc i , where δ ij = 1 indicates that m j belongs to tcc i , and δ ij = 0 indicates that m j does not belong to tcc i ;

[0023] ξ jk is the decision variable representing whether the representative group head node m j is covered by the k-th communication relay node zj k , where ξ jk = 1 indicates that m j is covered by zj k , and ξ jk = 0 indicates that m j is not covered by zj k ;

[0024] The logical-layer decision variables in step S201 include: ζ ik is the decision variable representing whether the i-th task group tcc i transmits information through the k-th communication relay node zj k , where ζ ik = 1 indicates that tcci Transmit information through zj k ζ = 0 indicates that tcc ik does not transmit information through zj i ζ = 1 is a necessary condition for k transmit information through zj ik The necessary condition for ζ = 1 is

[0025] (i≠i′ and k≠k′) represents that when realizing the information transmission from tcc i to tcc i′ (i′ = 1, 2,..., N tcc ), it is a decision variable indicating whether there is a transmission path from zj k to zj k′ (k′ = 1, 2,..., N zj ). Indicates that there is such a transmission path Indicates that there is no such transmission path; The necessary condition for ζ = 1 is that zj k and zj k′ must be physically connected, that is, ||b k′ -b k ||≤R1, where b k′ is the projection coordinate of the k′-th communication relay node zj k′ in X - Y, R1 is the communication radius between the communication relay node zj k and zj k′ , and ||·|| is the Euclidean distance calculation operator.

[0026] Furthermore, the physical layer constraint condition in step S202 is

[0027]

[0028] where R2 is the communication radius between the communication relay node zj k and the group header node m j ;

[0029] The logical layer constraint condition is

[0030]

[0031] where is the communication bandwidth requirement of the i-th task group tcc i , B max is the communication bandwidth threshold of all communication relay nodes; Bool(·) is a Boolean function.

[0032] Furthermore, the physical layer objective function in step S203 is min N zj, the optimal layout model of the physical layer is

[0033]

[0034] The objective function of the logic layer is The optimal layout model of the logic layer is

[0035]

[0036] Among them, is the average communication service quality obtained by all task groups, P i is the communication service quality obtained by tcc i , that is, the maximum value of the signal power received by the group head node within the task group;

[0037] is the average number of routing hops from all task groups to other task groups, For any task group tcc i to other task groups tcc i′ the number of routing hops is

[0038]

[0039] Furthermore, the specific operations of step S3 include the following steps

[0040] S301: Establish a standard artificial bee colony system;

[0041] S302: Establish a hierarchical artificial bee colony system corresponding to the communication relay node optimal layout model;

[0042] Take the coordinate positions of the communication relay nodes as code elements. In the population X, the number of rows is N ZQ , each row represents a solution; the number of columns is N W , the odd-numbered columns represent the abscissa, and the even-numbered columns represent the ordinate; determine the values of ξ jk and by calculating the relative position relationships between the communication relay nodes and the group head nodes, and between the communication relay nodes; design a hierarchical artificial bee colony system based on the double decision-maker game idea. During the evolutionary game process, the upper and lower layers perform optimization according to the evolutionary results of each other;

[0043] S303: Solve the hierarchical artificial bee colony system established in step S302 by the iterative method.

[0044] Furthermore, the specific operations of step S303 include the following steps

[0045] S3031: Initialize the upper-layer planning population X, set the number of iterations, the upper and lower bounds of the odd-numbered columns are the upper and lower bounds of the abscissa positions of all group head nodes, and the upper and lower bounds of the even-numbered columns are the upper and lower bounds of the ordinate positions of all group head nodes;

[0046] S3032: Use the artificial bee colony algorithm to obtain the optimal solution Y of the lower-layer planning with X as the input * , and use Y * as the input, and use the artificial bee colony algorithm to obtain the optimal solution X of the upper-layer planning * ;

[0047] S3033: If it is the first iteration, perform constraint processing on X * and Y * to obtain X1 * and Y1 * , then, calculate the corresponding fitness values for X1 * and Y1 * , perform non-dominated sorting and crowding distance calculation on all 2×N ZQ fitness values, and select the population corresponding to the first N ZQ fitness values as X; If it is not the first iteration, perform constraint processing on X, X * and Y * to obtain X1, X1 * and Y1 * , then, calculate the corresponding fitness values for X1, X1 * and Y1 * , perform non-dominated sorting and crowding distance calculation on all 3×N ZQ fitness values, and select the population corresponding to the first N ZQ fitness values as X;

[0048] S3034: If the maximum number of iterations is reached, after performing constraint processing on X to obtain X1, calculate the corresponding fitness value for X1, perform non-dominated sorting on all N ZQ fitness values, and extract the population corresponding to the Pareto front fitness value as the optimal solution set for output.

[0049] Further, the specific operations in step S4 include the following steps.

[0050] S401: Perform redundancy removal operations; delete the communication relay nodes that do not affect the connection relationship, and the judgment criterion is that after deleting the communication relay node, it satisfies:

[0051] The value of will not change before and after deletion;

[0052] The value of will not change before and after deletion.

[0053] S402: Perform the first step of the deficiency - making - up operation; determine whether all task groups are covered by at least one communication relay node, that is whether it holds. If it holds, perform the second step of the deficiency - making - up operation in step S403; if not, select the established task group tcc i , and randomly select m ij for which δ j = 1 holds. Take a circle with its coordinate position as the center and R1 as the radius. Take the coordinates of supplementary communication relay nodes at any position inside the circle until holds;

[0054] S403: Perform the second step of the deficiency - making - up operation; determine whether the undirected graph formed by all communication relay nodes is fully connected, that is whether it holds. If it holds, go to step S404 to perform the redundancy - removing operation again; if not, randomly select two from all connected components, select the two communication relay nodes with the shortest distance in the two connected components to form a line segment s. Starting from one communication relay node, take points on s with a step size of (0, 2R2) as the coordinates of supplementary communication relay nodes until holds;

[0055] S404: Perform the redundancy - removing operation again. The specific operation process is the same as that in step S401.

[0056] Furthermore, a communication relay node optimal layout system for a coordinated action system is characterized in that the system includes a coordinated action system model module, a communication relay node optimal layout model module, and a hierarchical artificial bee colony module; the coordinated action system model module is used to construct a coordinated action system model through attribute information, the communication relay node optimal layout model module is used to construct a communication relay node optimal layout model, and the hierarchical artificial bee colony module is used to establish a hierarchical artificial bee colony system corresponding to the communication relay node optimal layout model and use the iterative method to solve it, and perform feasible processing on non - feasible solutions;

[0057] The system executes the communication relay node optimal layout method for the coordinated action system as described above.

[0058] The beneficial effects of the present invention are:

[0059] 1. The present invention utilizes communication relay nodes to construct a backbone network among task groups. For the problem of optimizing the layout of communication relay nodes, a model construction + solution method is adopted to generate an optimized layout scheme for communication relay nodes. The problem is decoupled by distinguishing the physical layer and the logical layer. In the physical layer, communication connection constraints depending on the communication radius and node distance are considered, and in the logical layer, constraints on link bandwidth and the relationship between two-layer decision variables are considered. The objective function in the physical layer aims to minimize the use of communication relay nodes, and the objective function in the logical layer aims to achieve the best communication quality and the fewest relay hops. A hierarchical artificial bee colony algorithm with double-layer iterative game is used to solve the model. For infeasible solutions, forced conversion is carried out to ensure the search directionality of the optimized layout scheme for communication relay nodes. The present invention overcomes the defects of existing methods for optimizing the layout of communication relay nodes and improves the effectiveness and superiority of optimizing the layout of communication relay nodes.

[0060] 2. The optimized layout model of communication relay nodes in the collaborative action system of the present invention distinguishes physical layer communication connection and logical layer information interaction, fully reflecting the actual requirements of information communication guarantee in the collaborative action system, that is, the physical layer communication connection ensures "connection", and the logical layer information interaction ensures "smoothness", overcoming the problem of simply considering single-layer optimization in existing problem modeling.

[0061] 3. The method for optimizing the layout of communication relay nodes in the collaborative action system of the present invention adopts a hierarchical artificial bee colony algorithm with upper and lower layer iterative game optimization for the double-layer optimization problem including the physical layer and the logical layer, solving various problems such as insufficient optimality of existing methods, low quality and efficiency of communication relay node layout schemes, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is a schematic diagram of the structure of the collaborative action system in the present invention;

[0063] Figure 2 It is an equivalent communication topology structure diagram of the collaborative action system in the present invention;

[0064] Figure 3 It is a layout diagram of communication relay nodes satisfying communication connection constraints in the present invention;

[0065] Figure 4 It is a hierarchical artificial bee colony population cyclic game diagram in the present invention;

[0066] Figure 5 It is a layout diagram of communication relay nodes corresponding to a typical solution in the simulation experiment of the present invention;

[0067] Figure 6 It is a Pareto solution distribution diagram in the simulation experiment of the present invention;

[0068] Figure 7 It is a decomposition diagram of Pareto solutions in three objective function values in the simulation experiment of the present invention;

[0069] Figure 8 This is the comparison chart of the algorithm coverage index in the simulation experiment of the present invention;

[0070] Figure 9 This is the comparison chart of the algorithm uniformity index in the simulation experiment of the present invention;

[0071] Figure 10 This is the comparison chart of the algorithm breadth index in the simulation experiment of the present invention. Detailed implementation manners

[0072] In order to enable those of ordinary skill in the art to better understand the technical solution of the present invention, the technical solution of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0073] An optimized layout method for communication relay nodes in a coordinated action system includes the following steps,

[0074] S1: Obtain the attribute information of the coordinated action system;

[0075] Specifically, as shown in the appendix Figure 1 described, the coordinated action system includes communication relay nodes and task groups. A task group is a set of action nodes that perform the same type of tasks formed by a group leader node and member nodes. The division of task groups is mainly determined according to task types, node capabilities, and affiliation relationships. The group leader node is the superior node within the task group, responsible for commanding and controlling the member nodes, as well as communicating with other task groups and superior nodes; the member node is the inferior node within the task group, responsible for accepting the command and control of the group leader node and specifically executing the corresponding tasks.

[0076] Represent the set of task groups as where tcc i represents the i-th task group, i = 1, 2,..., N tcc , N tcc is the number of task groups.

[0077] Represent the set of group leader nodes as where N m is the number of group leader nodes, and the position of the j-th (j = 1, 2,..., N m ) group leader node m j in the X-Y coordinate system is It should be noted here that the set of group leader nodes includes all the group leader nodes in all task groups.

[0078] Represent the set of communication relay nodes as where N zj is the number of communication relay nodes, and the k-th (k = 1, 2,..., N zj ) communication relay node zjk The projected coordinates in X-Y are

[0079] Furthermore, step S2: Based on the collaborative action system attribute information obtained in step S1, establish an optimized layout model for communication relay nodes;

[0080] Specifically, S201: Determine the decision-making layer variables of the communication relay nodes;

[0081] Considering that communication relay nodes mainly achieve information interaction at the logical layer between the head nodes of different task groups on the basis of providing physical layer communication connections. Therefore, the decision variables mainly include two categories: physical layer decision variables and logical layer decision variables.

[0082] The physical layer decision variables include: (1) δ ij Is the decision variable representing whether the head node m j belongs to the i-th task group tcc i . δ ij =1 indicates that m j belongs to tcc i , and δ ij =0 indicates that m j does not belong to tcc i .

[0083] (2) ζ jk Is the decision variable representing whether the head node m j is covered by the k-th communication relay node zj k . ζ jk =1 indicates that m j is covered by zj k , and ξ jk =0 indicates that m j is not covered by zj k .

[0084] The logical layer decision variables include: (1) ζ ik Is the decision variable representing whether the i-th task group tcc i transmits information through the k-th communication relay node zj k . ζ ik indicates that tcc i transmits information through zj k , and ζ ik =0 indicates that tcc i does not transmit information through zj k . The necessary condition for ζ ik =1 is

[0085] (2) (i ≠ i' and k ≠ k') is used to represent whether there is a transmission path from zj i to tcc i′ (′ = 1, 2,..., N tcc ) during the information transmission from tcc k to zj k′ (k′ = 1, 2,..., N zj ). It is a decision variable for the transmission path. Indicates that there is such a transmission path. Indicates that there is no such transmission path. The necessary condition for k zj k′ and zj k′ -b k || ≤ R1, where b k′ is the projection coordinate of the k′-th communication relay node zj k′ in the X-Y plane, R1 is the communication radius between the communication relay node zj k and zj k′ . ||·|| is the Euclidean distance calculation operator.

[0086] S202: Determine the constraint conditions of the communication relay node optimization layout model;

[0087] In terms of constraint conditions, it also includes physical layer constraint conditions and logical layer constraint conditions.

[0088] Among the physical layer constraint conditions, the layout of communication relay nodes needs to meet the following two conditions: First, for any communication relay node zj k′ , there is at least one other communication relay node zj k connected to it; Second, for any task group tcc i , there is at least one group head node m j within the relay range of the communication relay node zj k .

[0089] Therefore, the physical layer communication connection constraint conditions are

[0090]

[0091] where R2 is the communication radius between the communication relay node zj k and the group head node m j . The first constraint ensures that all communication relay nodes are connected to at least one other communication relay node, and the second constraint ensures that any task group is covered by at least one communication relay node.

[0092] In the logical layer constraints, due to the limitation of the communication link bandwidth, the communication bandwidth provided by each communication relay node must not be greater than its communication bandwidth threshold, that is, there is

[0093]

[0094] Among them, is the communication bandwidth requirement of the i-th task group tcc i and B max is the communication bandwidth threshold of all communication relay nodes.

[0095] Considering the relationship between δ ij , ζ jk and ζ ik there is

[0096]

[0097] In addition, considering that the information interaction in the logical layer depends on the communication connection in the physical layer, there is

[0098]

[0099] Among them, Bool(·) is a Boolean function.

[0100] S203: Determine the objective function of the communication relay node optimization layout model and establish the communication relay node optimization layout model;

[0101] Corresponding to the decision variables and constraints, the objective function also includes two aspects: the physical layer and the logical layer.

[0102] In terms of the physical layer objective function, considering that the communication relay nodes are of high value and limited in number, therefore, it is necessary to use as few communication relay nodes as possible to achieve the communication relay service for the cooperative action system. Then, the objective function of the model is defined as

[0103] min N zj (6)

[0104] To sum up, the optimization layout model of the communication relay nodes in the physical layer of the cooperative action system can be established as

[0105]

[0106] In terms of the logical layer objective function, usually, the communication service quality provided by the communication relay nodes to each task group is different, and its value is related to the transmission power of the communication antenna of the communication relay platform received by each group head node, the allocation of the group head nodes within the task group, and the situation of the task group selecting a communication relay node for relay. The specific solution process is as follows.

[0107] As shown in the appendix Figure 2The equivalent communication topology structure diagram of the collaborative action system shown, assuming that the wireless communication channel between the communication relay node and the group head node is a Rice channel. Therefore, m j receives the signal power of zj k as

[0108]

[0109] where P k is the transmission power of zj k , θ is a constant, α is the path loss attenuation factor, and h is the height of the communication relay node.

[0110] If the group head node m j is simultaneously covered by multiple communication relay nodes, without considering the mutual interference, the effective signal power received by m j is

[0111]

[0112] Then the communication service quality obtained by tcc i is the maximum value of the signal power received by the group head node within the task group, that is, there is

[0113]

[0114] Therefore, maximizing the average communication service quality obtained by all task groups is

[0115]

[0116] In addition, in order to achieve the maximum efficiency of information flow in the collaborative action system, it is necessary to minimize the average number of routing hops from a certain task group to all other task groups. First, calculate the number of routing hops from any task group tcc i to other task groups tcc i′ as

[0117]

[0118] Then minimizing the average number of routing hops from all task groups to other task groups is

[0119]

[0120] To sum up, the optimization layout model of the communication relay node logic layer of the collaborative action system can be established as

[0121]

[0122] The layout diagram of the communication relay node that satisfies the communication connection constraint conditions is shown in the appendix Figure 3 .

[0123] Further, in step S3: establish a hierarchical artificial bee colony system corresponding to the communication relay node optimization layout model, and use the iterative method for solution;

[0124] Specifically, S301: establish a standard artificial bee colony system;

[0125] The artificial bee colony algorithm is an intelligent optimization algorithm that simulates the foraging behavior of bees. It divides the artificial bee colony into employed bees, onlooker bees, and scout bees, and the types of these three bees are converted under certain conditions. The position of each nectar source represents a solution, and the amount of nectar source characterizes the quality of the solution. Iterative optimization is carried out through the initialization stage, employed bee stage, onlooker bee stage, and scout bee stage.

[0126] Initialization stage: If the population size is N ZQ , and the nectar source dimension is N W , each nectar source position is randomly generated within the feasible interval, and its calculation formula is

[0127]

[0128] In the formula, x ij is the nectar source position, and are respectively the upper and lower bounds of the nectar source position value, and rand(0, 1) represents a random number with a value from 0 to 1.

[0129] Employed bee stage: The employed bees first perform local search near the nectar source and generate a new nectar source position according to formula (16)

[0130] x ij ′ = x ij + rand(-1, 1) × (x ij - x kj ) (16)

[0131] where i ≠ k, and x kj is a randomly selected number among all nectar sources, and rand(-1, 1) represents a random number with a value from -1 to 1. Compare the new nectar source with the old one. If the position of the new nectar source is better than that of the old one, replace it; otherwise, keep it unchanged.

[0132] Onlooker bee stage: After all employed bees complete local search, they share the nectar source position information with onlooker bees through the waggle dance. The onlooker bees adopt the binary tournament selection operator to select a nectar source, that is, randomly select two nectar sources Z1 and Z2, and compare the corresponding objective function (fitness function) values F1 and F2. If there is

[0133] F1 > F2 (17)

[0134] then select Z1 and update the nectar source position according to formula (16). If there is

[0135] F2 > F1 (18)

[0136] Then select Z2 and update the nectar source position according to Equation (16). Here, > means Pareto non-inferior to.

[0137] Scout bee stage: After performing Limit iterations, if the fitness value of a nectar source position does not improve, the employed bees at that nectar source are transformed into scout bees. After abandoning the original nectar source, a new nectar source position is generated according to Equation (15).

[0138] When using the artificial bee colony algorithm to solve multi-objective problems, the sorting of fitness is mainly through non-dominated sorting and crowding distance calculation, and the Pareto front is obtained.

[0139] S302: Establish a hierarchical artificial bee colony system corresponding to the communication relay node optimization layout model;

[0140] Considering that the communication relay node optimization layout problem shown in Equations (7) and (14) is a multi-decision variable two-layer programming problem, how to encode and decode is an important issue. In the present invention, the coordinate positions of the communication relay nodes are used as code elements. Taking the population X as an example, the number of rows is N ZQ , and each row represents a solution; the number of columns is N W , the odd-numbered columns represent the abscissa, and the even-numbered columns represent the ordinate. To simplify the problem, δ ij is determined in advance, and under a specific population X, by calculating the relative position relationships between the communication relay nodes and the group head nodes, and between the communication relay nodes and the communication relay nodes, ξ jk and the values of are determined, and the value of ξ jk depends on ξ jk and the settings in the specific solution process.

[0141] To solve Equations (7) and (14), a hierarchical artificial bee colony system is designed based on the double decision-maker game idea. In the evolutionary game process, the upper and lower layers (Equations (7) and (14)) always optimize according to the evolutionary results of each other. As the number of iterations increases, the entire system gradually reaches an equilibrium state.

[0142] S303: Solve the hierarchical artificial bee colony system established in step S302 by the iterative method, and the specific operations are as shown in the appendix Figure 4 as follows.

[0143] Initialize the upper-layer planning population X according to Equation (15), set the number of iterations, the upper and lower bounds of the odd-numbered column values are the upper and lower bounds of the abscissa positions of all group head nodes, and the upper and lower bounds of the even-numbered column values are the upper and lower bounds of the ordinate positions of all group head nodes.

[0144] The optimal solution Y of the lower-level planning is obtained by using the artificial bee colony algorithm with X as the input. * , and Y * is used as the input, and the artificial bee colony algorithm is used to obtain the optimal solution X of the upper-level planning. * .

[0145] If it is the first iteration, constraint processing is performed on X * and Y * respectively to obtain X1 * and Y1 * . Then, the corresponding fitness values of X1 * and Y1 * are calculated. Non-dominated sorting and crowding distance calculation are performed on all 2×N ZQ fitness values, and the population corresponding to the first N ZQ fitness values is selected as X. If it is not the first iteration, constraint processing is performed on X, X * and Y * respectively to obtain X1, X1 * and Y1 * . Then, the corresponding fitness values of X1, X1 * and Y1 * are calculated. Non-dominated sorting and crowding distance calculation are performed on all 3×N ZQ fitness values, and the population corresponding to the first N ZQ fitness values is selected as X.

[0146] If the maximum number of iterations is reached, after constraint processing is performed on X to obtain X1, the corresponding fitness value of X1 is calculated. Non-dominated sorting is performed on all N ZQ fitness values, and the population corresponding to the fitness values at the Pareto front end is extracted as the optimal solution set for output.

[0147] Furthermore, step S4: Feasibility processing is performed on the infeasible solutions obtained in step S3 to generate an optimized layout scheme for communication relay nodes;

[0148] Specifically, in the solution process, since the coordinate positions of communication relay nodes are used as code elements and the number of communication relay nodes is variable. Therefore, too large an N W value will increase the algorithm search space and reduce the algorithm solution speed; while too small an N W value will make it difficult to always satisfy the problem constraints, and the scientific setting of the N W value needs to be reasonably considered. In the present invention, N W is set to take a value 1.5 times that of the communication relay nodes required to cover all group head nodes.

[0149] In addition, Equations (7) and (14) represent a multi-constraint combinatorial optimization problem. Currently, the techniques for dealing with infeasible solutions in combinatorial optimization problems mainly include two categories: one is to retain infeasible solutions and handle them by defining the degree of constraint violation, constraint penalty functions, or by using constraint domination relationships and external archive techniques; the other is not to retain infeasible solutions and force the infeasible solutions to satisfy all constraints to transform them into feasible solutions. Which infeasible solution handling technique to adopt depends on the structure of the solution space. If the proportion of feasible solutions in the solution space is relatively large, the former is more applicable; conversely, the latter is more applicable. Considering that it is difficult to estimate the solution space of the problems in Equations (7) and (14) and it is difficult to determine the proportion of feasible solutions, from the perspective of algorithm feasibility, the present invention adopts the method of making infeasible solutions feasible.

[0150] S401: Perform redundancy removal operation. Delete the communication relay nodes that do not affect the connection relationship. The judgment criterion is that after deleting the communication relay node, the following conditions are satisfied:

[0151] (1) The value of will not change before and after deletion;

[0152] (2) The value of will not change before and after deletion.

[0153] S402: Perform the first step of the operation to make up for deficiencies. Judge whether all task groups are covered by at least one communication relay node, that is, whether it holds. If it holds, perform the second step of the operation to make up for deficiencies in step S403; if it does not hold, select the task group tcc for which holds i , and randomly select m ij for which δ j = 1 holds. Take the coordinates of the center of the circle with its coordinate position as the center and R1 as the radius, and take points at any position within the circle as the coordinates of the supplementary communication relay nodes until holds.

[0154] S403: Perform the second step of the operation to make up for deficiencies. Judge whether the undirected graph formed by all communication relay nodes is fully connected, that is, whether it holds. If it holds, go to step S404 to perform the redundancy removal operation again; if it does not hold, randomly select two from all connected components, select the two communication relay nodes with the shortest distance in the two connected components to form a line segment s, and starting from one communication relay node, take points on s with a step size of (0, 2R2) as the coordinates of the supplementary communication relay nodes until holds.

[0155] S404: Perform the redundancy removal operation again. The specific operation process is the same as that in step S401. Through the above steps, the non-feasible solutions can be effectively processed to obtain feasible solutions that meet all the constraint conditions, and finally an optimized layout scheme for communication relay nodes is obtained.

[0156] Simulation experiment:

[0157] All the simulation experiments in the present invention are carried out on a Lenovo computer configured with an Inter(R) Dual-Core CPU 3.06 GHz, using MATLAB 2019a as the simulation platform. In terms of the setting of scenario parameters: the number of task groups is N toc = 7, the number of group head nodes is N m = 13, the communication distance between communication relay nodes is R1 = 24, the communication distance between a communication relay node and a group head node is R2 = R1 / 2 = 12, the height of the communication relay node is h = 20, the signal transmission power of the communication relay node is P0 = 10, θ = 1, α = 2. All values are abstract values and can be assigned according to specific situations in the actual scenario. In terms of the algorithm setting: the number of nectar sources N zq = 20, the number of iterations Ite = 50, Limit = 5. As shown in Table 1, it is the coordinate information of the group head nodes.

[0158] Table 1 Coordinate information of group head nodes

[0159] Serial number Coordinate Serial number Coordinate Serial number Coordinate 1 (22.18,72.44) 6 (25.82,15.50) 11 (30.26,50.45) 2 (20.35,78.14) 7 (31.52,18.62) 12 (34.89,44.62) 3 (62.41,65.72) 8 (28.13,18.43) 13 (55.78,31.56) 4 (54.86,70.79) 9 (89.34,34.30) -- -- 5 (8.06,35.54) 10 (77.98,34.63) -- --

[0160] In addition, in terms of the attribution of group head nodes, m1 and m2 belong to tcc1, m3 and m4 belong to tcc2, m5 belongs to tcc3, m6, m7 and m8 belong to tcc4, m9 and m 10 belong to tcc5, m 11 and m 12 belong to tcc6, m 13 belongs to tcc7.

[0161] I. Verification of method effectiveness

[0162] To verify the effectiveness of the method of the present invention, an optimized layout scheme of communication relay nodes under a set of typical solutions is used for illustration. As shown in the appendix Figure 5 , it is the layout of communication relay nodes corresponding to this typical solution. As can be seen from the appendix Figure 5 , under this typical solution, for any task group, there is at least one group head node within the coverage range of at least one communication relay node, and any communication relay node is at least connected to at least one other communication relay node to form a communication connection relationship. Therefore, this typical solution meets the constraint conditions. As shown in the appendix Figure 6As shown, it is the distribution relationship of all Pareto solutions in three-dimensional space. From the appendix Figure 6 it can be seen that the Pareto solutions are evenly and reasonably distributed.

[0163] II. Verification of Parameter Influence

[0164] To verify the influence of key parameters on the objective function values, R1 is set to 9, 12, 15, 18, and 21 respectively. The Pareto solutions in the three objective function value decomposition diagrams are as shown in the appendix Figure 7 As shown, from the appendix Figure 7 it can be seen that the value of R1 has a great influence on the three optimized objective function values. As R1 increases, the average number of routing hops decreases, and the required communication relay nodes also decrease, while decreases accordingly, that is increases accordingly, indicating that the average communication information service quality becomes better, which is in line with the actual situation.

[0165] III. Verification of the Superiority of the Method

[0166] To verify the superiority of the hierarchical artificial bee colony algorithm compared with other algorithms, the hierarchical particle swarm optimization algorithm is selected as the comparison algorithm. The coverage index, uniformity index, and breadth index are selected as the comparison indexes respectively. Among them, the coverage index and the breadth index are benefit-type indexes, the larger the better; the uniformity index is a cost-type index, the smaller the better. As shown in the appendix Figure 8 ~Appendix Figure 10 As shown, they are the comparison situations of the coverage index, uniformity index, and breadth index respectively. From the appendix Figures 8 - 10 it can be seen that the hierarchical artificial bee colony algorithm is generally better than the hierarchical particle swarm optimization algorithm in all three indexes.

[0167] For the definition of the coverage index, the present invention measures it by comparing the coverage rates between the Pareto solution sets generated by the hierarchical artificial bee colony algorithm and the hierarchical particle swarm optimization algorithm. If the Pareto solution sets of the two algorithms are Y1 and Y2 respectively, then for Y1 and Y2, the coverage rate of Y1 to Y2 is defined as

[0168]

[0169] For the definition of the uniformity index, taking the uniformity index E(Y1) of the Pareto solution set Y1 as an example, its calculation formula is defined as

[0170]

[0171] where d i is the minimum value of the distance between the i-th Pareto solution and other Pareto solutions in the Pareto solution set, is all d iThe average value, define d i The calculation formula is N λ is the number of optimization objectives, and are the values of the i-th and i'-th Pareto solutions on the j-th optimization objective, respectively.

[0172] For the definition of the breadth index, taking the breadth index D(Y1) of the Pareto solution set Y1 as an example, its calculation formula is defined as

[0173]

[0174] Among them, and are the maximum and minimum values corresponding to the j-th optimization objective in the objective function values of the solution set Y1, respectively.

[0175] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing the layout of communication relay nodes in a coordinated action system, characterized in that, including the following steps, S1: Obtain the attribute information of the cooperative action system; S2: Establish an optimized layout model of communication relay nodes according to the attribute information of the cooperative action system obtained in step S1; S3: Establish a hierarchical artificial bee colony system corresponding to the optimized layout model of communication relay nodes, and use the iterative method for solution; S4: Perform feasibility processing on the infeasible solutions obtained in step S3 to generate an optimized layout plan for communication relay nodes; wherein, the cooperative action system described in step S1 includes communication relay nodes and task groups, and the task groups are formed by group leader nodes and group member nodes; Represent the task group set as wherein, tcc i represents the i-th task group, i = 1, 2, …, N tcc , N tcc is the number of task groups; The set of group head nodes is denoted as where N m is the number of group head nodes, and the j-th group head node m j has a position in the X-Y coordinate system as j = 1, 2,..., N m ; The set of communication relay nodes is represented as where N zj is the number of communication relay nodes, and the projection coordinates of the k-th communication relay node zj k in X-Y are The optimized layout model of the communication relay node described in step S2 includes the optimized layout model of the physical layer and the optimized layout model of the logical layer; the objective function of the physical layer is min N zj , and the objective function of the logical layer is In the formula, N zj is the number of communication relay nodes, is the average communication service quality obtained by all task groups, is the average number of routing hops from all task groups to other task groups; In step S3, a hierarchical artificial bee colony system is designed based on the double decision-maker game idea. During the evolutionary game process, the optimized layout models of the physical layer and the logical layer optimize according to the evolutionary results of each other. As the number of iterations increases, the entire system gradually reaches an equilibrium state.

2. The method for optimizing the layout of communication relay nodes in a coordinated action system according to claim 1, wherein The specific operations of step S2 include the following steps, S201: Determine the decision-making layer variables of the communication relay nodes, and the decision-making layer variables include physical layer decision variables and logical layer decision variables; S202: Determine the constraint conditions of the optimized layout model of communication relay nodes, and establish an optimized layout model of communication relay nodes; the constraint conditions include physical layer constraint conditions and logical layer constraint conditions; S203: Determine the objective function of the optimized layout model of communication relay nodes, and the objective function includes a physical layer objective function and a logical layer objective function.

3. The method for optimizing the layout of communication relay nodes in a coordinated action system according to claim 2, wherein, The physical layer decision variables in step S201 include: δ ij is the decision variable representing whether the group header node m j belongs to the i-th task group tcc i ; δ ij = 1 indicates that m j belongs to tcc i , and δ ij = 0 indicates that m j does not belong to tcc i ; ξ jk is the decision variable indicating whether the representative group head node m j is covered by the k-th communication relay node zj k ; ξ jk = 1 indicates that m j is covered by zj k ; ξ jk = 0 indicates that m j is not covered by zj k ; The logical layer decision variables in step S201 include: ζ ik To represent the i-th task group tcc i Whether to use the k-th communication relay node zj k The decision variable for information transmission, ζ ik ζ = 1 indicates that tcc i uses zj k for information transmission, ζ ik ζ = 0 indicates that tcc i does not use zj k for information transmission, ζ ik The necessary condition for ζ = 1 is To indicate whether there is a decision variable for the transmission path from zj i to tcc i′ when implementing the information transmission from tcc k to zj k' i ≠ i′ and k ≠ k′, i′ = 1, 2, …, N tcc , k′ = 1, 2, …, N zj , Indicates that there is such a transmission path, Indicates that there is no such transmission path; The necessary condition for k zj k′ and zj k′ -b k || ≤ R1, where b k′ is the projected coordinate of the k′-th communication relay node zj k′ in X - Y, R1 is the communication radius between the communication relay node zj k and zj k′ , and ||·|| is the Euclidean distance calculation operator.

4. The method for optimizing the layout of communication relay nodes in a coordinated action system according to claim 3, characterized in that, The physical layer constraint condition in step S202 is Among them, R2 is the communication relay node zj k and the group header node m j the communication radius therebetween; The logical layer constraint condition is wherein, is the communication bandwidth requirement of the i-th task group tcc i of which B max is the communication bandwidth threshold of all communication relay nodes; Bool(·) is a Boolean function.

5. A method for optimizing the layout of communication relay nodes in a collaborative action system according to claim 4, characterized in that In step S203, the physical layer objective function is min N zj , and the optimal layout model of the physical layer is The objective function of the logic layer is The optimization layout model of the logic layer is Among them, is the average communication service quality obtained for all task groups, P i is the communication service quality obtained by tcc i that is, the maximum value of the signal power received by the group leader node within the task group; is the average number of routing hops from all task groups to other task groups, any task group tcc i to other task groups tcc i′ the number of routing hops is 6. The method for optimizing the layout of communication relay nodes in a collaborative action system according to claim 5, wherein The specific operations of step S3 include the following steps, S301: Establish a standard artificial bee colony system; S302: Establish a hierarchical artificial bee colony system corresponding to the optimized layout model of communication relay nodes; Taking the coordinate position of the communication relay node as a symbol, in the population X, the number of rows is N ZQ , and each row represents a solution; the number of columns is N W , the odd columns represent the abscissa, and the even columns represent the ordinate; by calculating the relative position relationship between the communication relay node and the group head node, and between the communication relay nodes, ξ jk and the values of are determined; a hierarchical artificial bee colony system is designed based on the double decision-maker game idea. During the evolutionary game process, the upper and lower layers optimize according to the evolutionary results of each other S303: Solve the hierarchical artificial bee colony system established in step S302 by the iterative method.

7. A method for optimizing the layout of communication relay nodes in a coordinated action system according to claim 6, characterized in that, The specific operations of step S303 include the following steps, S3031: Initialize the upper-layer planning population X, set the number of iterations, the upper and lower bounds of the values of the odd-numbered columns are the upper and lower bounds of the abscissa positions of all group leader nodes, and the upper and lower bounds of the values of the even-numbered columns are the upper and lower bounds of the ordinate positions of all group leader nodes; S3032: Use the artificial bee colony algorithm to obtain the optimal solution Y of the lower-level planning with X as the input * , and use Y * as the input, and use the artificial bee colony algorithm to obtain the optimal solution X of the upper-level planning * ; S3033: If it is the first iteration, perform constraint processing on X * and Y * respectively to obtain X1 * and Y1 * . Then, calculate the corresponding fitness values for X1 * and Y1 * . Perform non-dominated sorting and crowding distance calculation on all 2×N ZQ fitness values, and select the population corresponding to the top N ZQ fitness values as X. If it is not the first iteration, perform constraint processing on X, X * and Y * respectively to obtain X1, X1 * and Y1 * . Then, calculate the corresponding fitness values for X1, X1 * and Y1 * . Perform non-dominated sorting and crowding distance calculation on all 3×N ZQ fitness values, and select the population corresponding to the top N ZQ fitness values as X; S3034: If the maximum number of iterations is reached, after performing constraint processing on X to obtain X1, calculate the corresponding fitness value for X1, and for all N ZQ perform non-dominated sorting on the fitness values, and extract the population corresponding to the fitness values at the Pareto front as the optimal solution set for output.

8. A method for optimizing the layout of communication relay nodes in a coordinated action system according to claim 7, characterized in that The specific operations of step S4 include the following steps, S401: Perform redundancy removal operations; delete the communication relay nodes that do not affect the connection relationship, and the judgment criterion is that after deleting the communication relay node, it satisfies: The value will not change before and after deletion; The value will not change before and after deletion; S402: Perform the first step of the deficiency compensation operation; determine whether all task groups are covered by at least one communication relay node, that is whether it holds. If it holds, perform the second step of the deficiency compensation operation in step S403; if not, select the established task group tcc i , and randomly select m ij such that δ j = 1 holds. Take the coordinates of the center of the circle with its coordinate position as the center and R1 as the radius, and take the coordinates of the supplementary communication relay node at any position within the circle until holds; S403: Perform the second step of the deficiency compensation operation; determine whether the undirected graph formed by all communication relay nodes is fully connected, that is Whether it holds. If it holds, go to step S404 to perform the redundancy removal operation again; if it does not hold, randomly select two from all connected components, select the two communication relay nodes with the shortest distance in the two connected components to form a line segment s, starting from one communication relay node, take points on s with a step size of (0, 2R2) as the coordinates of the supplementary communication relay nodes until Holds; S404: Perform redundancy removal operations again, and the specific operation process is the same as that of step S401.

9. A communication relay node optimization deployment system for a collaborative action system, characterized in that The layout system includes a cooperative action system model module, an optimized layout model module of communication relay nodes, and a hierarchical artificial bee colony module; the cooperative action system model module is used to construct a cooperative action system model through attribute information, the optimized layout model module of communication relay nodes is used to construct an optimized layout model of communication relay nodes, and the hierarchical artificial bee colony module is used to establish a hierarchical artificial bee colony system corresponding to the optimized layout model of communication relay nodes, and use the iterative method for solution, and perform feasibility processing on infeasible solutions; The layout system executes the optimized layout method for communication relay nodes of the cooperative action system as described in any one of claims 1-8.

Images