CN116867025A

CN116867025A - Sensor node clustering method and device in wireless sensor network

Info

Publication number: CN116867025A
Application number: CN202310743119.7A
Authority: CN
Inventors: 郭少勇; 王冰沁; 熊翱; 陈钰; 才智; 姚辉; 彭凯
Original assignee: Xingyaoneng Beijing Technology Co ltd; Beijing University of Posts and Telecommunications
Current assignee: Xingyaoneng Beijing Technology Co ltd; Beijing University of Posts and Telecommunications
Priority date: 2023-06-21
Filing date: 2023-06-21
Publication date: 2023-10-10

Abstract

The application provides a sensor node clustering method and device in a wireless sensor network, wherein the method comprises the following steps: adopting a flow prediction means, and obtaining overall communication prediction energy consumption of the sensor node set in a preset future time period based on communication flow data sets of all the sensor nodes in the wireless sensor network; wherein the set of sensor nodes includes each of the sensor nodes; constructing an adaptability function for selecting an optimal cluster head according to the overall communication prediction energy consumption; and obtaining optimal cluster head distribution data corresponding to the sensor node set according to the fitness function and a butterfly optimization algorithm. The method and the device can select the optimal cluster head based on the predicted energy consumption, so that the rationality of selecting the cluster head node can be effectively improved, and the communication energy consumption in the wireless sensor network can be effectively reduced.

Description

Sensor node clustering method and device in wireless sensor network

Technical Field

The application relates to the field of communication of the Internet of things, in particular to a sensor node clustering method and device in a wireless sensor network.

Background

In recent years, with the rapid development of 5G and Internet of things technologies, more and more Internet of things terminal devices are applied to the production and life of people. While the intelligent terminal equipment is convenient for people to live, the intelligent terminal equipment is continuously advancing to the direction of higher quality and higher performance. The performance of the intelligent terminal equipment is greatly improved, the power consumption is increased, the size of the intelligent terminal equipment is small, and the capacity of a placed battery is limited, so that how to fully utilize the limited energy of the intelligent terminal equipment and prolong the service life of the intelligent terminal equipment is always a problem to be solved. These intelligent terminal devices constitute a wireless ad hoc network, and the wireless sensor network, as a special branch in the field of wireless ad hoc networks, has characteristics similar to those of wireless ad hoc networks: the sensor nodes have limited energy. If the nodes in the network cannot work due to energy consumption, the problems of network topology change, route reestablishment and the like can be caused, and even communication interruption can be caused. Therefore, how to save battery energy of the wireless sensor network as much as possible without affecting functions becomes a core problem in the design of software and hardware of the wireless sensor network.

In the prior art, when a cluster head is selected in a clustering stage of a clustering routing protocol, the problem of future communication energy consumption possibly generated after a sensor node becomes the cluster head is not considered, and the problem that only historical data is referred to and the future energy consumption is not considered exists.

Disclosure of Invention

In view of this, embodiments of the present application provide a method and apparatus for clustering sensor nodes in a wireless sensor network, so as to eliminate or improve one or more drawbacks existing in the prior art.

The first aspect of the application provides a sensor node clustering method in a wireless sensor network, which comprises the following steps:

adopting a flow prediction means, and obtaining overall communication prediction energy consumption of the sensor node set in a preset future time period based on communication flow data sets of all the sensor nodes in the wireless sensor network; wherein the set of sensor nodes includes each of the sensor nodes;

constructing an adaptability function for selecting an optimal cluster head according to the overall communication prediction energy consumption;

and obtaining optimal cluster head distribution data corresponding to the sensor node set according to the fitness function and a butterfly optimization algorithm.

In some embodiments of the present application, the means for using traffic prediction in the wireless sensor network to obtain, based on the communication traffic data set of each sensor node, overall communication predicted energy consumption of the sensor node set in a preset future time period includes:

Predicting the communication flow data of each sensor node according to the communication flow prediction model to obtain a corresponding prediction flow set;

and obtaining the whole communication prediction energy consumption of the sensor node set according to each prediction flow set.

In some embodiments of the application, said deriving said overall communication predicted energy consumption of said set of sensor nodes from each of said set of predicted traffic comprises:

obtaining the transmission energy consumption of each sensor node serving as a non-cluster head node in the sensor node set for transmitting communication data to each corresponding sensor node serving as a cluster head node according to each predicted traffic set, and the receiving energy consumption of each cluster head node for receiving the communication data from each corresponding non-cluster head node and the forwarding energy consumption of each cluster head node for forwarding the communication data to a communication base station;

and obtaining the whole communication prediction energy consumption according to each sending energy consumption, each forwarding energy consumption and each receiving energy consumption.

In some embodiments of the present application, before the constructing an fitness function for selecting an optimal cluster head according to the overall communication predicted energy consumption, the method further includes:

Calculating cluster head residual energy of each cluster head node, and obtaining overall cluster head residual energy according to each cluster head residual energy;

calculating first communication distances from the corresponding non-cluster head nodes to the cluster head nodes in the cluster head nodes, obtaining second communication distances corresponding to the cluster head nodes according to the first communication distances in the cluster head nodes, and obtaining third communication distances according to the second communication distances;

obtaining a base station communication distance according to the distance between each cluster head node and the communication base station;

and obtaining the number of the whole nodes according to the number of the non-cluster head nodes in each cluster head node.

In some embodiments of the present application, the constructing an fitness function for selecting an optimal cluster head according to the overall communication predicted energy consumption includes:

and obtaining the fitness function according to the overall communication prediction energy consumption, the overall cluster head remaining energy, the third communication distance, the base station communication distance, the overall node number and the corresponding preset priority coefficients.

In some embodiments of the present application, the obtaining, according to the fitness function and a butterfly optimization algorithm, optimal cluster head distribution data corresponding to the sensor node set includes:

Initializing a plurality of butterflies and iteration parameters; wherein the iteration parameters include: switching probability, maximum iteration number and sensing mode;

performing a preset iteration step on each butterfly to obtain current optimal cluster head distribution data;

and if the current iteration number is smaller than or equal to the maximum iteration number, continuing to execute the iteration step on each updated butterfly until the iteration number is larger than the maximum iteration number, and outputting the current cluster head distribution data as the optimal cluster head distribution data.

In some embodiments of the application, the optimizing step comprises:

obtaining the stimulus intensity of each butterfly at the current position according to the fitness function;

obtaining the fragrance concentration corresponding to each butterfly according to the stimulus intensity, and selecting cluster head distribution data corresponding to the butterfly with the minimum fragrance concentration as the current optimal cluster head distribution data;

generating a random number, and updating the current positions corresponding to the butterflies according to a first position updating formula if the random number is larger than the switching probability; if the random number is smaller than or equal to the switching probability, updating the current positions corresponding to the butterflies according to a second position updating formula;

And updating the sensing mode.

A second aspect of the present application provides a sensor node clustering apparatus in a wireless sensor network, the apparatus comprising:

the energy consumption prediction module is used for obtaining the whole communication prediction energy consumption of the sensor node set in a preset future time period based on the communication flow data set of each sensor node in the wireless sensor network by adopting a flow prediction means; wherein the set of sensor nodes includes each of the sensor nodes;

the fitness function construction module is used for constructing a fitness function for selecting an optimal cluster head according to the overall communication prediction energy consumption;

and the optimal cluster head distribution screening module is used for obtaining optimal cluster head distribution data corresponding to the sensor node set according to the fitness function and a butterfly optimization algorithm.

A third aspect of the present application provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method for clustering sensor nodes in a wireless sensor network according to the first aspect when executing the computer program.

A fourth aspect of the present application provides a computer readable storage medium having stored thereon a computer program which when executed by a processor implements the method for clustering sensor nodes in a wireless sensor network according to the first aspect.

Additional advantages, objects, and features of the application will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the application. The objectives and other advantages of the application may be realized and attained by the structure particularly pointed out in the written description and drawings.

It will be appreciated by those skilled in the art that the objects and advantages that can be achieved with the present application are not limited to the above-described specific ones, and that the above and other objects that can be achieved with the present application will be more clearly understood from the following detailed description.

Drawings

The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate and together with the description serve to explain the application. The components in the figures are not necessarily to scale, emphasis instead being placed upon illustrating the principles of the application. Corresponding parts in the drawings may be exaggerated, i.e. made larger relative to other parts in an exemplary device actually manufactured according to the present application, for convenience in showing and describing some parts of the present application. In the drawings:

fig. 1 is a flow chart of a sensor node clustering method in a wireless sensor network according to an embodiment of the application.

Fig. 2 is a schematic structural diagram of a sensor node clustering device in a wireless sensor network according to another embodiment of the present application.

Fig. 3 is a schematic diagram of a wireless communication energy consumption model of a sensor node according to another embodiment of the present application.

Fig. 4 is a schematic diagram of a butterfly optimization algorithm implementation in another embodiment of the application.

FIG. 5 is a flow chart of ARIMA model modeling in another embodiment of the present application.

Detailed Description

The present application will be described in further detail with reference to the following embodiments and the accompanying drawings, in order to make the objects, technical solutions and advantages of the present application more apparent. The exemplary embodiments of the present application and the descriptions thereof are used herein to explain the present application, but are not intended to limit the application.

It should be noted here that, in order to avoid obscuring the present application due to unnecessary details, only structures and/or processing steps closely related to the solution according to the present application are shown in the drawings, while other details not greatly related to the present application are omitted.

It should be emphasized that the term "comprises/comprising" when used herein is taken to specify the presence of stated features, elements, steps or components, but does not preclude the presence or addition of one or more other features, elements, steps or components.

It is also noted herein that the term "coupled" may refer to not only a direct connection, but also an indirect connection in which an intermediate is present, unless otherwise specified.

Hereinafter, embodiments of the present application will be described with reference to the accompanying drawings. In the drawings, the same reference numerals represent the same or similar components, or the same or similar steps.

The following examples are provided to illustrate the invention in more detail.

The wireless sensor mainly comprises four modules of a sensor, a processor, wireless communication and energy supply, and can complete functions of real-time monitoring, data acquisition, wireless communication and the like. Research shows that in the first three modules, the energy consumption of the wireless communication module is far greater than that of the sensor and the processor module, and the wireless communication module becomes a main power consumption source of the wireless sensor. At present, the communication energy consumption of the wireless sensor network can be reduced from various aspects such as hardware, data fusion, hierarchical protocol and the like. In terms of hardware energy conservation, the Active state and the Sleep state of the processor MCU, the Active state and the Sleep state of the digital-to-analog converter ADC and the transmitting and receiving state of the wireless data transmitting/receiving module WTRU can be respectively set in the wireless sensor network, and after the control command designed according to the primitive is sent from the monitoring center through the processor, the conversion of the states is controlled, so that the application layer, the network layer, the link layer and the physical layer of the wireless sensor network are subjected to cross-layer energy conservation optimization, the effect of reducing the whole energy consumption of the node is achieved on the premise that the application performance and the function are not influenced, and the service life of a battery used by the node and the network life time are prolonged. In the aspect of data fusion energy saving, the current data fusion technology is more researched, and besides the existing cluster-based, chain-based and tree-based data fusion technology, researchers also adopt a BP neural network-based data fusion algorithm to apply the BP neural network to a clustering routing protocol for feature extraction, so that redundant data is reduced, and node death time is slowed down; the BP neural network data fusion algorithm GA-BP based on Genetic Algorithm (GA) optimizes the weight and threshold parameters of the BP neural network by using the genetic algorithm, reduces redundant data and prolongs the network life cycle; the WSN data fusion algorithm PSO-BP based on the particle swarm optimization BP neural network utilizes the particle swarm algorithm to optimize BP neural network parameters, and the optimized BP neural network and the sensor network clustering routing protocol are organically combined, so that the data fusion efficiency is effectively improved, and the network energy consumption is balanced. In the aspect of hierarchical protocol energy conservation, a basic hierarchical routing protocol, such as a low-energy-consumption self-adaptive distributed routing protocol LEACH and a hybrid clustering protocol HEED, and a plurality of hierarchical protocols which are improved on the basis of the basic hierarchical routing protocol, such as LEACH-C, ALEACH, LEACH-MAC, M-LEACH, MMR and the like, are generally adopted, and the hierarchical protocols are distributed by optimizing cluster preference and balancing cluster head, so that a plurality of modes of 'hot zone' and the like are adopted, the energy utilization rate of the wireless sensor network is improved, the node energy consumption is reduced, and the service life of the network is prolonged. The clustering routing protocol is a key point of research because of simplicity, high efficiency and good balance of node energy consumption.

The embodiment of the application provides a sensor node clustering method in a wireless sensor network, which can be executed by a sensor node clustering device in the wireless sensor network, and referring to fig. 1, the sensor node clustering method in the wireless sensor network specifically comprises the following contents:

step 110: adopting a flow prediction means, and obtaining overall communication prediction energy consumption of the sensor node set in a preset future time period based on communication flow data sets of all the sensor nodes in the wireless sensor network; wherein the set of sensor nodes includes each of the sensor nodes.

Step 110 includes:

step 112: and predicting the communication flow data of each sensor node according to the communication flow prediction model to obtain a corresponding prediction flow set.

Step 113: and obtaining the whole communication prediction energy consumption of the sensor node set according to each prediction flow set.

Step 113 comprises:

step 114: and obtaining the transmission energy consumption of each sensor node serving as a non-cluster head node in the sensor node set for transmitting communication data to each corresponding sensor node serving as a cluster head node according to each predicted traffic set, the receiving energy consumption of each cluster head node for receiving the communication data from each corresponding non-cluster head node, and the forwarding energy consumption of each cluster head node for forwarding the communication data to a communication base station.

Step 115: and obtaining the whole communication prediction energy consumption according to each sending energy consumption, each forwarding energy consumption and each receiving energy consumption.

Specifically, in step 110, communication traffic data of each sensor node is first predicted according to a communication traffic prediction model to obtain a corresponding predicted traffic set. Secondly, according to each predicted flow set, obtaining the transmission energy consumption of each sensor node serving as a non-cluster head node in the sensor node set for transmitting communication data to each corresponding sensor node serving as a cluster head node, the receiving energy consumption of each cluster head node for receiving the communication data from each corresponding non-cluster head node, and the forwarding energy consumption of each cluster head node for forwarding the communication data to a communication base station; and finally, obtaining the whole communication prediction energy consumption according to each sending energy consumption, each forwarding energy consumption and each receiving energy consumption, so that the communication energy consumption in a preset future time period can be effectively predicted, the rationality of cluster head node selection is improved, and the communication energy consumption in the wireless sensor network is effectively reduced.

The communication traffic prediction model can be called a differential autoregressive moving average model by using an ARIMA model (Autoregressive Integrated Moving Average Model).

The overall communication predicted energy consumption can then be calculated from equation (3-1).

Wherein m represents the number of cluster head nodes, I _j Represents CH _j Cluster member node (i.e. non-cluster head node),indicating that in a certain cluster, the cluster member node transmits communication data to the cluster head node CH _j Energy consumption by->Representing cluster head node CH _j And receiving and transmitting the communication data to the communication base station.

In addition, the following provisions are made for the sensor nodes and base stations in a Wireless Sensor Network (WSN):

(1) Once all sensor nodes and Base Stations (BS) in the WSN are deployed, the location is fixed, and each node has a unique ID.

(2) The base station has infinite energy and is located at the top of a square area with known locations.

(3) All nodes in the WSN are identical in characteristics, such as capacity, volume, etc.

(4) All nodes have the same and limited initial energy and can report their own location information.

(5) All nodes should periodically transmit data to the target node, and may also autonomously communicate with the base station.

Referring to fig. 3, a sensor node sends an L-bit data packet, and when the communication distance is d, a cluster member node transmits communication data to a cluster head node CH _j The energy consumption is calculated by the formulas (3-2) and (3-3).

Wherein E is _elec The energy consumption of a transceiving circuit, namely the energy consumed by transceiving 1 bit data; epsilon _fs Amplifier energy loss coefficient epsilon as a free space attenuation model _mp An amplifier energy loss coefficient which is a multipath fading model; d, d ₀ For the communication distance threshold, from ε _fs And epsilon _mp And determining the value as a fixed value.

Cluster head node CH _j Energy consumption by receiving and transmitting communication data to communication base stationCalculated by formulas (3-4) and (3-5).

E _RX (L)＝L×E _elec (3-4)

E _DA (L)＝L×E _da (3-5)

Wherein E is _da The energy consumed by the 1-bit data is fused for the nodes.

Step 120: and constructing an adaptability function for selecting the optimal cluster head according to the overall communication prediction energy consumption.

Also included before step 120 is:

Step 120 includes:

Specifically for step 120, in the wireless sensor network, the cluster head node performs various tasks, collects communication data from non-cluster head nodes and transmits the communication data to the communication base station. The cluster head needs higher energy to complete the task, so the node with higher residual energy is preferably selected as the cluster head node. The overall cluster head residual energy is calculated by the formula (3-6).

Wherein m is the number of cluster heads, E _CHi Is the remaining energy of the ith cluster head.

The first communication distance defines a distance between a common node and its own cluster head. When the transmission distance from the common node to the cluster head is smaller, the node energy consumption is smaller. The third communication distance is calculated by the formula (3-7).

Wherein m is the number of cluster heads and node _i Is of CH _j Is (node) _i ,CH _j ) For nodes i and CH _j Distance of I _j Representing a CH _j Is defined by the number of nodes.

The base station communication distance defines the transmission distance from the cluster head node to the communication base station. The energy consumption of a node depends on the distance of the transmission path. If the communication base station is located far from the cluster head node, more energy is required to transmit the data. Thus, a sudden drop in cluster head node energy may occur due to an increase in energy consumption. Therefore, in the data transmission process, the sensor node with smaller distance from the communication base station is preferred to be used as the cluster head node. The base station communication distance is calculated by the equation (3-8).

Wherein m is the number of cluster heads, dis (CH) _j BS) is CH _j Distance from BS.

The overall node number defines the number of common nodes (i.e., non-cluster head nodes) belonging to each cluster head node, i.e., the number of cluster members, and fewer cluster heads with fewer cluster members are selected because cluster heads with more node members lose energy in a shorter time. The number of the integral nodes is calculated by the formula (3-9).

Wherein m is the number of cluster head nodes, I _i Is belonging to cluster head node CH _j Is not the number of cluster head nodes.

By calculating the overall cluster head residual energy, the third communication distance, the base station communication distance and the overall node number, the rationality of cluster head node selection can be effectively improved, and therefore communication energy consumption in the wireless sensor network is effectively reduced.

And finally, obtaining a fitness function according to the overall communication prediction energy consumption, the overall cluster head residual energy, the third communication distance, the base station communication distance, the overall node number and the corresponding preset priority coefficients, wherein the fitness function is obtained by calculating (3-10).

Wherein, take alpha ₁ ，α ₂ ，α ₃ ，α ₄ ，α ₅ The values of (2) are 0.25, 0.2, 0.1, 0.25, respectively. Consider alpha ₁ And alpha ₅ And taking the residual energy and the predicted energy consumption as the first consideration factors for the first priority, and avoiding the node from being invalid as the cluster head. Subsequently, consider alpha ₂ And alpha ₃ And for the second priority, taking a shorter distance from the cluster head node or the communication base station as a secondary consideration, thereby reducing the communication distance and enabling the energy loss to be smaller. Node degree alpha ₄ And selecting cluster head nodes with smaller node degree as the third priority.

Step 130: and obtaining optimal cluster head distribution data corresponding to the sensor node set according to the fitness function and a butterfly optimization algorithm.

Step 130 includes:

Wherein, the optimization steps are as follows:

and updating the sensing mode.

Specifically, for step 130, in the butterfly optimization algorithm, the fragrance of the butterfly as a function of the physical stimulus intensity is formulated as formula (3-11):

F＝cI ^a (3-11)

wherein: f is the size of the fragrance perceived by other butterflies, i.e. the fragrance concentration, c is the perceived mode, I is the stimulus intensity, a is the power of the mode dependent exponent, which can be understood as the different degree of fragrance absorption. The values of a and c are in the range of 0,1]. The iterative variation formula of c is formula (3-12), wherein T _max The maximum number of iterations is:

in addition, the butterfly optimization algorithm has three stages:

1. in the initialization stage, the algorithm defines an objective function and a solution space thereof, distributes parameter values, initializes the butterfly population, randomly generates butterfly positions in a search space, and calculates and stores stimulus intensity values I and fragrance concentration values f of the butterfly positions.

2. In the iteration stage, the algorithm executes a plurality of iterations, in each iteration, all butterfly vectors in the search space are moved to a new position, the vector of each butterfly is updated, then the respective stimulus intensity I is recalculated, and then the fragrance f generated by each butterfly at the new position is calculated according to the formula (3-11). Among them, there are two kinds of butterfly movement modes:

1) Global search mode: butterfly-to-optimal butterfly (current optimal position g) ^* ) For the ith butterfly, the location update is expressed by equation (3-13).

Wherein:for the position vector of butterfly individual i at iteration number t,/>For the new position vector g of butterfly individual i after iteration ^* For the current global optimal solution, also a vector, r E [0,1 ]]Is a random number, f _i Is the fragrance concentration of butterfly i.

2) Local search mode: if the individual butterfly cannot feel the fragrance concentration of other butterflies, the individual butterfly can move locally and randomly, and the position updating mode is expressed by the formula (3-14).

Wherein:for the position vector of butterfly individual i at iteration number t,/>New position vector for butterfly individual i after iteration, < ->And->The positions of any jth butterfly and kth butterfly in the solution space are respectively a position vector, f _i For butterfly i fragrance concentration, r.epsilon.0, 1]Is a random number. If->And->Belonging to the same butterfly population, formula (3-11) represents a local random mobile search.

The butterfly population adopts a global search or local search mode, and depends on the switching probability p, p epsilon [0,1]. By generating random numbers r ₁ Compared with p, the search mode is switched.

3. And finally, when the stopping condition is met, ending the iteration. The stop condition may be defined in different ways, such as whether a maximum CPU usage time is reached, whether a maximum number of iterations is reached, whether a particular error rate value is reached, or any other suitable criteria. When the iteration phase ends, the algorithm outputs the optimal solution vector with the best fitness.

Referring first to fig. 4, the following steps are performed: wherein the iteration steps are S2-S10, N in the figure indicates no, and Y indicates yes.

S1: initializing a plurality of butterflies and iteration parameters; wherein the iteration parameters include: the switching probability p, the maximum iteration number max_gen and the sensing mode c, and the butterfly number k.

Where k is 100, each butterfly is represented as a 40-dimensional row vector, and the elements of each vector are 0 or 1, denoted BF _i ＝[x ₁ ,x ₂ ,x ₃ ,...,x ₄₀ ](i＝1，2，3...100)，x _j (j=1, 2, 3..40) is 0 or 1, and specifies the butterfly vector BF at the time of initialization _i Some and only 5 elements are 1 and the remaining elements are 0. Assume that each element x of the butterfly vector _j Are all corresponding to one node, x, in the wireless sensor network _j =0 represents a common node, x _j =1 represents the cluster head node. Each butterfly corresponds to a solution of cluster head distribution in the wireless sensor network, and corresponds to a determined clustering mode.

S2: the fitness function f is derived using equation (3-10).

S3: calculate that each butterfly individual is currently x _i Stimulation intensity i=f (x _i )。

S4: and (3) calculating the fragrance concentration F of each butterfly individual by using a formula (3-11) to obtain the current optimal solution (namely the current optimal cluster head distribution data).

S5: generating random number r ₁ If r1 > the switching probability p, jumping to the step S7; otherwise, step S6 is skipped. R 1? Representing whether the random number r1 is less than or equal to the handover probability p.

S6: and (3) entering a global search mode, moving the butterfly to the optimal solution by using a formula (3-13), and jumping to a step S8 after the execution is finished.

S7: entering the local search mode, the butterfly moves randomly using equation (3-14).

S8: and updating the butterfly position, and updating the sensing mode c by using the formula (3-12).

S9: if the current iteration number gen is less than or equal to max_gen, jumping to the step S3; otherwise, the process goes to step S10.

S10: and outputting the optimal solution to obtain the current optimal cluster head distribution data serving as the optimal cluster head distribution data.

The cluster head is selected on the basis of calculating the node residual energy, the distance from the node to the cluster head, the distance from the node to the base station, the node degree and the predicted communication energy consumption through a butterfly optimization algorithm, so that the rationality of cluster head node selection is improved, and the communication energy consumption in the wireless sensor network is effectively reduced.

The application also provides an embodiment of a sensor node clustering method in the wireless sensor network, which comprises the following steps:

1.1 related parameter settings

40 sensor nodes are randomly deployed in a space of size 100×100, and 1 base station. Some attributes are defined for the node, such as the remaining Energy power of the node, the cluster head distance Rc, the distance d from the base station, the node degree num_join, the predicted Energy consumption form_energy_con, etc. The relevant parameter settings are as follows:

the settings of the wireless sensor network parameters are shown in table 1.

TABLE 1

Parameter name	Numerical value
		Sensor network coverage area m×m	100m*100m
Number of sensor nodes n	40
		Base station location (BS) _x ,BS _y )	(50m,125m)
Sensor node initial energy E ₀	1.25J
		Parameter name	Numerical value
Energy loss E of transceiver circuit _elec	50nJ/bit

The parameter settings of the wireless communication energy consumption model are shown in table 2.

TABLE 2

Free space attenuation energy consumption coefficient epsilon _fs	10pJ/bit/m ²
		Multipath fading energy consumption coefficient epsilon _mp	0.0013pJ/bit/m ⁴
Communication distance threshold d ₀	87m

The relevant parameter settings for the butterfly optimization algorithm are shown in table 3.

TABLE 3 Table 3

Parameter name	Numerical value
		Population number k	100
Probability of handover p	0.8
		Power exponent a	0.1
Sensing modality c	0.01

1.2ARIMA model principle and modeling

1.2.1ARIMA model principle

Converting the non-stationary time sequence into stationary time sequence, and finding the real situation of the historical data by using a model established by regression of the dependent variable only on the hysteresis value of the dependent variable and the current value and the hysteresis value of the random error term, thereby predicting. It is a more common method of processing time series predictions. The ARIMA model evolved on the ARMA model, which actually differenced the data before using the ARMA model. Table 4 describes the parameters and meanings in ARIMA (p, d, q).

TABLE 4 Table 4

Parameters (parameters)	Parameter description
		p	Autoregressive order
d	The time series becomes the differential order required for the smooth series
		q	Moving average order
AR(p)	Autoregressive model
		MA(q)	Moving average model

The ARIMA model has the following form:

Y _t ＝β ₁ Y _t-1 +...+β _p Y _t-p +e _t +α ₁ e _t-1 +α ₂ e _t-2 +...+α _q e _t-q (3-6)

wherein, in the formula (3-5),represents d-order difference, ++>Representing a time series X that will not be stationary _t D step difference is made to obtain a stable time sequence Y _t Can only be processed with the ARMA model. In the formula (3-6), beta ₁ Y _t-1 +...+β _p Y _t-p +e _t Representing the AR (p) autoregressive model, the sequence value Y of the previous p-phase _t-1 ,Y _t-2 ,...,Y _t-p As independent variable, random variable Y _t Is the value y of (2) _t As a dependent variable, a linear regression model is established, and an AR (p) model can be understood as that the time sequence value at the current moment can be represented by the linear combination of p time sequence values in the past plus white noise, wherein the white noise refers to a random stable sequence with zero mean constant variance; in the formula (3-6), alpha ₁ e _t-1 +α ₂ e _t-2 +...+α _q e _t-q Representing MA (q) moving average model, random variable Y _t Is the value y of (2) _t Independent of the sequence values of the preceding phases, only random disturbances (white noise) e are established with the preceding q phases _t-1 ,e _t-2 ,...,e _t-q The MA (q) model can be understood as a linear combination of white noise of past q-order with the timing value of the current moment of the sequence.

Thus, the ARIMA model is to first convert the non-stationary sequence X _t Obtaining a stable sequence Y after d-order difference _t Combining AR and MA models to represent random variable Y _t Is the value y of (2) _t Not only to the sequence values of the previous p-phase but also to random perturbations of the previous q-phase.

1.2.2ARIMA model modeling process, see fig. 5, where N represents no and Y represents yes, described in detail below:

firstly, carrying out stability detection on an observation value sequence, and if the observation value sequence is not stable, carrying out differential operation on the observation value sequence until the differential data is stable;

after the data is stable, performing white noise test on the data, wherein the white noise refers to a random stable sequence with zero mean constant variance, and if the data is white noise, the analysis is finished and the time sequence prediction cannot be performed;

if the white noise test is passed, a stable non-white noise sequence is obtained, ACF and PACF images are drawn, ARMA and other model identification is carried out, and the specific identification method is shown in Table 5;

TABLE 5

Model	ACF	PACF
			AR(p)	Trailing tail	p-order tail cutting
MA(q)	q-order tail cutting	Trailing tail
			ARMA(p,q)	Trailing tail	Trailing tail
Unsuitable model	Tail cutting	Tail cutting

For the identified model, determining model parameters according to the ACF and PACF, if the q-order attenuation of the ACF is close to zero (geometric or oscillation type) and the p-order attenuation of the PACF is close to zero (geometric or oscillation type), determining the p and q parameters of the ARMA (p, q) model.

And then checking the model effect, visualizing the residual error of the trained model by residual error checking, and judging whether the residual error accords with normal distribution. If the distribution is normal, finishing the model establishment; if the distribution is not normal, useful information is described in the residual error, and an optimization model is needed.

Finally, in the following step 1.3, the communication traffic is predicted by the established ARIMA model.

Among them, the following description is given for ACF, PACF, tail-cutting, and tail:

ACF graph (autocorrelation graph): is a graph with the hysteresis order as the horizontal axis and the autocorrelation coefficient as the vertical axis. The horizontal axis is 1, representing X _t And X is _t-1 Is a correlation coefficient value of (1); the horizontal axis is 2, representing X _t And (X) _t-1 ，X _t-2 ) Is a correlation coefficient value of (1); the horizontal axis is n, representing Xt and (X) _t-1 ，X _t-2 ，…，X _t-n ) Is used for the correlation coefficient value of (a).

PACF plot (partial autocorrelation plot): the hysteresis order is plotted on the horizontal axis and the partial autocorrelation coefficient is plotted on the vertical axis. The horizontal axis is 1, representingX _t And X is _t-1 Is a correlation coefficient value of (1); the horizontal axis is 2, representing X _t And X is _t-2 Is a correlation coefficient value of (1); the horizontal axis is n, representing Xt and X _t-n Is used for the correlation coefficient value of (a).

Tail cutting: the sample autocorrelation/bias autocorrelation coefficients are significantly larger than 2 standard deviations in the initial d-order, while almost 95% of the autocorrelation/bias autocorrelation coefficients are within two standard deviations, and the process of decay from non-zero coefficients to small value fluctuations is very abrupt, like being "truncated". I.e. it quickly goes to 0 k-order tail-biting after being larger than a certain order (k), which can be simply understood as changing from a certain order to 0 directly after.

Tailing: more than 5% of the autocorrelation/bias autocorrelation coefficients are outside of the 2 standard deviation range, or decay to small fluctuations from significantly non-zero coefficients is relatively slow or continuous, or decays exponentially (sinusoidal waveform), "drags. That is, there is always a non-zero value, and the value will not quickly approach 0 (but will fluctuate around 0) after being larger than a certain order, and it can be simply understood that the value will not be 0 anyway, but will randomly change around 0 after a certain order.

1.3 description of time-sequential prediction model data

In order to simulate randomness of data traffic transmitted by each node in a wireless sensor network, a data set used for research is generated by random numbers, 40 x 2000 random numbers are generated, 40 is the number of the nodes, 2000 is a simulation round, each number is represented as transmission traffic of one node in a certain round, the data range is 1-150 bits, and the data are stored in a data. The dataset was split into a training set of 40 x 1500 and a simulation set of 40 x 500. The training set of 40 x 1500 was considered as a two-dimensional array, denoted T1[40,1500]. When each node i is trained, the data of the previous 1500 rounds, namely T1[ i,: ], are taken and trained to obtain ARIMA_i. Traversing i from 1 to 40, training completes 40 ARIMA models. Then, the prediction is started, and for each node i, the data T1[ i ] of the previous 1500 rounds is taken as input, and 500 data are output and obtained, namely the data are taken as the transmission flow of 500 rounds after the predicted node i. Traversing i from 1 to 40 yields 40 x 500 predicted traffic flows, denoted as a two-dimensional array form_tra [40,500]. The time sequence predicting part is finished to obtain a two-dimensional array form_tra [40,500], wherein form_tra [ i, j ] represents the predicted flow of the jth round of the ith node in the BOA_LEACH simulation. Meanwhile, the data set which is left after being split, namely a simulation set of 40×500, is denoted as real_tra [40,500] and is used as actual transmission flow in the simulation process of the data transmission stage.

1.4 butterfly optimization Algorithm implementation

The specific implementation process of finding cluster heads by the butterfly optimization algorithm is described herein:

the butterfly population size n=100 is defined first.

Then defining a solution space, representing each butterfly as a 40-dimensional row vector, wherein the element of each vector is 0 or 1, and the element is represented as BF _i ＝[x ₁ ,x ₂ ,x ₃ ,...,x ₄₀ ](i＝1，2，3...100)，x _j (j=1, 2, 3..40) is 0 or 1, and specifies the butterfly vector BF at the time of initialization _i Some and only 5 elements are 1 and the remaining elements are 0. Assume that each element x of the butterfly vector _j Are all corresponding to one node, x, in the wireless sensor network _j =0 represents a common node, x _j =1 represents the cluster head node. Each butterfly corresponds to a solution of cluster head distribution in the wireless sensor network, and corresponds to a determined clustering mode. Each solution, namely each butterfly, can calculate the stimulus intensity value I and the fragrance concentration value F determined by the butterfly according to the formula (3-17), then sequence according to the fragrance concentration value to obtain the current optimal solution, and update the positions of all the butterflies according to the moving formulas (3-10) and (3-11), namely change the element values of butterfly vectors of the non-optimal solution, thereby completing one iteration of the butterfly optimization algorithm. Repeating the steps, entering the next iteration until the maximum iteration times are reached, and outputting the optimal solution to obtain the selection distribution condition of the optimal cluster head.

1.5 data Transmission phase

The stage imitates LEACH protocol, after each round of optimal cluster head distribution selection is completed, common nodes are clustered nearby, and clustering is completed. Entering a data transmission stage, sending data to a cluster head by a common node in each cluster, wherein the size of the data is the transmission flow (note that the predicted flow is not the actual flow here) in a simulation data set real_tra [40,500], and solving the energy consumption sent by each common node according to an energy consumption formula (3-2). And then the cluster head node receives and fuses the data, and the energy consumption of the cluster head node is calculated according to the energy consumption formulas (3-4) and (3-5). And finally, the cluster head node transmits data to the base station, and the transmission energy consumption of the cluster head node is calculated according to an energy consumption formula (3-2). And after the optimal cluster head distribution selection of one round is finished, updating the residual energy of all the nodes, and entering the next round of simulation until all the nodes die.

From the software aspect, the present application further provides a device for performing sensor node clustering in a wireless sensor network in all or part of the method for clustering sensor nodes in the wireless sensor network, referring to fig. 2, where the device for clustering sensor nodes in the wireless sensor network specifically includes the following contents:

The energy consumption prediction module 10 is configured to obtain, in the wireless sensor network, overall communication prediction energy consumption of the sensor node set in a preset future time period based on the communication traffic data set of each sensor node by using a traffic prediction means; wherein the set of sensor nodes includes each of the sensor nodes;

the fitness function construction module 20 is configured to construct a fitness function for selecting an optimal cluster head according to the overall communication predicted energy consumption;

and the optimal cluster head distribution screening module 30 is used for obtaining optimal cluster head distribution data corresponding to the sensor node set according to the fitness function and a butterfly optimization algorithm.

The embodiment of the sensor node clustering device in the wireless sensor network provided by the application can be particularly used for executing the processing flow of the embodiment of the sensor node clustering method in the wireless sensor network in the embodiment, and the functions of the embodiment of the sensor node clustering device in the wireless sensor network are not repeated herein, and can be referred to the detailed description of the embodiment of the sensor node clustering method in the wireless sensor network.

The application provides a sensor node clustering device in a wireless sensor network, and the method executed by the device comprises the following steps: adopting a flow prediction means, and obtaining overall communication prediction energy consumption of the sensor node set in a preset future time period based on communication flow data sets of all the sensor nodes in the wireless sensor network; wherein the set of sensor nodes includes each of the sensor nodes; constructing an adaptability function for selecting an optimal cluster head according to the overall communication prediction energy consumption; and obtaining optimal cluster head distribution data corresponding to the sensor node set according to the fitness function and a butterfly optimization algorithm. The method and the device can select the optimal cluster head based on the predicted energy consumption, so that the rationality of selecting the cluster head node can be effectively improved, and the communication energy consumption in the wireless sensor network can be effectively reduced.

The embodiment of the application also provides an electronic device, such as a central server, which may include a processor, a memory, a receiver and a transmitter, where the processor is configured to execute the sensor node clustering method in the wireless sensor network mentioned in the foregoing embodiment, and the processor and the memory may be connected by a bus or other manners, for example, through a bus connection. The receiver may be connected to the processor, memory, by wire or wirelessly.

The processor may be a central processing unit (Central Processing Unit, CPU). The processor may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), field-Programmable gate arrays (FPGA) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination thereof.

The memory is used as a non-transitory computer readable storage medium and can be used for storing non-transitory software programs, non-transitory computer executable programs and modules, such as program instructions/modules corresponding to a sensor node clustering method in a wireless sensor network in an embodiment of the application. The processor executes various functional applications and data processing of the processor by running non-transitory software programs, instructions and modules stored in the memory, that is, the sensor node clustering method in the wireless sensor network in the above method embodiment is implemented.

The memory may include a memory program area and a memory data area, wherein the memory program area may store an operating system, at least one application program required for a function; the storage data area may store data created by the processor, etc. In addition, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory may optionally include memory located remotely from the processor, the remote memory being connectable to the processor through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The one or more modules are stored in the memory that, when executed by the processor, perform the sensor node clustering method in the wireless sensor network of the embodiments.

In some embodiments of the present application, a user equipment may include a processor, a memory, and a transceiver unit, which may include a receiver and a transmitter, the processor, the memory, the receiver, and the transmitter may be connected by a bus system, the memory being configured to store computer instructions, the processor being configured to execute the computer instructions stored in the memory to control the transceiver unit to transmit and receive signals.

As an implementation manner, the functions of the receiver and the transmitter in the present application may be considered to be implemented by a transceiver circuit or a dedicated chip for transceiver, and the processor may be considered to be implemented by a dedicated processing chip, a processing circuit or a general-purpose chip.

As another implementation manner, a manner of using a general-purpose computer may be considered to implement the server provided by the embodiment of the present application. I.e. program code for implementing the functions of the processor, the receiver and the transmitter are stored in the memory, and the general purpose processor implements the functions of the processor, the receiver and the transmitter by executing the code in the memory.

The embodiment of the application also provides a computer readable storage medium, on which a computer program is stored, which when being executed by a processor, is used for realizing the steps of the sensor node clustering method in the wireless sensor network. The computer readable storage medium may be a tangible storage medium such as Random Access Memory (RAM), memory, read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, floppy disks, hard disk, a removable memory disk, a CD-ROM, or any other form of storage medium known in the art.

Those of ordinary skill in the art will appreciate that the various illustrative components, systems, and methods described in connection with the embodiments disclosed herein can be implemented as hardware, software, or a combination of both. The particular implementation is hardware or software dependent on the specific application of the solution and the design constraints. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application. When implemented in hardware, it may be, for example, an electronic circuit, an Application Specific Integrated Circuit (ASIC), suitable firmware, a plug-in, a function card, or the like. When implemented in software, the elements of the application are the programs or code segments used to perform the required tasks. The program or code segments may be stored in a machine readable medium or transmitted over transmission media or communication links by a data signal carried in a carrier wave.

It should be understood that the application is not limited to the particular arrangements and instrumentality described above and shown in the drawings. For the sake of brevity, a detailed description of known methods is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method processes of the present application are not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between steps, after appreciating the spirit of the present application.

In this disclosure, features that are described and/or illustrated with respect to one embodiment may be used in the same way or in a similar way in one or more other embodiments and/or in combination with or instead of the features of the other embodiments.

The above description is only of the preferred embodiments of the present application and is not intended to limit the present application, and various modifications and variations can be made to the embodiments of the present application by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. The sensor node clustering method in the wireless sensor network is characterized by comprising the following steps of:

2. The method for clustering sensor nodes in a wireless sensor network according to claim 1, wherein the means for using traffic prediction obtains, in the wireless sensor network, overall communication prediction energy consumption of the sensor node set in a preset future time period based on communication traffic data sets of the respective sensor nodes, and the method comprises:

3. The method for clustering sensor nodes in a wireless sensor network according to claim 2, wherein said obtaining said overall communication predicted energy consumption of said set of sensor nodes from each of said predicted traffic sets comprises:

4. A method of clustering sensor nodes in a wireless sensor network according to claim 3, further comprising, prior to said constructing an fitness function for selecting an optimal cluster head based on said overall communication predicted energy consumption:

5. The method for clustering sensor nodes in a wireless sensor network according to claim 4, wherein the constructing an fitness function for selecting an optimal cluster head according to the overall communication predicted energy consumption comprises:

6. The method for clustering sensor nodes in a wireless sensor network according to claim 1, wherein the obtaining the optimal cluster head distribution data corresponding to the sensor node set according to the fitness function and a butterfly optimization algorithm includes:

7. The method for clustering sensor nodes in a wireless sensor network according to claim 6, wherein the optimizing step comprises:

and updating the sensing mode.

8. A sensor node clustering device in a wireless sensor network, comprising:

9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the sensor node clustering method in a wireless sensor network according to any one of claims 1 to 7 when the computer program is executed by the processor.

10. A computer readable storage medium having stored thereon a computer program, which when executed by a processor implements a method of clustering sensor nodes in a wireless sensor network according to any one of claims 1 to 7.