A mobile ad hoc network clustering networking strategy based on GWO in a star-ground fusion scenario

Abstract

The present application relates to the field of mobile ad hoc network communication technology in the star-ground fusion, and particularly relates to a mobile ad hoc network clustering networking strategy based on GWO in the star-ground fusion scene, the networking strategy comprising: an information collection stage, an information transmission stage, a cluster establishment stage and a cluster maintenance stage; a dual-mode communication node transmits network information to a command center through a low-orbit satellite to perform a grey wolf optimization algorithm with an optimized fitness function from a global perspective to cluster; the present application solves the slow convergence speed and large resource consumption caused by distributed calculation in the traditional mobile ad hoc network clustering algorithm, simultaneously optimizes the fitness function of the grey wolf optimization algorithm, avoids the algorithm from falling into local optimum, and improves the efficiency and rationality of the algorithm. The present application can reduce the energy consumption of the entire network, thereby improving the life of the entire network.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of mobile ad hoc network communication in the star-ground fusion, and particularly relates to a mobile ad hoc network clustering networking strategy based on GWO in a star-ground fusion scenario. BACKGROUND

[0002] Mobile ad hoc network is widely used in disaster relief, military communication and distributed cooperative communication scenarios due to its self-organization, temporality, multi-hop transmission and independence on infrastructure. At the same time, low-orbit satellite network has the advantages of wide coverage, large communication capacity and flexible network construction, and can effectively expand and supplement the deficiency of ground ad hoc network: in a dynamically changing environment (such as battlefield or disaster area), the wide coverage of satellite network can support the flexibility and mobility of mobile nodes, and adapt to the rapid change of network topology; the dual-mode communication terminal integrating low-orbit satellite and ad hoc network communication can avoid the network limitation caused by the need of special satellite gateway in traditional star-ground communication, and improve the flexibility and scalability of the whole network.

[0003] In the fusion scenario of satellite network and ground mobile ad hoc network, the existing research often uses network technologies such as SDN and NFV to decouple the heterogeneous network, so as to design the mobile ad hoc network and the satellite network respectively. For the ground ad hoc network, in order to obtain global network information, nodes will perform a large amount of probe information flooding, which not only increases the computing overhead of nodes, but also causes slow network convergence. Therefore, researchers have proposed a hierarchical network structure, that is, nodes are divided into groups called clusters, and one node in the cluster is elected as the cluster head. Current clustering algorithms mainly include: based on minimum ID algorithm, maximum degree algorithm, etc., considering multiple factors and performing on-demand weighting, and finally clustering according to weight. However, the factors considered in the algorithm weight vary with the environment, and when the network situation changes, the weight calculation cost will also increase accordingly. At the same time, the mobile ad hoc network nodes are distributed, and each node decides to establish clustering, enter clustering and elect cluster head according to the information interacted with neighbor nodes. When the number of ad hoc network nodes increases, the ad hoc network nodes will consume a large amount of resources for clustering operation, resulting in slow network convergence and increased energy consumption. With the development of intelligent algorithms, bionic clustering algorithms based on biological colony idea have attracted widespread attention. They not only guarantee the advantages of traditional clustering algorithms, but also have the characteristics of strong self-adaptation and self-organization. Algorithms such as ant colony clustering algorithm, grey wolf optimization algorithm (GWO) and firefly algorithm have been applied to clustering algorithms. However, these algorithms are mostly used in wireless sensor networks with poor mobility, and may easily fall into local optimum. At the same time, in the emergency scenario where energy may not be supplied, the ad hoc network nodes running such algorithms may consume a large amount of resources, thereby reducing the lifetime of the whole network. SUMMARY

[0004] To solve the problems in the background art, the application provides a mobile ad hoc network clustering networking strategy based on GWO in a star-ground fusion scenario, which includes an information collection stage, an information transmission stage, a cluster establishment stage and a cluster maintenance stage.

[0005] In the information collection stage, the mobile ad hoc network node broadcasts a HELLO information packet within its communication range; the mobile ad hoc network node receives the HELLO message of other nodes and maintains a node information table, wherein the node information table includes the number of the node, the remaining energy of the node, the position of the node, the moving speed of the node, the communication strength of the node with the satellite node, and the number of the next hop node of the node.

[0006] In the information transmission stage, the dual-mode communication node in the mobile ad hoc network selects the optimal satellite node to send its node information table to the command center according to the uplink load of the dual-mode communication node with the satellite node, the signal strength of the dual-mode communication node with the satellite node, the visibility of the dual-mode communication node with the satellite node, and the communication delay of the dual-mode communication node with the satellite node; the command center analyzes and integrates the node information of the mobile ad hoc network;

[0007] In the cluster establishment stage, the command center clusters the nodes in the mobile ad hoc network by using an optimized grey wolf optimization algorithm according to the node information of the mobile ad hoc network, and sends the clustering result to the dual-mode communication node through the satellite node, which broadcasts the clustering result in the mobile ad hoc network;

[0008] In the cluster maintenance stage, the node information table of each mobile ad hoc network node maintains the node information of the cluster to which it belongs, and simultaneously performs the leaving, joining and cluster head updating operations of the cluster members according to the maintained information, wherein the cluster head node simultaneously maintains the information of the nodes within the cluster and the nodes outside the cluster.

[0009] Preferably, the selection of the optimal satellite node by the dual-mode communication node includes:

[0010] The dual-mode communication node N i selects the optimal satellite node according to the uplink load of the dual-mode communication node N j with the satellite node S j , the signal strength of the dual-mode communication node N j with the satellite node S j , the visibility of the dual-mode communication node N j with the satellite node S j , and the communication delay of the dual-mode communication node N ij with the satellite node S ij .

[0011] The score of the satellite node S ij includes: ij- ω4·D ij

[0012] wherein SV ij denotes the score of the satellite node S i for the dual-mode communication node N j ; ω1, ω2, ω3 and ω4 denote weight parameters of the respective factors; U ij denotes the uplink load of the satellite node S i for the dual-mode communication node N j ; RSS ij denotes the signal strength of the satellite node S i for the dual-mode communication node N j ; V ij denotes the visibility of the satellite node S i for the dual-mode communication node N j ; D ij denotes the communication delay of the satellite node S i for the dual-mode communication node N j .

[0013] Preferably, the visibility V i of the satellite node S j for the dual-mode communication node N ij comprises:

[0014]

[0015] wherein θ ij denotes the zenith angle of the satellite node S i for the dual-mode communication node N j , and is given by:

[0016]

[0017] wherein and denote the longitude and latitude of the dual-mode communication node N i and the satellite node S j , respectively, and θ threshold denotes a set threshold value.

[0018] Preferably, the clustering of the nodes in the mobile ad-hoc network using the optimized grey wolf optimization algorithm comprises:

[0019] Step 101: initializing a population size, a population dimension and a maximum number of iterations, and randomly initializing a position of each grey wolf individual, wherein the position of each grey wolf individual corresponds to a cluster head node selection scheme;

[0020] Step 102: constructing an optimized fitness function, the optimized fitness function F being as follows:

[0021] F = ψ1·f1 + ψ2·f2 + ψ3·f3 + ψ4·f4 - ψ5·f5

[0022] wherein, f1 represents the relative speed of the cluster head node and its neighbor nodes; f2 represents the residual energy fraction of the cluster head node; f3 represents the shortest distance between the cluster head nodes; f4 and f5 represent the reward factor and the penalty factor, for ensuring that there is a dual-mode communication node in each cluster and encouraging the dual-mode communication node to become a cluster head; ψ1, ψ2, ψ3, ψ4 and ψ5 are weights thereof, and ψ1 + ψ2 + ψ3 + ψ4 + ψ5 = 1;

[0023] Step 103: according to the geographical position distance, the nearest cluster head node of other nodes is selected to join, based on the constructed fitness function, the fitness value of each gray wolf is calculated; according to the fitness value of each gray wolf, the alpha wolf, beta wolf and delta wolf are determined;

[0024] Step 104: updating the position of the gray wolf according to the position of the alpha wolf, beta wolf and delta wolf;

[0025] Step 105: calculating the fitness value of the gray wolf after position updating; according to the fitness value of the gray wolf after updating, the alpha wolf, beta wolf and delta wolf are re-determined;

[0026] Step 106: if the current round reaches the maximum iteration number, the iteration is ended, the optimal individual is output, and the final clustering scheme with the maximum fitness value is obtained;

[0027] Step 107: if the current round does not reach the maximum iteration number, go to step 104.

[0028] Preferably, the updating of the position of the gray wolf according to the position of the alpha wolf, beta wolf and delta wolf comprises:

[0029]

[0030] wherein, F1, F2 and F3 are the optimized weight factors, F α , F β and F δ respectively represent the fitness of the alpha wolf, beta wolf and delta wolf in the gray wolf algorithm; represent the position of the gray wolf individual after updating.

[0031] Preferably, the relative speed of the cluster head node and its neighbor nodes comprises:

[0032]

[0033] wherein, NUM i represents the number of neighbor nodes of the cluster head node N i ; V j represents the moving speed of the cluster head node; V max represents the moving speed of the cluster head node the moving speed of the neighbor node of the node.

[0034] Preferably, the residual energy fraction of the node comprises:

[0035]

[0036] wherein E max represents the maximum energy of the cluster head node; represents the residual energy of the cluster head node.

[0037] Preferably, the distance between the cluster head nodes comprises:

[0038]

[0039] wherein, and represent the longitude and latitude coordinates of the cluster head node CHi′ , and represent the longitude and latitude coordinates of the cluster head node CHi , arcsin represents the inverse sine function; R earth represents the radius of the earth.

[0040] Preferably, the reward factor and the penalty factor comprise:

[0041] f4 = 1 + p1 · t

[0042] f5 = p2 (1 - t)

[0043]

[0044] wherein p1 and p2 represent a constant greater than 0.

[0045] Preferably, the leaving, joining and cluster head updating operations of the cluster members comprise:

[0046] After clustering is completed, the cluster head CH i inserts the cluster head identifier in the HELLO information, and the ordinary node N i obtains the position and moving speed of the cluster head according to the HELLO information, and learns the distance and relative speed

[0047] Leaving of the cluster member: when there exists a cluster node satisfying , v threshold is the relative speed threshold, the node is removed from the cluster;

[0048] Joining of the cluster member: when there exists a node outside the cluster satisfying , the node outside the cluster is added to the cluster;

[0049] Cluster head update: when the residual energy of the cluster head node is lower than the threshold value, the cluster head broadcasts cluster head update information, and the nodes in the cluster calculate the score R according to the residual energy of the nodes and the relative moving speed of the nodes i , and the specific calculation formula is as follows:

[0050]

[0051] Wherein, R i represents the score of the node N i in the cluster, σ1 and σ2 represent the weight of each factor, NOC i represents the number of nodes in the cluster; V i and V j represent the moving speed of N i and N j of the nodes in the cluster; E max is the maximum energy of the node; E i represents the residual energy of the node N i in the cluster;

[0052] After each node in the cluster calculates the score of itself, the node broadcasts the HELLO information with the score of itself;

[0053] When each node in the cluster receives the HELLO information with a higher score than itself, the node does not broadcast the HELLO information with the score of itself;

[0054] When a node broadcasts the HELLO information with the score, and does not receive the HELLO information with the score from the remaining nodes within a preset time period, the node becomes a new cluster head.

[0055] The present application has at least the following beneficial effects

[0056] In the star-ground fusion scene, network information is collected from the dual-mode communication terminal device by using the wide coverage of the satellite network, and then transmitted to the command center with high computing power, and the command center uses GWO for clustering networking from the global network perspective, so that the efficiency of the mobile ad hoc network clustering operation is improved, the clustering calculation overhead and energy consumption of the ad hoc network node are reduced, and the life of the entire network is improved. In addition, the fitness function of the grey wolf optimization algorithm is adaptively optimized by the optimized grey wolf optimization algorithm, so that the algorithm is prevented from falling into local optimum, and the efficiency and rationality of the algorithm are improved. BRIEF DESCRIPTION OF DRAWINGS

[0057] Figure 1 is a schematic diagram of a star-ground fusion scene in an embodiment of the present application. DETAILED DESCRIPTION

[0058] Following, the embodiments of the present application will be described in detail by specific examples, and other advantages and effects of the present application can be easily understood by those skilled in the art from the description. The present application can also be implemented or applied by other different embodiments, and various modifications or changes can be made to the details in the description based on different views and applications without departing from the spirit of the present application. It should be noted that the drawings provided in the following examples only illustrate the basic concept of the present application in a schematic manner, and the following examples and features in the examples can be combined with each other without conflict.

[0059] Referring to Figure 1 The present application provides a mobile ad hoc network clustering networking strategy based on GWO in a satellite-ground fusion scenario, comprising:

[0060] The networking strategy comprises an information collection stage, an information transmission stage, a cluster establishment stage and a cluster maintenance stage.

[0061] In the information collection stage, the mobile ad hoc network node broadcasts a HELLO information packet within its communication range; the mobile ad hoc network node receives the HELLO message of other nodes and maintains a node information table, wherein the node information table comprises the number of the node, the remaining energy of the node, the position of the node, the moving speed of the node, the communication strength of the node with the satellite node, and the number of the next hop node of the node.

[0062] In the information transmission stage, the dual-mode communication node in the mobile ad hoc network selects the optimal satellite node to send its node information table to the command center according to the uplink load of the dual-mode communication node with the satellite node, the signal strength of the dual-mode communication node with the satellite node, the visibility of the dual-mode communication node with the satellite node, and the communication delay of the dual-mode communication node with the satellite node, and the command center analyzes and integrates the node information of the mobile ad hoc network.

[0063] In the cluster establishment stage, the command center clusters the nodes in the mobile ad hoc network using the optimized grey wolf optimization algorithm according to the node information of the mobile ad hoc network, and sends the clustering result to the dual-mode communication node through the satellite node, and the dual-mode communication node broadcasts the clustering result in the mobile ad hoc network.

[0064] In the cluster maintenance stage, the node information table of each mobile ad hoc network node maintains the node information of the cluster to which it belongs, and simultaneously performs the leaving, joining and cluster head updating operations of the cluster members according to the maintained information, wherein the cluster head node simultaneously maintains the information of the nodes within the cluster and the nodes outside the cluster.

[0065] In the present application, the mobile ad hoc network clustering networking strategy comprises an information collection stage, an information transmission stage, a cluster establishment stage and a cluster maintenance stage.

[0066] In the information collection phase, the mobile ad hoc network nodes transmit their own node information by broadcasting HELLO information packets.

[0067] In the information transmission phase, the dual-mode communication nodes in the mobile ad hoc network transmit network node information to the command center through satellites.

[0068] In the cluster establishment phase, the command center performs gray wolf optimization algorithm for clustering calculation, and selects cluster head nodes through an optimized fitness function.

[0069] In the cluster maintenance phase, the node information table of each mobile ad hoc network node only needs to maintain the node information within the cluster, thereby saving energy consumption.

[0070] Referring to Figure 1 , the satellite-ground fusion scene used in the embodiment of the present application is deployed with N mobile ad hoc network nodes N i (i=1, 2,..., N), which can be divided into dual-mode communication nodes Nd i (i=1, 2,..., d) and ordinary ad hoc network nodes Ns i (i=1, 2,..., s) according to communication ability, and the communication range of each is r.

[0071] In the satellite-ground fusion scene, M low-orbit satellites S j (j=1, 2,..., M) are deployed, which have wide coverage and can establish communication connection with the dual-mode communication nodes in the mobile ad hoc network.

[0072] In order to ensure that the mobile ad hoc network is more comprehensive and accurate in the information collection phase, the dual-mode communication nodes are evenly distributed in the entire mobile ad hoc network, which ensures that each cluster has a dual-mode communication node as much as possible during subsequent clustering operations.

[0073] Considering the beneficial effect of low-orbit satellites on the effective expansion of the communication range of the mobile ad hoc network, Figure 1A flow chart for clustering mobile ad hoc networks based on GWO in a satellite-ground fusion scenario is presented. This invention fully leverages this fusion scenario by offloading clustering computation tasks to a command center with nearly unlimited energy through dual-mode communication nodes and low-orbit satellites. Mobile ad hoc network nodes only need to perform information collection, thus removing the computational burden from the mobile ad hoc network nodes and reducing energy consumption across the entire network. This, in turn, increases the lifespan of the entire network.

[0074] When the mobile ad hoc network is just established, all nodes are in an independent state. Each node first obtains its own node information, including the node number and the node's remaining energy E. i , the location of the node ( and φ i They are the ad hoc network nodes N i geographic latitude and longitude) and the node's movement speed Each dual-mode communication node N i Additional acquisition of its satellite node S j Communication Strength RSS ij , the present invention mainly considers free space path loss and atmospheric attenuation:

[0075] RSS ij =P t +G t +G r -FSPL-AAL

[0076] Among them, P t , G t and G r They are the transmit power of the transmitter on the satellite, the transmit antenna gain and the receive antenna gain of the dual-mode node, FSPL represents the free space path loss, and AAL represents the atmospheric attenuation loss.

[0077] FSPL is the signal strength attenuation caused by the increase in distance, specifically expressed as:

[0078] FSPL=20lg(d)+20lg(f)+20lg(4π / c)

[0079] Where d is the distance between the ground dual-mode communication node and the satellite, f is the frequency of the communication signal, and c is the speed of light, which is approximately 3×10 8 m / s, lg represents the logarithmic function with a base of natural numbers.

[0080] The distance between the ground dual-mode communication node and the satellite node can be expressed as:

[0081]

[0082] in, It is a satellite node Sj The longitude and latitude of ij Satellite node S j With the ground dual-mode communication node N i Zenith angle between earth represents the radius of the earth; h represents the satellite node S j orbital altitude.

[0083] AAL mainly considers rain attenuation rain and cloud vapor attenuation A cloud The impact on the signal is specifically expressed as:

[0084] ALL=A rain +A cloud

[0085] A rain =γ r ·d r

[0086] A cloud =γ c ·d c

[0087] Among them, γ r and γ c is the rain and cloud attenuation coefficient, d r and d c Indicates the distance the signal travels through rain and clouds.

[0088] After the mobile ad hoc network node obtains its own node information, each node periodically broadcasts the HELLO message, which includes the node's own number and the node's own remaining energy E i , the location of the node itself RSS of the communication strength between the node itself and the satellite i To avoid HELLO message flooding, its TTL is set to 1.

[0089] In ordinary ad hoc network nodes, the communication strength between them and satellite nodes is set to empty. If the dual-mode communication node and the satellite node cannot communicate due to terrain or other reasons, the communication strength is 0. This can distinguish the two and is also helpful for identifying whether the node is a dual-mode communication node.

[0090] The nodes in the mobile ad hoc network build and maintain a node information table by receiving HELLO messages from other nodes, which includes the node number, the node's remaining energy E i , the location of the node Node movement speed Communication strength RSS between the node and the satellite node ij, and the number of the next-hop node. The number of the next-hop node represents the relay node that sends information to the current entry node. If the number of the current entry node and the number of the next-hop node are consistent, it represents a one-hop neighbor node.

[0091] The mobile ad hoc node periodically broadcasts the node information table. By receiving the node information table of other nodes, the mobile ad hoc node can obtain the information of other nodes more than two hops away and update it to the node information table itself, and the number of the next-hop node is filled in the number of the node sending the node information table.

[0092] In order to avoid the flooding of such message packets, the TTL is set to 2, at which time the node can accurately know the correct previous-hop node of the three-hop node.

[0093] In the information transmission phase: after the information collection phase, the dual-mode communication node will periodically send its node information table to the command center through the satellite. When the dual-mode communication node selects a transmission satellite, it comprehensively considers the uplink load U ij , signal strength RSS ij , visibility V ij , and communication delay D ij . For the uplink load level U ij , it is specifically depicted as:

[0094]

[0095] Wherein, B used represents the current link bandwidth, and B total represents the total bandwidth capacity.

[0096] The higher the load level U ij , the more congested the link may be, so a satellite node with a low load level is usually selected.

[0097] For signal strength RSS ij , the collective depiction is the same as in the information collection phase. The stronger the signal strength, the better the communication quality, and the lower the bit error rate.

[0098] For visibility V ij , it can be judged by the zenith angle θ ij between the dual-mode communication node and the satellite:

[0099]

[0100] Wherein, the zenith angle θ ij is depicted the same as in the information collection phase. When θ i is less than a certain threshold θ threshold , it indicates that the satellite is within the visible range.

[0101]

[0102] wherein, and respectively represent the longitude and latitude of the dual-mode communication node N i and the satellite node S j , and θ threshold represents a set threshold value.

[0103] For the communication delay D ij , it is specifically characterized as:

[0104]

[0105] The dual-mode communication node will preferentially select a satellite with a shorter distance and lower delay, and c represents the speed of light.

[0106] Therefore, by comprehensively considering the above factors, a multi-factor model is used to weight and sum to obtain the comprehensive score of each satellite, which is collectively characterized as follows:

[0107] SV ij = ω1·(1-U ij )+ ω2·RSS ij + ω3·V ij - ω4·D ij

[0108] wherein ω1, ω2, ω3 and ω4 are the weights of the respective factors, and ω1+ ω2+ ω3+ ω4=1.

[0109] The ground dual-mode communication terminal selects a satellite with a high comprehensive score for transmission by judging one or more satellites SV ij within the communication range.

[0110] In the cluster establishment phase: after satellite relay transmission, the command center integrates the received node information table and restores the entire network topology, and then executes the grey wolf optimization algorithm for mobile ad hoc network cluster head election, cluster formation, and the specific steps are as follows:

[0111] The grey wolf optimization algorithm simulates the grey wolf hunting process and finds the optimal solution through the cooperative hunting of leader wolves and follower wolves. In cluster formation, each grey wolf represents a cluster head selection scheme, and the optimal grey wolf is the optimal cluster head allocation.

[0112] The introduction of the grey wolf optimization algorithm into the cluster head election of the mobile ad hoc network in the space-ground fusion scenario needs to fully consider the influence of mobility and dual-mode communication nodes, so the fitness function in the grey wolf optimization algorithm needs to be adaptively improved.

[0113] The fitness function design comprehensively considers the node mobility, node residual energy score, node position, and speciality of dual-mode nodes in the mobile ad hoc network, and is designed as follows:

[0114] F = ψ1·f1 + ψ2·f2 + ψ3·f3 + ψ4·f4 - ψ5·f5

[0115] Wherein, f1 represents the relative speed of the cluster head node and its neighbor nodes; f2 represents the residual energy fraction of the cluster head node; f3 represents the shortest distance between the cluster head nodes; f4 and f5 represent the reward factor and the penalty factor, for ensuring that there is a dual-mode communication node in each cluster and encouraging the dual-mode communication node to become a cluster head; ψ1, ψ2, ψ3, ψ4 and ψ5 are weights thereof, and ψ1 + ψ2 + ψ3 + ψ4 + ψ5 = 1; wherein, when initializing the position of each grey wolf individual, the mobile ad hoc network node can be first mapped to the position space, for example, all nodes in the mobile ad hoc network are sorted, and the sorting serial number thereof is used as the position, and then the position of each grey wolf individual is initialized, while the positions of the grey wolf individuals in all dimension spaces are required to be different.

[0116] For example, the clustering of the nodes in the mobile ad hoc network by using the optimized grey wolf optimization algorithm comprises:

[0117] When the command center collects the ad hoc network node information, the number of cluster heads is determined according to the following steps:

[0118] The communication range coverage area of the ad hoc network node is A = πr 2 , and the area A total of the entire ad hoc network is approximately calculated according to the collected position information, so that the number of cluster heads is preliminarily estimated as This value can be dynamically adjusted according to the specific situation later.

[0119] The steps of executing the grey wolf optimization algorithm are as follows:

[0120] Step 101: initializing the population size, population dimension and maximum iteration number T, and initializing and wherein is a convergence coefficient linearly decreasing in the interval [2, 0], and are coefficient vectors, and the calculation formulae are as follows:

[0121]

[0122] Wherein, r1 and r2 are random numbers between 0 and 1.

[0123] Meanwhile, the position of each grey wolf individual is randomly initialized, wherein the position of each grey wolf individual corresponds to a cluster head selection scheme, which is represented as a k pre ×1 matrix, each parameter corresponds to the ad hoc network node id, and the id is arranged from small to large.

[0124] Step 102: Then an optimized fitness function is constructed, and the optimized fitness function F is as follows:

[0125] F = ψ1·f1 + ψ2·f2 + ψ3·f3 + ψ4·f4 - ψ5·f5

[0126] Step 103: According to the geographical position distance, the nearest cluster head node of other nodes is selected to join, the fitness value of each gray wolf is calculated based on the constructed fitness function, and the alpha wolf, beta wolf and delta wolf are determined according to the fitness value of each gray wolf;

[0127] Step 104: The position of the gray wolf is updated according to the position of the alpha wolf, beta wolf and delta wolf;

[0128] Step 105: The fitness value of the gray wolf after position updating is calculated, the alpha wolf, beta wolf and delta wolf are re-determined according to the fitness value of the gray wolf after updating, and and

[0129] Step 106: If the current round reaches the maximum iteration number, the iteration is ended, the optimal individual is output, and the final clustering scheme with the maximum fitness value is obtained;

[0130] Step 107: If the current round does not reach the maximum iteration number, the step 104 is turned to.

[0131] Preferably, the position of the gray wolf is updated according to the position of the alpha wolf, beta wolf and delta wolf, and the position of the gray wolf is updated according to the position of the alpha wolf, beta wolf and delta wolf.

[0132]

[0133] Wherein, F1, F2 and F3 are optimized weight factors, F α , F β and F δ respectively represent the fitness of the alpha wolf, beta wolf and delta wolf in the gray wolf algorithm; represent the position of the gray wolf individual after updating.

[0134] After the iteration is completed, the final alpha wolf is obtained, that is, the final cluster head set is obtained.

[0135] For node mobility, the relative speed of the cluster head candidate node and other neighbor nodes is mainly considered, when the relative speed of the node and its neighbor nodes is low, the node should be more suitable to become a cluster head, and the cluster connection will be more stable, thereby reducing the cluster head reselection, which is specifically depicted as:

[0136]

[0137] Wherein, NUM i represents the number of neighbor nodes of the cluster head node N i ; and V represents the moving speed of the cluster head node j V represents the moving speed of the cluster head node V represents the moving speed of the neighbor node of the cluster head node.

[0138] For the node residual energy fraction, the node with higher residual energy fraction should be selected as much as possible, so as to ensure that the cluster head can work continuously, which is specifically depicted as:

[0139]

[0140] E represents the maximum energy of the cluster head node; max E represents the maximum energy of the cluster head node; E represents the residual energy of the cluster head node.

[0141] For the node position, the cluster head should be as evenly distributed as possible in the mobile ad hoc network, so the distance between cluster heads is introduced into the fitness function, and the distance between nodes N i and N i′ is specifically depicted as follows:

[0142]

[0143] and and represent the longitude and latitude coordinates of the cluster head node CHi′ , and and represent the longitude and latitude coordinates of the cluster head node CHi , arcsin represents the inverse sine function; R earth represents the radius of the earth.

[0144] For the dual-mode communication node, it has special properties, and it should be as much as possible to make every cluster exist a dual-mode communication node to maintain the contact between the mobile ad hoc network and the low-orbit satellite. Therefore, a reward factor is introduced to encourage the dual-mode communication node to become a cluster head, and a penalty factor is introduced to increase the penalty when there is no dual-mode communication node in the cluster, which is specifically depicted as:

[0145] The reward factor and the penalty factor include:

[0146] f4=1+ρ1·τ

[0147] f5=ρ2(1-τ)

[0148]

[0149] wherein ρ1 and ρ2 represent a constant greater than 0.

[0150] In summary, the final optimized fitness function F can be written as:

[0151]

[0152] The optimized fitness function F is used to update the original position of the grey wolf optimization algorithm The update is as follows:

[0153]

[0154] Wherein, F1, F2 and F3 are optimized weight factors, F α , F β and F δ respectively represent the fitness of the alpha wolf, beta wolf and delta wolf in the grey wolf algorithm; The position of the updated grey wolf individual.

[0155] According to the optimized fitness function, the fitness of each grey wolf (i.e. the node of the mobile ad hoc network) is calculated, and the cluster head election is finally completed.

[0156] After the cluster head election is completed, other nodes select the nearest cluster head node according to the geographical distance to join.

[0157] Subsequently, the command center broadcasts the clustering result to the mobile ad hoc network through the satellite, and the nodes of the mobile ad hoc network execute the clustering result.

[0158] Through the information collection stage, the nodes of the mobile ad hoc network obtain the node information of the entire network, but the nodes of the mobile ad hoc network are randomly moving, the topology connection has time-varying nature, and the nodes maintain frequently changing network information, which will seriously consume the energy of the nodes. Therefore, in order to reduce the storage overhead of the nodes, the cluster member nodes only need to maintain the node information within the cluster.

[0159] Step 4. After the cluster establishment stage, the nodes of the mobile ad hoc network know the cluster and the cluster head in which they are located. However, the nodes of the mobile ad hoc network have random mobility, and the establishment and disconnection of links between nodes are very common. Based on the relationship between position and relative speed, the reconstruction of the cluster or the link can be reduced as much as possible, thereby reducing the number of times of sending control information and reducing the overhead of the entire network.

[0160] 9. The leaving, joining and cluster head updating operations of the cluster members include:

[0161] After the clustering of the mobile ad hoc network is completed, the cluster head CH i will insert the cluster head identifier in the HELLO information, and the ordinary node N i will obtain the position and moving speed of the cluster head according to the HELLO information, and obtain the distance and relative speed

[0162] Leaving of the cluster member: when there is a node in the cluster that satisfies , v thresholdis a relative speed threshold value, the node is removed from the cluster;

[0163] Joining of cluster members: when there is a node outside the cluster satisfying , the node outside the cluster is added to the cluster;

[0164] Cluster head update: when the residual energy of the cluster head node is lower than a threshold value, the cluster head broadcasts cluster head update information, and the in-cluster nodes calculate a score R i according to their own residual energy and the relative moving speed with the remaining in-cluster nodes, and the score is calculated as follows:

[0165]

[0166] wherein R i represents the score of the in-cluster node N i , σ1 and σ2 represent the weight of each factor, NOC i represents the number of in-cluster nodes; V i and V j represent the moving speed of N i and N j of the in-cluster nodes; E max is the maximum energy of the node; E i represents the residual energy of the in-cluster node N i ;

[0167] After each in-cluster node calculates its own score, it broadcasts a HELLO message with its own score;

[0168] When each in-cluster node receives a HELLO message with a higher score than its own, the node does not broadcast a HELLO message with its own score;

[0169] When a node broadcasts a HELLO message with a score, and within a preset time period, it does not receive a HELLO message with a score from the remaining nodes, the node becomes a new cluster head.

[0170] In summary, in the star-ground fusion scenario, the application collects network information from a dual-mode communication terminal device by using the wide coverage of a satellite network, and then transmits the information to a command center with high computing power. The command center uses GWO for clustering and networking from a global network perspective, thereby improving the efficiency of mobile ad hoc network clustering operations, reducing the clustering calculation overhead and energy consumption of ad hoc network nodes, and thereby prolonging the life of the entire network. In addition, the fitness function of the grey wolf optimization algorithm is adaptively optimized by the optimized grey wolf optimization algorithm, which avoids the algorithm from falling into local optimum and improves the efficiency and rationality of the algorithm.

[0171] Finally, it is to be explained that the above embodiments are only used to illustrate the technical solutions of the present application but not to limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced without departing from the purpose and scope of the present application, and all should be covered in the scope of the claims of the present application.

Claims

1. A mobile ad hoc network clustering strategy based on GWO in a satellite-ground fusion scenario, characterized in that, The networking strategy comprises an information collection stage, an information transmission stage, a cluster establishment stage and a cluster maintenance stage; In the information collection stage, the mobile ad hoc network node broadcasts a HELLO information packet within its communication range; the mobile ad hoc network node receives the HELLO message of other nodes and maintains a node information table, wherein the node information table comprises the number of the node, the residual energy of the node, the position of the node, the moving speed of the node, the communication strength between the node and the satellite node and the number of the next hop node of the node; In the information transmission stage, the dual-mode communication node in the mobile ad hoc network selects the optimal satellite node to send the node information table to the command center according to the uplink load of the dual-mode communication node and the satellite node, the signal strength between the dual-mode communication node and the satellite node, the visibility of the dual-mode communication node and the satellite node and the communication delay between the dual-mode communication node and the satellite node; the command center analyzes and integrates the node information of the mobile ad hoc network; In the cluster establishment stage, the command center clusters the nodes in the mobile ad hoc network by using the optimized grey wolf optimization algorithm according to the node information of the mobile ad hoc network, and sends the clustering result to the dual-mode communication node through the satellite node; the dual-mode communication node broadcasts the clustering result in the mobile ad hoc network; The clustering of the nodes in the mobile ad hoc network by using the optimized grey wolf optimization algorithm comprises: Step 101: initializing the population size, the population dimension and the maximum iteration number, and randomly initializing the position of each grey wolf individual, wherein the position of each grey wolf individual corresponds to a cluster head node selection scheme; Step 102: constructing an optimized fitness function, the optimized fitness function F is as follows: F = ψ1·f1 + ψ2·f2 + ψ3·f3 + ψ4·f4 - ψ5·f5 Wherein, f1 represents the relative speed of the cluster head node and its neighbor nodes; f2 represents the residual energy fraction of the cluster head node; f3 represents the shortest distance between the cluster head nodes; f4 and f5 represent the reward factor and the penalty factor, which are used to ensure that there is a dual-mode communication node in each cluster and encourage the dual-mode communication node to become a cluster head; ψ1, ψ2, ψ3, ψ4 and ψ5 are weights, and ψ1 + ψ2 + ψ3 + ψ4 + ψ5 = 1; Step 103: selecting the nearest cluster head node of other nodes according to the geographical position distance, calculating the fitness value of each grey wolf based on the constructed fitness function, and determining the alpha wolf, the beta wolf and the delta wolf according to the fitness value of each grey wolf; Step 104: updating the position of the grey wolf according to the position of the alpha wolf, the beta wolf and the delta wolf; Step 105: calculating the fitness value of the grey wolf after the position is updated, and re-determining the alpha wolf, the beta wolf and the delta wolf according to the fitness value of the grey wolf after the position is updated; Step 106: if the current round reaches the maximum iteration number, the iteration is ended, the optimal individual is outputted, and the final clustering scheme with the maximum fitness value is obtained; Step 107: if the current round does not reach the maximum iteration number, the step 104 is turned to; The updating of the position of the grey wolf according to the position of the alpha wolf, the beta wolf and the delta wolf comprises: wherein F1, F2 and F3 are optimized weight factors, F α , F β and F δ represent the fitness of the alpha wolf, beta wolf and delta wolf in the grey wolf algorithm, respectively; represents the position of the updated grey wolf individual; In the cluster maintenance phase, the node information table of each mobile ad hoc network node maintains the node information of the cluster to which it belongs, and performs cluster member departure, joining and cluster head update operations according to the maintained information, wherein the cluster head node simultaneously maintains the information of the nodes in the cluster and the nodes outside the cluster. 2.The mobile ad hoc network clustering strategy based on GWO in a star-ground fusion scenario according to claim 1, characterized in that, The dual-mode communication node selects the optimal satellite node, including: Dual-mode communication node N i According to its uplink load with the satellite node S j , its signal strength with the satellite node S j , its visibility with the satellite node S j , and its communication latency with the satellite node S j , the score of the satellite node S j , the score of the satellite node S j includes: SV ij = ω1 · (1 - U ij ) + ω2 · RSS ij + ω3 · V ij - ω4 · D ij Among them, SV ij For dual-mode communication node N i , satellite node S j The score of each factor; ω1, ω2, ω3 and ω4 represent the weight parameters of each factor; U ij Represents dual-mode communication node N i With satellite node S j Uplink load; RSS ij Represents dual-mode communication node N i With satellite node S j Signal strength; V ij Represents dual-mode communication node N i With satellite node S j Visibility; D ij Represents dual-mode communication node N i With satellite node S j communication delay. 3.The GWO-based mobile ad hoc network clustering strategy in a satellite-ground integrated scenario according to claim 2, characterized in that, The dual-mode communication node N i with the satellite node S j visibility V ij comprises: where θ ij represents the zenith angle of the satellite node S i with respect to the dual-mode communication node N j , and is represented as: wherein, and respectively represent the longitude and latitude of the dual-mode communication node N i and the satellite node S j , and θ threshold represents a set threshold value.

4. The mobile ad hoc network clustering strategy based on GWO in a star-ground fusion scenario according to claim 1, characterized in that, The relative speed of the cluster head node and its neighbor nodes includes: wherein, NUM i represents the number of neighbor nodes of the cluster head node N i ; V represents the moving speed of the cluster head node; and j represents the moving speed of the neighbor nodes of the cluster head node .

5. The mobile ad hoc network clustering strategy based on GWO in a star-ground fusion scenario according to claim 1, characterized in that, The residual energy fraction of the node includes: wherein E max represents the maximum energy of the cluster head node; represents the residual energy of the cluster head node.

6. The mobile ad hoc network clustering strategy based on GWO in a star-ground fusion scenario according to claim 1, characterized in that, The distance between the cluster head nodes includes: wherein, and denotes the latitude and longitude coordinates of the cluster head CH i′ , and denotes the latitude and longitude coordinates of the cluster head CH i , arcsin denotes the inverse sine function; R earth denotes the radius of the earth.

7. The mobile ad hoc network clustering strategy based on GWO in a star-ground fusion scenario according to claim 1, characterized in that, The reward factor and the penalty factor include: f4 = 1 + p1 * tau f5 = p2 * (1 - tau) Wherein, p1 and p2 represent a constant greater than 0.

8. The mobile ad hoc network clustering strategy based on GWO in a star-ground fusion scenario according to claim 1, characterized in that, The cluster member departure, joining and cluster head update operations include: Mobile Ad hoc Network, after clustering is completed, cluster head CH i will insert cluster head identification in HELLO information, ordinary node N i According to the HELLO information, the position and moving speed of the cluster head are obtained, and the distance and relative speed between the cluster head and itself are known and relative speed Departure of cluster members: When there exists a node within the cluster that satisfies v threshold is a relative speed threshold, then the node is removed from the cluster; Joining of cluster members: When there exists a cluster-out node satisfying then the cluster-out node is joined to the cluster; Cluster head update: when the residual energy of the cluster head node is lower than the threshold, the cluster head broadcasts cluster head update information, and the nodes in the cluster calculate the score R according to their own residual energy and the relative moving speed with the rest of the nodes in the cluster i The specific calculation formula of the score is as follows: wherein R i represents the score of a node N i within a cluster, σ1 and σ2 represent the weight of each factor, NOC i represents the number of nodes within a cluster; V i and V j represent the moving speed of N i and N j of nodes within a cluster; E max is the maximum energy of a node; E i represents the residual energy of a node N i within a cluster. After each node in the cluster calculates its own score, it broadcasts the HELLO information with its own score; When each node in the cluster receives the HELLO information with a higher score than its own, the node does not broadcast the HELLO information with its own score; When a node broadcasts the HELLO information with a score, and does not receive the HELLO information with a score from the remaining nodes within a predetermined time period, the node becomes a new cluster head.

Images