MPR set selection method based on binary penguin group optimization algorithm

Abstract

The invention discloses an MPR set selection method based on a binary penguin group optimization algorithm, and belongs to the technical field of computer network routing protocols. According to the method, the problems of local optimum, single selection standard and the like in MPR set selection in an OLSR protocol are solved, and the method comprises the steps that coordinate information and speed information are added to a HELLO message by an unmanned aerial vehicle node, and the link life is obtained through calculation; a fitness function is obtained through a binary emperor penguin group optimization algorithm according to four parameters of node connectivity, link life, node willingness and the number of MPR nodes, and then an MPR set with the maximum comprehensive quality is selected.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle (UAV) ad-hoc routing protocols, and particularly relates to a method for selecting a multipoint relay (MPR) set based on a binary emperor penguin colony optimization algorithm. Background Art

[0002] With the continuous progress of self-organizing network technology and the development of multi-UAV cooperative combat technology, UAV ad-hoc networks have become one of the key research areas. Since UAV ad-hoc networks do not rely on ground infrastructure and have strong anti-interference capabilities, self-organizing characteristics, and good scalability, they have been widely used in scenarios such as emergency rescue and military operations.

[0003] Compared with traditional mobile ad-hoc networks, UAV ad-hoc networks exhibit more significant advantages. First, their network coverage is wider. Second, as network nodes, UAVs have higher mobility and adaptability, and can cross complex terrains and avoid obstacles. In addition, UAVs can carry a variety of payload devices, including advanced devices such as sensors, high-definition cameras, and communication modules. These characteristics give UAV ad-hoc networks unique advantages in complex and changing working environments. However, although UAV ad-hoc networks show great potential in practice, their core technical difficulties are still similar to those of traditional mobile ad-hoc networks, and the most critical one is the design and optimization of efficient routing protocols. Among them, the Optimized Link State Routing (OLSR) protocol has become one of the most representative and widely used routing protocols in self-organizing network environments due to its excellent performance and wide applicability.

[0004] The OLSR protocol is a proactive routing protocol. In the MPR selection mechanism of the traditional OLSR protocol, in addition to ensuring the inclusion of necessary nodes (i.e., the case where some two-hop nodes can only reach through specific relay nodes), when constructing the MPR set, the number of two-hop nodes directly connected to candidate nodes (i.e., node connectivity) is usually used as the only selection criterion, and one-hop nodes with higher connectivity are preferentially selected. Although this greedy strategy-based selection mechanism can optimize performance within a region, in the UAV ad-hoc network environment, the movement speed of nodes and the topology change rate increase significantly, making candidate nodes with high connection density likely to quickly leave the current communication neighborhood, resulting in lag in the adjacency relationship table information in the local topology awareness system. This lag phenomenon may cause routing path failures within the adjacent topology control message update period, leading to data forwarding failures or a significant increase in transmission delay, exposing the limitations of the MPR selection method that only relies on connectivity as the selection criterion. Therefore, a more practical MPR selection method for UAV ad-hoc networks is needed to select an MPR set with the largest comprehensive quality. Summary of the Invention

[0005] In view of this, the present invention provides an MPR set selection method based on a binary emperor penguin colony optimization algorithm. In the present invention, in order to improve the performance of the routing protocol in the UAV ad hoc network environment, the present invention obtains a fitness function by modifying the HELLO message format to add information such as the speed and position of the UAV, and then uses a swarm intelligence algorithm to screen the MPR set for the one-hop neighbor nodes in the UAV ad hoc network.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] An MPR set selection method based on a binary emperor penguin colony optimization algorithm, characterized in that: the method includes the following steps:

[0008] Step 1) Use the Beidou satellite navigation system (BDS) installed on the UAV to obtain the position information and speed information of the current UAV F i ;

[0009] Step 2) Calculate the link lifetime between UAVs according to the obtained position information and speed information of UAV F i ;

[0010] Step 3) Complete the preliminary screening of MPR in the standard OLSR protocol, and add the selected one-hop nodes to the MPR set S MPR ;

[0011] Step 4) For the one-hop nodes not selected, use the binary emperor penguin colony optimization algorithm to initialize the population. Initially define N emperor penguin individuals, each individual corresponding to an MPR set, and each emperor penguin is represented by an array, that is, each dimension position of each emperor penguin is randomly assigned a state of 1 or 0;

[0012] Step 5) Calculate the fitness function Consider four factors, including neighbor willingness, link lifetime, neighbor coverage, and the number of MPRs;

[0013] Step 6) Update the positions of the emperor penguins, and then perform algorithm iteration. After reaching the maximum number of times, the emperor penguin individual with the highest fitness function value is regarded as the final solution to obtain S MPR * , and finally add it to the selected MPR set to finally obtain the complete S MPR .

[0014] Further, in step 1, UAV F i obtains accurate position information at time t + 1 according to the Beidou satellite navigation system (BDS) carried by itself and the position information of UAV F i at time t Next is the speed information of UAV F i , that is, speed and direction of motion where V i t (V min ≤V i t ≤V max ) represents the magnitude of the UAV's velocity vector, represents the angle between the projection of the UAV velocity vector on the XY plane and the X-axis, represents the angle between the UAV velocity vector and the Z-axis.

[0015] Furthermore, in step 2, the link lifetime is calculated. Here, the link lifetime is represented by TL n , representing the connection time maintained between two nodes within the communication range. The calculation formula is

[0016]

[0017] In the formula

[0018]

[0019] Furthermore, in step 3, the initial selection of the MPR set is completed, mainly referring to screening out the one-hop nodes that are always willing to be relay nodes and the unique reachable one-hop nodes corresponding to the two-hop nodes; first, set the one-hop neighbor set of the UAV node F i as Secondly, the two-hop neighbor set of the node is The symbol S MPR = {mpr1, mpr2,..., mpr j} is the set of selected MPR nodes; the screening steps for standard MPR nodes, the MPR algorithm process in RFC3626 is as follows:

[0020] Step 3.1 Initialize the set S MPR as an empty set;

[0021] Step 3.2 Add the nodes with the willingness value of WILL_ALWAYS in to it; The willingness is a state parameter designed in the standard OLSR protocol (where WILL_NEVER means the node is completely unwilling to be a relay node; WILL_ALWAYS: means the node is always willing to be a relay node), used to represent the willingness degree of relevant nodes to transmit and send messages to other nodes;

[0022] Step 3.3 Calculate the connection degree D(y) of all one-hop nodes, where the symbol D(y) represents the number of two-hop symmetric nodes covered by the initial one-hop nodes;

[0023] Step 3.4 Add the uniquely reachable one-hop nodes corresponding to the two-hop nodes to S MPR ; Delete the two-hop nodes covered by the selected MPR nodes from .

[0024] Furthermore, in Step 4, the binary emperor penguin colony optimization algorithm is used; the specific content of the binary emperor penguin colony optimization algorithm is described here;

[0025] Step 4.1 Emperor penguins are social animals living in Antarctica; during the Antarctic winter, the temperature drops to minus forty degrees. To cope with such harsh weather, they gather together to resist the cold; during this period, each emperor penguin individual will move towards the direction with higher temperature in the group and find the optimal position to reach the highest temperature; and the emperor penguin individual will adjust according to the distance between itself and the optimal position, continuously updating the optimal position, so as to achieve the optimal movement and position update of individuals within the group; the temperature T1 distribution during the iteration process:

[0026]

[0027] In the formula, t max is the maximum number of iterations, k is the current number of iterations, and the temperature T expression is:

[0028]

[0029] Step 4.2 After determining the temperature range, calculate the distance between each individual in the population and the optimal individual:

[0030]

[0031] In the formula and represent the vector parameters used for the volume collision of emperor penguins; represents a random number on [0,1]; represents the optimal individual position in the kth iteration; represents the current position of the individual; S(·) represents the movement trend of the emperor penguin towards the optimal individual, defining the main social status of the emperor penguin to distinguish the optimal individual from the ordinary individual:

[0032]

[0033] In the formula, f and v represent important control parameters for better exploration and development, and their value ranges are [2,3] and [1.5,2] respectively; P grid(Accuracy) represents the absolute value between the optimal individual and the positions of other individuals; M represents the movement step size parameter, which is used for the spacing between individuals to avoid collisions, and generally takes a value of 2;

[0034] Step 4.3 Then reposition the movement of the emperor penguins. The iterative formula for the position of the emperor penguin individuals:

[0035]

[0036] Step 4.4 The original Emperor Penguin Optimization (EPO) algorithm aims to solve continuous optimization problems. Due to discrete optimization problems in the discrete search space, its performance has limitations. The binary version of EPO (Binary EPO, BEPO) is introduced; in the BEPO algorithm, the position update of the emperor penguins is constrained to produce only discrete solutions of 0 and 1, and this process is achieved by introducing a sigmoid transfer function; the mathematical expression of the sigmoid transfer function is as follows:

[0037]

[0038] In the formula represents the control parameter of the nth emperor penguin at the i-th dimension in the k-th iteration; represents the position of the nth emperor penguin at the i-th dimension in the (k + 1)-th iteration;

[0039] Step 4.5 Initialize the population: The problem of whether to select a certain hop node as an MPR node is described by a binary value; the case of being selected as an MPR node is represented by the value 1, while the case of not being selected is represented by the value 0; initially, N emperor penguin individuals are defined, each individual corresponds to an MPR set, and each emperor penguin is represented by an array, that is, each dimension position of each emperor penguin is randomly assigned a state of 1 or 0; the number of elements in this array matches the number of one-hop neighbors after filtering out the unique reachable one-hop nodes corresponding to the two-hop nodes in Step 4; each initialized emperor penguin must be able to cover all strictly symmetric two-hop neighbor nodes.

[0040] Furthermore, in Step 5, calculate the fitness function Consider four factors, including neighbor willingness, link lifetime, neighbor coverage, and the number of MPRs; here, the calculation is for one-hop nodes without adding S MPR of the fitness f i n , which includes its neighbor willingness, link lifetime, neighbor coverage, and the number of MPRs;

[0041] Four variable factors of the fitness function:

[0042] The first variable is the neighbor willingness (willingness(NGW n )): It is extracted from the neighbor table in each drone F i 1-hop (n), where NGW max = 7 and NGW min = 1;

[0043] The second variable is the link lifetime (TL n ), which is calculated in step 2 and can be obtained in the HELLO message;

[0044] The third variable is the coverage of neighbors (D(n)): The number of two-hop nodes that a one-hop node can connect to;

[0045] The number of MPRs (NUM): The number of MPR nodes selected in the emperor penguin algorithm;

[0046] Fitness expression:

[0047]

[0048] In the formula, NGW max , NGW min , TL max , TL min , D(max), D(min) represent the maximum and minimum values of the corresponding functions;

[0049] Fitness function expression:

[0050]

[0051] In the formula, f i mpr,q represents the fitness of the MPR nodes selected by the emperor penguin individual in the emperor penguin algorithm; n * represents the number of remaining one-hop nodes after screening in step 3; ω1, ω2 are the weight values in the fitness function, and both are taken as 0.5 here.

[0052] Furthermore, in step 6, update the position and calculate the final solution set; According to the position equation in the binary emperor penguin optimization algorithm

[0053]

[0054] Update;

[0055] Output result: Here, the algorithm iterates 100 times. After reaching the maximum number of times, the emperor penguin individual with the highest fitness function value is regarded as the final solution, obtaining S MPR * , and finally add it to the selected MPR set, and finally obtain the complete SMPR 。

[0056] The effective effect of the present invention lies in: aiming at the problem that the state information of neighbor nodes fails due to the single-index selection of the traditional MPR selection algorithm, a method for selecting MPR based on the binary emperor penguin optimization algorithm is proposed. This method uses a swarm intelligence algorithm to screen the MPR set for the one-hop neighbor nodes in the UAV ad hoc network. This method no longer relies solely on a single connectivity metric, but comprehensively considers various dynamic characteristics of nodes (such as moving speed, topological stability, etc.), and selects the MPR set with the optimal comprehensive quality through an intelligent optimization algorithm. This method can better adapt to the highly dynamic environment of the UAV ad hoc network, improve the real-time performance of topological perception and the robustness of routing, thereby effectively reducing the data transmission failure rate and optimizing the network performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] To make the objectives, technical solutions and beneficial effects of the present invention clearer, the following drawings are provided for illustration:

[0058] Figure 1 It is the flowchart of MPR selection based on the binary emperor penguin optimization algorithm according to the embodiment of the present invention;

[0059] Figure 2 It is the schematic diagram of traditional MPR selection according to the embodiment of the present invention;

[0060] Figure 3 It is the diagram of the HELLO message format according to the embodiment of the present invention;

[0061] Figure 4 It is the schematic diagram of the initialization population of the swarm optimization algorithm according to the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0062] Next, the preferred embodiments of the present invention will be described in detail in conjunction with the drawings. The described embodiments are a part of the embodiments of the present invention, rather than all of the embodiments.

[0063] Embodiment

[0064] Through the traditional MPR screening mechanism, the initialization must select the one-hop nodes to be included in the MPR set (i.e., the case where some two-hop nodes must be reachable through specific relay nodes); then, through the binary emperor penguin colony optimization algorithm and the fitness function obtained according to the four parameters of node connectivity, link lifetime, node willingness, and the number of MPR nodes, determine the boundary range enclosed during the emperor penguin clustering process; secondly, define the temperature of different gradients in the clustering process according to the fact that the temperature of the warming environment is different at different gradient positions during the clustering process; finally, calculate the distance between a random emperor penguin and the emperor penguin at the center of the cluster. Since the positions of the emperor penguins change continuously during the clustering process, the function continuously updates the position of the emperor penguin at the center of the cluster, and finally can quickly converge to the MPR set with the fewest and most stable nodes that meet the MPR set conditions. Furthermore, select the MPR set with the largest comprehensive quality.

[0065] The above MPR selection process can be passed through Figure 1 Briefly represented.

[0066] Among them, in step 3, the preliminary screening of MPR in the standard OLSR protocol is completed, and the selected one-hop nodes are added to the MPR set S MPR In, it can be clearly shown in Figure 2 That is, select the one-hop nodes in the case where some two-hop nodes must be reachable through specific relay nodes.

[0067] Among them, in step 4, the population of the optimization algorithm is initialized, as shown in Figure 3 . Here, we describe the problem of whether to select a certain one-hop node as an MPR node through a binary value. Specifically, the case of being selected as an MPR node is represented by the value 1, while the unselected case is represented by the value 0. We initially define N emperor penguin individuals, each corresponding to an MPR set, and each emperor penguin is represented by an array, that is, each dimension position of each emperor penguin is randomly assigned a state of 1 or 0. It should be noted here that the number of elements in this array matches the number of one-hop neighbors after step 4 screens out the uniquely reachable one-hop nodes corresponding to the two-hop nodes; each initialized emperor penguin must be able to cover all strictly symmetric two-hop neighbor nodes.

[0068] Among them, Figure 4 Represents the modified HELLO message format, in which the position information, speed information, and link lifetime of the drone are added, respectively used to calculate the fitness function.

[0069] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made in form and details without departing from the scope defined by the claims of the present invention.

Claims

1. An MPR set selection method based on a binary emperor penguin colony optimization algorithm, characterized in that: The following steps are involved: Step 1) Use the Beidou Satellite Navigation System (BDS) installed on the drone to obtain the current drone F i Position information and speed information; Step 2) According to the obtained drone F i The location information and speed information of the drones are used to calculate the link lifetime between the drones. Step 3) Complete the preliminary screening of MPR in the standard OLSR protocol and add the selected one-hop node to the MPR set S MPR middle; Step 4) For the one-hop nodes that are not selected, use the binary emperor penguin population optimization algorithm to initialize the population, initially define N emperor penguin individuals, each individual corresponds to an MPR set, and each emperor penguin is represented by an array, that is, each dimensional position of each emperor penguin is randomly assigned a 1 or 0 state; Step 5) Calculate the fitness function Four factors are considered, including neighbor willingness, link lifetime, neighbor coverage, and the number of MPRs; Step 6) Update the position of the emperor penguin, and then iterate the algorithm. After reaching the maximum number of times, the emperor penguin individual with the highest fitness function value is regarded as the final solution, and S is obtained. MPR * , and finally add it to the selected MPR set to get the complete S MPR .

2. The MPR set selection method based on the binary emperor penguin colony optimization algorithm according to claim 1 is characterized in that: The specific process in step 1 includes: According to step 1, drone F i According to the Beidou satellite navigation system (BDS) carried by the drone, obtain the F i Position information at time t and drone F i Velocity information, i.e. speed and direction of movement in Represents the magnitude of the UAV's velocity vector, It represents the angle between the projection of the UAV velocity vector on the XY plane and the X axis. Indicates the angle between the drone's velocity vector and the Z axis.

3. The MPR set selection method based on the binary emperor penguin colony optimization algorithm according to claim 1 is characterized in that: The specific process in step 2 includes: According to step 2, calculate the link lifetime; here the link lifetime is TL n It represents the connection time between two nodes within the communication range; the calculation formula is In the formula 4. The MPR set selection method based on the binary emperor penguin colony optimization algorithm according to claim 1, characterized in that: The specific process in step 3 includes: According to step 3, the initial selection of the MPR set is completed, which mainly refers to screening out the one-hop node that is always willing to serve as a relay node and the only reachable one-hop node corresponding to the two-hop node; first, set the drone node F i The set of one-hop neighbors of The set of two-hop neighbors of the second node is Symbol S MPR ={mpr1,mpr2,...,mpr j } is the set of selected MPR nodes; the standard MPR node screening steps, the MPR algorithm process in RFC3626 is as follows: Step 3.1 Initialize the set S MPR is an empty set; Step 3.2 The nodes with willingness value of WILL_ALWAYS in the OLSR protocol are added to the OLSR protocol. Willingness is a state parameter designed in the standard OLSR protocol (WILL_NEVER means that the node is completely unwilling to serve as a relay node; WILL_ALWAYS means that the node is always willing to serve as a relay node), which is used to indicate the willingness of the relevant node to transmit and send messages to other nodes. Step 3.3 calculates the connectivity D(y) of all one-hop nodes, where D(y) represents the number of two-hop symmetric nodes covered by the initial one-hop node; Step 3.4 Add the only reachable one-hop node corresponding to the two-hop node to S MPR in; from Delete the two-hop nodes covered by the selected MPR node.

5. The MPR set selection method based on the binary emperor penguin colony optimization algorithm according to claim 1, characterized in that: The specific process in step 4 includes: According to step 4, the binary emperor penguin population optimization algorithm is used; the specific content of the binary emperor penguin population optimization algorithm is explained; Step 4.1 In order to cope with this bad weather, emperor penguins will gather together to resist the cold; during bad weather, each emperor penguin will move towards the direction of higher temperature in the group and find the optimal position to reach the highest temperature; and the emperor penguin individual will adjust according to the distance between itself and the optimal position, and continuously update the optimal position, so as to achieve the optimized movement and position update of individuals in the group; temperature T1 distribution during the iteration process: Where t max is the maximum number of iterations, k is the current number of iterations, and the temperature T expression is: After determining the temperature range in step 4.2, calculate the distance between each individual in the population and the optimal individual: In the formula and Represents the vector parameters used for the emperor penguin's volume collision; Represents a random number on [0,1]; represents the optimal individual position in the kth iteration; represents the current position of the individual; S(·) represents the movement trend of the emperor penguin towards the optimal individual, defines the main social status of the emperor penguin, and is used to distinguish the optimal individual from the ordinary individual: Where f and v represent important control parameters for better exploration and development, and their value ranges are [2,3] and [1.5,2] respectively; P grid (Accuracy) represents the absolute value between the positions of the optimal individual and other individuals; M represents the moving step parameter, which is used to avoid collisions between individuals, and its value is generally 2; Step 4.3 Then reposition the movement of the emperor penguin, the iterative formula for the individual position of the emperor penguin is: Step 4.4 The original Emperor Penguin Optimization Algorithm (EPO) is designed to solve continuous optimization problems. Due to the discrete optimization problem in the discrete search space, its performance is limited. The binary version of EPO (Binary EPO, BEPO) is introduced. In the BEPO algorithm, the position update of the emperor penguin is constrained to only produce discrete solutions of 0 and 1. This process is achieved by introducing the S-type transfer function. The mathematical expression of the S-type transfer function is as follows: In the formula represents the control parameters of the nth emperor penguin at iteration k in the i-th dimension; represents the position of the nth emperor penguin at iteration k+1 of the i-th dimension; Step 4.5 initializes the population: the question of whether a certain hop node is selected as an MPR node is described by a binary value; the case of being selected as an MPR node is represented by a value of 1, and the case of not being selected is represented by a value of 0; initially define N emperor penguin individuals, each individual corresponds to an MPR set, and each emperor penguin is represented by an array, that is, each dimensional position of each emperor penguin is randomly assigned a 1 or 0 state; the number of elements in the array matches the number of one-hop neighbors after the only reachable one-hop node corresponding to the two-hop node is screened out in step 4; each initialized emperor penguin must be able to cover all strictly symmetric two-hop neighbor nodes.

6. The MPR set selection method based on the binary emperor penguin colony optimization algorithm according to claim 1, characterized in that: The specific process in step 5 includes: Calculate the fitness function as described in step 5 Four factors are considered, including neighbor willingness, link lifetime, neighbor coverage, and the number of MPRs; this calculation does not include S MPR One-hop node Fitness These include the willingness of its neighbors, link lifetime, neighbor coverage, and the number of MPRs; The fitness function has four variable factors: The first variable is neighbor willingness (NGW n ): Through each drone The neighbor table in NGW is extracted, where max =7 and NGW min =1; The second variable is the link lifetime (TL n ), calculated in step 2, can be obtained in the HELLO message; The third variable is the neighbor coverage (D(n)): the number of two-hop nodes to which a one-hop node can connect; MPR number (NUM): the number of MPR nodes selected in the Emperor Penguin algorithm; Fitness expression: Where NGW max ,NGW min ,TL max ,TL min ,D(max),D(min) represent the maximum and minimum values ​​of the corresponding functions; Fitness function expression: In the formula Indicates the fitness of the MPR node selected by the emperor penguin individual in the emperor penguin algorithm; n * It represents the number of remaining one-hop nodes after screening in step 3; ω1, ω2 are the weight values ​​in the fitness function, both of which are 0.

5.

7. The MPR set selection method based on the binary emperor penguin colony optimization algorithm according to claim 1, characterized in that: The specific process in step 6 includes: According to step 6, update the position and calculate the final solution set, according to the position equation in the binary emperor penguin optimization algorithm To update, Output result: The algorithm iterates 100 times. After reaching the maximum number of times, the emperor penguin individual with the highest fitness function value is regarded as the final solution, and S is obtained. MPR * , and finally add it to the selected MPR set to get the complete S MPR .

Images