CN107949029A - A kind of leader cluster node method for routing, device and medium based on wireless sense network - Google Patents
A kind of leader cluster node method for routing, device and medium based on wireless sense network Download PDFInfo
- Publication number
- CN107949029A CN107949029A CN201711407162.7A CN201711407162A CN107949029A CN 107949029 A CN107949029 A CN 107949029A CN 201711407162 A CN201711407162 A CN 201711407162A CN 107949029 A CN107949029 A CN 107949029A
- Authority
- CN
- China
- Prior art keywords
- power consumption
- leader cluster
- cluster node
- gradient
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W40/00—Communication routing or communication path finding
- H04W40/02—Communication route or path selection, e.g. power-based or shortest path routing
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W40/00—Communication routing or communication path finding
- H04W40/02—Communication route or path selection, e.g. power-based or shortest path routing
- H04W40/04—Communication route or path selection, e.g. power-based or shortest path routing based on wireless node resources
- H04W40/10—Communication route or path selection, e.g. power-based or shortest path routing based on wireless node resources based on available power or energy
Abstract
The invention discloses a kind of leader cluster node method for routing, device and medium based on wireless sense network, the step of this method, includes：Obtain each characteristic value collection of target leader cluster node；According to gradient descent algorithm, the characteristic value in each characteristic value collection and power consumption gradient parameter are fitted to power consumption anticipation function；Power consumption threshold value is set, and according to power consumption anticipation function and power consumption threshold value generational loss function to weigh the difference between the value of power consumption anticipation function and power consumption threshold value；Calculate when the value of loss function is minimum, the parameter value of power consumption gradient parameter；The interative computation of parameter value is carried out on the basis of default power consumption gradient, to obtain result power consumption gradient, and is directed toward result power consumption gradient as the route of target leader cluster node.This method improves leader cluster node to the efficiency being route between aggregation node.In addition, the present invention also provides a kind of leader cluster node route device based on wireless sense network and medium, beneficial effect are as described above.
Description
Technical field
The present invention relates to wireless sense network field, more particularly to a kind of leader cluster node route side based on wireless sense network
Method, device and medium.
Background technology
Wireless sense network is the wireless network that one group of sensor node is formed in a manner of adaptive, self-organizing.It is wireless to pass
Feel collection, processing and transmission that net can be used for particular surroundings live signal.Wireless sense network be a kind of brand-new acquisition of information and
Treatment technology, just by more and more extensive use in actual life.Wireless sense network is by thousands of sensor node
Self-organizing is formed, and sensor node is powered by battery, and the behaviour such as data acquisition, fusion and transmission is carried out in sensor node
As when be required for consumption the energy content of battery, after thousands of sensor node is disposed, be often not easy to replace, it is therefore desirable to protect
Card is under the limited electricity of sensor node, using the teaching of the invention it is possible to provide the service of longer time.
Current wireless sense network has a hierarchical structure at work, the sensor node in hierarchical wireless sense network point
For three kinds, i.e. member node, leader cluster node, aggregation node.There are N number of member node, member node under each leader cluster node
By the data transfer collected to leader cluster node, and then leader cluster node handles data, is merged and is transferred to aggregation node.
But under existing large-scale wireless sense network detection environment, the distance between most of leader cluster node and aggregation node compared with
Far, can not direct communication, it is therefore desirable to which the other leader cluster nodes passed through between target leader cluster node and aggregation node carry out data
Transmission finally to realize the communication of target leader cluster node and aggregation node.Since the data transfer between sensor node is wireless
Necessary action in sensing network, and corresponding energy expenditure can be brought, thus leader cluster node to aggregation node it
Between carry out the energy that rationally efficient route opposite can save leader cluster node and be consumed, and then extend the whole of wireless sensor network
The life cycle of body.
As can be seen here, there is provided a kind of leader cluster node method for routing based on wireless sense network, to improve leader cluster node to remittance
The efficiency being route between poly- node, and then the opposite energy expenditure for saving leader cluster node entirety, it is overall to extend wireless sensor network
Life cycle, be those skilled in the art's urgent problem to be solved.
The content of the invention
The object of the present invention is to provide a kind of leader cluster node method for routing, device and medium based on wireless sense network, with
Leader cluster node is improved to the efficiency being route between aggregation node, and then the opposite energy expenditure for saving leader cluster node entirety, extend
The life cycle of wireless sensor network entirety.
In order to solve the above technical problems, the present invention provides a kind of leader cluster node method for routing based on wireless sense network, bag
Include：
Obtain each characteristic value collection of target leader cluster node；Wherein, each characteristic value collection corresponds to target leader cluster node
Data transfer scene, and the destinations traffic path between target leader cluster node and aggregation node is included at least in characteristic value collection
Distance；
According to gradient descent algorithm, the characteristic value in each characteristic value collection and power consumption gradient parameter are fitted to power consumption prediction
Function；
Power consumption threshold value is set, and letter is predicted to weigh power consumption according to power consumption anticipation function and power consumption threshold value generational loss function
Difference between several values and power consumption threshold value；
Calculate when the value of loss function is minimum, the parameter value of power consumption gradient parameter；
The interative computation of parameter value is carried out on the basis of default power consumption gradient, to obtain result power consumption gradient, and will knot
Fruit power consumption gradient is directed toward as the route of target leader cluster node.
Preferably, the dump energy of target leader cluster node and the data of target leader cluster node are further included in characteristic value collection
The traffic.
Preferably, before carrying out the interative computation of parameter value on the basis of default power consumption gradient, this method further comprises：
Constrained parameters are set；
Correspondingly, the interative computation that parameter value is carried out on the basis of default power consumption gradient is specially：
The amplitude of interative computation is controlled by constrained parameters, to be iterated computing.
Preferably, the value of constrained parameters is specially 0.0001.
Preferably, after result power consumption gradient is obtained, this method further comprises：
Result power consumption gradient is write into daily record.
Preferably, this method further comprises：
The result power consumption gradient of each target leader cluster node is obtained to generate overall routing table.
In addition, the present invention also provides a kind of leader cluster node route device based on wireless sense network, including：
Characteristic value acquisition module, for obtaining each characteristic value collection of target leader cluster node；
Anticipation function generation module, for according to gradient descent algorithm, by the characteristic value and power consumption in each characteristic value collection
Gradient parameter is fitted to power consumption anticipation function；
Loss function generation module, is damaged for setting power consumption threshold value, and according to power consumption anticipation function and the generation of power consumption threshold value
Function is lost to weigh the difference between the value of power consumption anticipation function and power consumption threshold value；
Parameter value calculation module, for calculating when the value of loss function is minimum, the parameter value of power consumption gradient parameter；
Interative computation module, for carrying out the interative computation of parameter value on the basis of default power consumption gradient, to be tied
Fruit power consumption gradient, and be directed toward result power consumption gradient as the route of target leader cluster node.
Preferably, which further comprises：
Setup module is constrained, for setting constrained parameters.
In addition, the present invention also provides a kind of leader cluster node route device based on wireless sense network, including：
Memory, for storing computer program；
Processor, the leader cluster node route side based on wireless sense network described above is realized during for performing computer program
The step of method.
In addition, the present invention also provides a kind of computer-readable recording medium, meter is stored with computer-readable recording medium
Calculation machine program, realizes the leader cluster node method for routing based on wireless sense network described above when computer program is executed by processor
The step of.
Leader cluster node method for routing provided by the present invention based on wireless sense network, first obtains each of target leader cluster node
Characteristic value collection, and the power consumption prediction for including characteristic value and power consumption gradient parameter is built according to the thought of gradient descent algorithm
Function, for predicting the data transmission power consumption of the target leader cluster node under different scenes, and then by power consumption anticipation function and in advance
If power consumption threshold value form loss function, and calculate when the value of loss function is minimum, i.e. the value of power consumption anticipation function meets pre-
If during the requirement of power consumption threshold value, the parameter value of power consumption gradient parameter, and then every time when parameter value is calculated, i.e., in default power consumption
The interative computation of the parameter value is carried out on the basis of gradient, and then obtains result power consumption gradient using the road as target leader cluster node
By being directed toward.Due in the method, the transmitting scene for the target leader cluster node that each characteristic value collection correspond to, it is seen that we
Method makes leader cluster node carry out the integrated learning of scene equivalent to the simulation for carrying out substantial amounts of data transfer scene to leader cluster node,
Learn all to carry out power consumption gradient according to the thought of gradient descent algorithm more efficient convergence each time, so that power consumption gradient is more
Add the data transfer scene for adapting to current composite, and then be directed toward result power consumption gradient as the route of target leader cluster node, energy
Enough reasonable and efficient progress route transmission of the target leader cluster node to aggregation node.As it can be seen that this method improves leader cluster node
To the efficiency being route between aggregation node, and then the energy expenditure of leader cluster node entirety is provide further relative savings, extend wireless sense network
The life cycle of network entirety.In addition, the present invention also provides a kind of leader cluster node route device and medium based on wireless sense network,
Beneficial effect is as described above.
Brief description of the drawings
In order to illustrate the embodiments of the present invention more clearly, attached drawing needed in the embodiment will be done simply below
Introduce, it should be apparent that, drawings in the following description are only some embodiments of the present invention, for ordinary skill people
For member, without creative efforts, other attached drawings can also be obtained according to these attached drawings.
Fig. 1 is a kind of flow chart of the leader cluster node method for routing based on wireless sense network provided in an embodiment of the present invention；
Fig. 2 is a kind of leader cluster node route device structure chart based on wireless sense network provided in an embodiment of the present invention.
Embodiment
Below in conjunction with the attached drawing in the embodiment of the present invention, the technical solution in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment only part of the embodiment of the present invention, rather than whole embodiments.Based on this
Embodiment in invention, without making creative work, what is obtained is every other by those of ordinary skill in the art
Embodiment, belongs to the scope of the present invention.
The core of the present invention is to provide a kind of leader cluster node method for routing based on wireless sense network, improves leader cluster node and arrives
The efficiency being route between aggregation node, and then the opposite energy expenditure for saving leader cluster node entirety, it is whole to extend wireless sensor network
The life cycle of body.Another core of the present invention is to provide a kind of leader cluster node route device based on wireless sense network and Jie
Matter.
In order to make those skilled in the art more fully understand the present invention program, with reference to the accompanying drawings and detailed description
The present invention is described in further detail.
Embodiment one
Fig. 1 is a kind of flow chart of the leader cluster node method for routing based on wireless sense network provided in an embodiment of the present invention.
Please refer to Fig.1, the specific steps of the leader cluster node method for routing based on wireless sense network include：
Step S10：Obtain each characteristic value collection of target leader cluster node.
Wherein, each characteristic value collection corresponds to the data transfer scene of target leader cluster node, and in characteristic value collection at least
Include the distance in the destinations traffic path between target leader cluster node and aggregation node.
It should be noted that the characteristic value in characteristic value collection in this step can represent that target leader cluster node leads to
State during letter in a certain respect, and then each characteristic value collection should correspond to a kind of target leader cluster node and other leader cluster nodes
Operative scenario during into row data communication, should in characteristic value collection since operative scenario is related to the communications of data
This includes at least the distance in a certain destinations traffic path between target leader cluster node and aggregation node.And then multiple characteristic value collection
The communication scenes that can be just provided to target leader cluster node under different situations are closed, in order to which target leader cluster node is to characteristic value collection
Carry out comprehensive analysis and study.
Step S11：According to gradient descent algorithm, the characteristic value in each characteristic value collection and power consumption gradient parameter are fitted to
Power consumption anticipation function.
Constructed power consumption anticipation function is the prediction for carrying out power consumption to target leader cluster node in this step.Due to
The power consumption situation of target leader cluster node to it into row data communication when operative scenario it is related, and characteristic value collection can characterize mesh
The operative scenario of leader cluster node is marked, therefore there is the characteristic value in characteristic value collection in power consumption anticipation function；In addition, power consumption gradient
Parameter is a unknown variable in power consumption anticipation function, for reflecting target leader cluster node under different operating scene to remittance
Gradient situation during poly- node-node transmission data.Since the power consumption of target leader cluster node characterizes each side factor in by characteristic value collection
Characteristic value influence, it is therefore desirable to the characteristic value in each characteristic value collection and power consumption gradient parameter are fitted to power consumption prediction letter
Number.It is emphasized that power consumption gradient parameter is in the nature vector, therefore it is with directive property.
Step S12：Power consumption threshold value is set, and according to power consumption anticipation function and power consumption threshold value generational loss function to weigh work(
Consume the difference between the value and power consumption threshold value of anticipation function.
Set power consumption threshold value is maximum of the user according to the integrality defined of target leader cluster node in this step
Power consumption number, what is shown is that to send data to the energy that aggregation node is consumed be at most more for leader cluster node in certain region
It is few, it is impossible to which that, more than the power consumption threshold value, no it will cause the network energy being not necessarily to waste.In addition, pass through power consumption anticipation function
The loss function expression generated with power consumption threshold value is existing difference between the value of power consumption anticipation function and power consumption threshold value.
Step S13：Calculate when the value of loss function is minimum, the parameter value of power consumption gradient parameter.
It should be noted that the basic goal of gradient descent algorithm, which is the mode declined along gradient, calculates function most
Small value, and then obtain the value of the gradient in the case where function is minimum value.Therefore the purpose of this step is that counting loss function takes
When difference between minimum value, the i.e. value of power consumption anticipation function and power consumption threshold value is minimum, the parameter value of power consumption gradient parameter.This step
The parameter value of power consumption gradient parameter obtained in rapid can meet the difference of the value of power consumption anticipation function and power consumption threshold value most
It is small.
Step S14：The interative computation of parameter value is carried out on the basis of default power consumption gradient, to obtain result power consumption ladder
Degree, and be directed toward result power consumption gradient as the route of target leader cluster node.
The only current signature value set feelings reflected by the parameter value of the power consumption gradient parameter obtained in above-mentioned steps
Corresponding power consumption gradient under condition, in order to make power consumption gradient constantly accurately and rationally, it is necessary to which target leader cluster node is analyzed and learnt
More data transfer scenes, and then the parameter value of the power consumption gradient under more scenes is obtained, and then in default power consumption gradient
On the basis of carry out parameter value interative computation, each time interative computation be equivalent to the once renewal to power consumption gradient, every time more
New capital makes power consumption gradient more accurate, and then makes the route direction of target leader cluster node more reasonable, finally obtains result power consumption
Gradient.It should be noted that result power consumption gradient is opposite concept, the quantity determination result of current whole characteristic value collection
The degree of convergence of power consumption gradient, but if the quantity of increase characteristic value collection is to continue the iteration of power consumption gradient, then result
It is accurate that power consumption gradient still may proceed to convergence.
Leader cluster node method for routing provided by the present invention based on wireless sense network, first obtains each of target leader cluster node
Characteristic value collection, and the power consumption prediction for including characteristic value and power consumption gradient parameter is built according to the thought of gradient descent algorithm
Function, for predicting the data transmission power consumption of the target leader cluster node under different scenes, and then by power consumption anticipation function and in advance
If power consumption threshold value form loss function, and calculate when the value of loss function is minimum, i.e. the value of power consumption anticipation function meets pre-
If during the requirement of power consumption threshold value, the parameter value of power consumption gradient parameter, and then every time when parameter value is calculated, i.e., in default power consumption
The interative computation of the parameter value is carried out on the basis of gradient, and then obtains result power consumption gradient using the road as target leader cluster node
By being directed toward.Due in the method, the transmitting scene for the target leader cluster node that each characteristic value collection correspond to, it is seen that we
Method makes leader cluster node carry out the integrated learning of scene equivalent to the simulation for carrying out substantial amounts of data transfer scene to leader cluster node,
Learn all to carry out power consumption gradient according to the thought of gradient descent algorithm more efficient convergence each time, so that power consumption gradient is more
Add the data transfer scene for adapting to current composite, and then be directed toward result power consumption gradient as the route of target leader cluster node, energy
Enough reasonable and efficient progress route transmission of the target leader cluster node to aggregation node.As it can be seen that this method improves leader cluster node
To the efficiency being route between aggregation node, and then the energy expenditure of leader cluster node entirety is provide further relative savings, extend wireless sense network
The life cycle of network entirety.
Embodiment two
On the basis of above-described embodiment, the present invention also provides following preferred embodiment.
As a preferred embodiment, the dump energy and mesh of target leader cluster node are further included in characteristic value collection
Mark the data traffic of leader cluster node.
It is understood that due to characteristic value collection characterization be target leader cluster node operative scenario, characteristic value
The characteristic value species operative scenario that more at most this feature value set is characterized in set is more careful.And target leader cluster node is surplus
An important factor for data traffic of complementary energy and target leader cluster node is the route communication of joint effect target leader cluster node, because
This using the dump energy of target leader cluster node and the data traffic of target leader cluster node as the element in characteristic value collection,
It can make the calculating of power consumption gradient in view of the factor of more various dimensions, and it is more careful and accurate.For example, in a kind of scene
Under, target leader cluster node can there are two other leader cluster nodes within transmission range, but the two other leader cluster nodes
The distance of respective distance objective leader cluster node differs, and the other leader cluster nodes distance convergence of distance objective leader cluster node farther out
Node is closer to but target leader cluster node sends data to power consumption bigger required for other leader cluster nodes farther out, special at this time
The dump energy of target leader cluster node and data traffic can serve as the restraining factors on different dimensions in value indicative set,
With the evaluation integrated and choose most rational routed path.
In addition, as a preferred embodiment, the iteration fortune of parameter value is carried out on the basis of default power consumption gradient
Before calculation, this method further comprises：
Constrained parameters are set；
Correspondingly, the interative computation that parameter value is carried out on the basis of default power consumption gradient is specially：
The amplitude of interative computation is controlled by constrained parameters, to be iterated computing.
It should be noted that the effect for setting constrained parameters is when being iterated computing, the work(to participating in interative computation
The parameter value of consumption gradient parameter reduces corresponding multiple to achieve the purpose that the amplitude for controlling interative computation.Constrained parameters be in order to
Prevent that constrained parameters are to each due to the decline for differing greatly and causing overall operational precision between a few features value set
The parameter value of the power consumption gradient parameter of iteration is reduced, with caused resultant error caused by opposite reduction the above situation.Separately
Outside, for constrained parameters value setting should according to specific need of the otherness between each characteristic value collection and user and
It is fixed, it is not specifically limited herein.
On the basis of the above embodiment, as a preferred embodiment, the value of constrained parameters is specially
0.0001。
It is opposite since the value of constrained parameters is arranged to the 0.0001 diminution degree for the parameter value of power consumption gradient parameter
It is larger, therefore the error for reducing interative computation has the function that universality.User can be using 0.0001 as a constraint
The setting basis of parameter, and the value of constrained parameters is increased and decreased according to specific requirements under basis herein, do not do specific limit herein
It is fixed.
In addition, as a preferred embodiment, after result power consumption gradient is obtained, this method further comprises：
Result power consumption gradient is write into daily record.
It is understood that after result power consumption gradient is write daily record, user can be straight if it need to obtain result power consumption gradient
It is connected in daily record and is obtained, without being iterated computing again, realizes the effect of " write-once, repeatedly obtains ".In addition,
, then can be directly by reading result power consumption gradient in daily record when needing further interative computation to continue to restrain power consumption gradient
It is further accurate to be carried out to power consumption gradient using as new default power consumption gradient.It can be seen that result power consumption gradient is write into day
Will is provided convenience for its subsequent use.
In addition, as a preferred embodiment, this method further comprises：
The result power consumption gradient of each target leader cluster node is obtained to generate overall routing table.
It is understood that since each target leader cluster node can only be concerned about the next-hop cluster head node of itself
Route direction, thus be directed to for user only in a certain target leader cluster node acquisition route direction can not get it is whole
The route planning situation of body, and the result power consumption gradient of each target leader cluster node is obtained to generate overall routing table, Neng Gouyou
Help the route planning situation that user understands wireless sensor network entirety leader cluster node.
Below with regard to a certain embodiment of this programme, it is further detailed in the form of function and formula：
Power consumption anticipation function is h (x)=hθ(x)=θ0x0+θ1x1+θ2x2, wherein θ expression power consumption gradient parameters, Χ expressions
The characteristic value collection of input, x0Represent the dump energy of target leader cluster node, x1Represent between target leader cluster node and aggregation node
Destinations traffic path distance, x2Represent the data traffic of target leader cluster node, corresponding θ0、θ1、θ2Respectively x0、x1、
x2Corresponding power consumption gradient parameter component；
Loss function isWherein j represents the current quantity of characteristic value collection, and m is represented
The total quantity of characteristic value collection, y represent power consumption threshold value.
The formula of interative computation isWherein i represents the number of iteration, θiAfter representing iteration renewal
Power consumption gradient parameter,Represent in J (θ) to θiLocal derviation is sought, to calculate when the value of loss function is minimum, power consumption
The parameter value of gradient parameter.
Above-mentioned formula and function are only opened up only as the operation content under a kind of specific embodiment of the present invention
Show explanation, each parameter and formula can be changed accordingly in the case where following present subject matter and being intended to,
This is not specifically limited.
Embodiment three
Hereinbefore it is described in detail for the embodiment of the leader cluster node method for routing based on wireless sense network,
The present invention also provides a kind of leader cluster node route device based on wireless sense network corresponding with this method, due to device part
Embodiment is corresponded with the embodiment of method part, therefore the embodiment of device part refers to the embodiment of method part
Description, wouldn't repeat here.
Fig. 2 is a kind of leader cluster node route device structure chart based on wireless sense network provided in an embodiment of the present invention.This
The leader cluster node route device based on wireless sense network that inventive embodiments provide, specifically includes：
Characteristic value acquisition module 10, for obtaining each characteristic value collection of target leader cluster node.
Anticipation function generation module 11, for according to gradient descent algorithm, by the characteristic value and work(in each characteristic value collection
Consumption gradient parameter is fitted to power consumption anticipation function.
Loss function generation module 12, generates for setting power consumption threshold value, and according to power consumption anticipation function and power consumption threshold value
Loss function is to weigh the difference between the value of power consumption anticipation function and power consumption threshold value.
Parameter value calculation module 13, for calculating when the value of loss function is minimum, the parameter value of power consumption gradient parameter.
Interative computation module 14, for carrying out the interative computation of parameter value on the basis of default power consumption gradient, to obtain
As a result power consumption gradient, and be directed toward result power consumption gradient as the route of target leader cluster node.
Leader cluster node route device provided by the present invention based on wireless sense network, first obtains each of target leader cluster node
Characteristic value collection, and the power consumption prediction for including characteristic value and power consumption gradient parameter is built according to the thought of gradient descent algorithm
Function, for predicting the data transmission power consumption of the target leader cluster node under different scenes, and then by power consumption anticipation function and in advance
If power consumption threshold value form loss function, and calculate when the value of loss function is minimum, i.e. the value of power consumption anticipation function meets pre-
If during the requirement of power consumption threshold value, the parameter value of power consumption gradient parameter, and then every time when parameter value is calculated, i.e., in default power consumption
The interative computation of the parameter value is carried out on the basis of gradient, and then obtains result power consumption gradient using the road as target leader cluster node
By being directed toward.Due in the present apparatus, the transmitting scene for the target leader cluster node that each characteristic value collection correspond to, it is seen that this dress
Put equivalent to the simulation that substantial amounts of data transfer scene is carried out to leader cluster node, leader cluster node is carried out the integrated learning of scene,
Learn all to carry out power consumption gradient according to the thought of gradient descent algorithm more efficient convergence each time, so that power consumption gradient is more
Add the data transfer scene for adapting to current composite, and then be directed toward result power consumption gradient as the route of target leader cluster node, energy
Enough reasonable and efficient progress route transmission of the target leader cluster node to aggregation node.As it can be seen that the present apparatus improves leader cluster node
To the efficiency being route between aggregation node, and then the energy expenditure of leader cluster node entirety is provide further relative savings, extend wireless sense network
The life cycle of network entirety.
On the basis of embodiment three, which further includes：
Setup module is constrained, for setting constrained parameters.
Example IV
The present invention also provides a kind of leader cluster node route device based on wireless sense network, including：
Memory, for storing computer program；
Processor, the leader cluster node route side based on wireless sense network described above is realized during for performing computer program
The step of method.
Leader cluster node route device provided by the present invention based on wireless sense network, first obtains each of target leader cluster node
Characteristic value collection, and the power consumption prediction for including characteristic value and power consumption gradient parameter is built according to the thought of gradient descent algorithm
Function, for predicting the data transmission power consumption of the target leader cluster node under different scenes, and then by power consumption anticipation function and in advance
If power consumption threshold value form loss function, and calculate when the value of loss function is minimum, i.e. the value of power consumption anticipation function meets pre-
If during the requirement of power consumption threshold value, the parameter value of power consumption gradient parameter, and then every time when parameter value is calculated, i.e., in default power consumption
The interative computation of the parameter value is carried out on the basis of gradient, and then obtains result power consumption gradient using the road as target leader cluster node
By being directed toward.Due in the present apparatus, the transmitting scene for the target leader cluster node that each characteristic value collection correspond to, it is seen that this dress
Put equivalent to the simulation that substantial amounts of data transfer scene is carried out to leader cluster node, leader cluster node is carried out the integrated learning of scene,
Learn all to carry out power consumption gradient according to the thought of gradient descent algorithm more efficient convergence each time, so that power consumption gradient is more
Add the data transfer scene for adapting to current composite, and then be directed toward result power consumption gradient as the route of target leader cluster node, energy
Enough reasonable and efficient progress route transmission of the target leader cluster node to aggregation node.As it can be seen that the present apparatus improves leader cluster node
To the efficiency being route between aggregation node, and then the energy expenditure of leader cluster node entirety is provide further relative savings, extend wireless sense network
The life cycle of network entirety.
The present invention also provides a kind of computer-readable recording medium, computer journey is stored with computer-readable recording medium
Sequence, realizes the step of the leader cluster node method for routing based on wireless sense network described above when computer program is executed by processor
Suddenly.
The computer-readable recording medium of leader cluster node route provided by the present invention based on wireless sense network, first obtains
Each characteristic value collection of target leader cluster node, and characteristic value and power consumption ladder are included according to the thought of gradient descent algorithm structure
The power consumption anticipation function of parameter is spent, for predicting the data transmission power consumption of the target leader cluster node under different scenes, Jin Eryou
Power consumption anticipation function forms loss function with default power consumption threshold value, and calculates when the value of loss function is minimum, i.e., power consumption is pre-
When the value of survey function meets the requirement of default power consumption threshold value, the parameter value of power consumption gradient parameter, and then parameter is calculated every time
During value, i.e., the interative computation of the parameter value is carried out on the basis of default power consumption gradient, and then obtain result power consumption gradient to make
It is directed toward for the route of target leader cluster node.Since in this computer-readable recording medium, each characteristic value collection correspond to
Target leader cluster node transmitting scene, it is seen that this computer-readable recording medium is largely counted equivalent to leader cluster node
According to the simulation of transmitting scene, leader cluster node is carried out the integrated learning of scene, learn each time all according to gradient descent algorithm
Thought carries out power consumption gradient more efficient convergence, so that power consumption gradient more adapts to the data transfer scene of current composite,
And then be directed toward result power consumption gradient as the route of target leader cluster node, it can rationally and efficiently carry out target leader cluster node
To the route transmission of aggregation node.As it can be seen that this computer-readable recording medium improves leader cluster node to road between aggregation node
By efficiency, and then provide further relative savings the energy expenditure of leader cluster node entirety, extend the life cycle of wireless sensor network entirety.
Above to a kind of leader cluster node method for routing, device and medium based on wireless sense network provided by the present invention into
Go and be discussed in detail.Each embodiment is described by the way of progressive in specification, what each embodiment stressed be with
The difference of other embodiment, between each embodiment identical similar portion mutually referring to.For disclosed in embodiment
For device, since it is corresponded to the methods disclosed in the examples, so description is fairly simple, related part is referring to method portion
Defend oneself bright.It should be pointed out that for those skilled in the art, the premise of the principle of the invention is not being departed from
Under, some improvement and modification can also be carried out to the present invention, these are improved and modification also falls into the protection of the claims in the present invention
In the range of.
It should also be noted that, in the present specification, relational terms such as first and second and the like be used merely to by
One entity or operation are distinguished with another entity or operation, without necessarily requiring or implying these entities or operation
Between there are any actual relationship or order.Moreover, term " comprising ", "comprising" or its any other variant meaning
Covering non-exclusive inclusion, so that process, method, article or equipment including a series of elements not only include that
A little key elements, but also including other elements that are not explicitly listed, or further include for this process, method, article or
The intrinsic key element of equipment.In the absence of more restrictions, the key element limited by sentence "including a ...", is not arranged
Except also there are other identical element in the process, method, article or apparatus that includes the element.
Claims (10)
- A kind of 1. leader cluster node method for routing based on wireless sense network, it is characterised in that including：Obtain each characteristic value collection of target leader cluster node；Wherein, each characteristic value collection corresponds to the target cluster head section The data transfer scene of point, and the mesh between the target leader cluster node and aggregation node is included at least in the characteristic value collection Mark the distance of communication path；According to gradient descent algorithm, the characteristic value in each characteristic value collection and power consumption gradient parameter are fitted to power consumption prediction Function；Power consumption threshold value is set, and according to the power consumption anticipation function and the power consumption threshold value generational loss function to weigh the work( Consume the difference between the value of anticipation function and the power consumption threshold value；Calculate when the value of the loss function is minimum, the parameter value of the power consumption gradient parameter；The interative computation of the parameter value is carried out on the basis of default power consumption gradient, to obtain result power consumption gradient, and by institute The route that result power consumption gradient is stated as the target leader cluster node is directed toward.
- 2. according to the method described in claim 1, it is characterized in that, the target cluster head section is further included in the characteristic value collection The dump energy of point and the data traffic of the target leader cluster node.
- 3. according to the method described in claim 1, it is characterized in that, described carry out the ginseng on the basis of default power consumption gradient Before the interative computation of numerical value, this method further comprises：Constrained parameters are set；Correspondingly, it is described the parameter value is carried out on the basis of default power consumption gradient interative computation be specially：The amplitude of the interative computation is controlled by the constrained parameters, to carry out the interative computation.
- 4. according to the method described in claim 3, it is characterized in that, the value of the constrained parameters is specially 0.0001.
- 5. according to the method described in claim 1, it is characterized in that, it is described obtain result power consumption gradient after, this method is into one Step includes：The result power consumption gradient is write into daily record.
- 6. according to the method described in claim 1-5 any one, it is characterised in that this method further comprises：The result power consumption gradient of each target leader cluster node is obtained to generate overall routing table.
- A kind of 7. leader cluster node route device based on wireless sense network, it is characterised in that including：Characteristic value acquisition module, for obtaining each characteristic value collection of target leader cluster node；Anticipation function generation module, for according to gradient descent algorithm, by the characteristic value and power consumption in each characteristic value collection Gradient parameter is fitted to power consumption anticipation function；Loss function generation module, gives birth to for setting power consumption threshold value, and according to the power consumption anticipation function and the power consumption threshold value Into loss function to weigh the difference between the value of the power consumption anticipation function and the power consumption threshold value；Parameter value calculation module, for calculating when the value of the loss function is minimum, the parameter value of the power consumption gradient parameter；Interative computation module, for carrying out the interative computation of the parameter value on the basis of default power consumption gradient, to be tied Fruit power consumption gradient, and be directed toward the result power consumption gradient as the route of the target leader cluster node.
- 8. device according to claim 7, it is characterised in that the device further comprises：Setup module is constrained, for setting constrained parameters.
- A kind of 9. leader cluster node route device based on wireless sense network, it is characterised in that including：Memory, for storing computer program；Processor, is realized during for performing the computer program as claim 1 to 6 any one of them is based on wireless sensing The step of leader cluster node method for routing of net.
- 10. a kind of computer-readable recording medium, it is characterised in that be stored with computer on the computer-readable recording medium Program, is realized when the computer program is executed by processor as claim 1 to 6 any one of them is based on wireless sense network Leader cluster node method for routing the step of.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711407162.7A CN107949029B (en) | 2017-12-22 | 2017-12-22 | Cluster head node routing method, device and medium based on wireless sensor network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711407162.7A CN107949029B (en) | 2017-12-22 | 2017-12-22 | Cluster head node routing method, device and medium based on wireless sensor network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107949029A true CN107949029A (en) | 2018-04-20 |
| CN107949029B CN107949029B (en) | 2021-03-16 |
Family
ID=61941826
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711407162.7A Expired - Fee Related CN107949029B (en) | 2017-12-22 | 2017-12-22 | Cluster head node routing method, device and medium based on wireless sensor network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107949029B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101394321A (en) * | 2008-09-25 | 2009-03-25 | 重庆邮电大学 | Non-uniform clustering method for cluster wireless sensor network based on energy balance |
| CN101729331A (en) * | 2008-10-28 | 2010-06-09 | 华为技术有限公司 | Clustering method and device, routing method and device of cluster head and base station |
| CN104080144A (en) * | 2014-07-18 | 2014-10-01 | 重庆邮电大学 | Gradient based energy-efficient uneven clustering data forwarding method |
| WO2014198120A1 (en) * | 2013-06-14 | 2014-12-18 | 电子科技大学 | Wireless mobile network data transmission method based on coverage maintenance and minimum spanning tree |
| CN104994554A (en) * | 2015-05-04 | 2015-10-21 | 四川大学 | Mobile assistance WSNs routing method based on unequal clustering |
-
2017
- 2017-12-22 CN CN201711407162.7A patent/CN107949029B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101394321A (en) * | 2008-09-25 | 2009-03-25 | 重庆邮电大学 | Non-uniform clustering method for cluster wireless sensor network based on energy balance |
| CN101729331A (en) * | 2008-10-28 | 2010-06-09 | 华为技术有限公司 | Clustering method and device, routing method and device of cluster head and base station |
| WO2014198120A1 (en) * | 2013-06-14 | 2014-12-18 | 电子科技大学 | Wireless mobile network data transmission method based on coverage maintenance and minimum spanning tree |
| CN104080144A (en) * | 2014-07-18 | 2014-10-01 | 重庆邮电大学 | Gradient based energy-efficient uneven clustering data forwarding method |
| CN104994554A (en) * | 2015-05-04 | 2015-10-21 | 四川大学 | Mobile assistance WSNs routing method based on unequal clustering |
Non-Patent Citations (5)
| Title |
|---|
| MOSTAFA MOFARREH-BONAB ET AL.: "A Low Complexity and High Speed Gradient Descent Based Secure Localization in Wireless Sensor Networks", 《3RD INTERNATIONAL CONFERENCE ON COMPUTER AND KNOWLEDGE ENGINEERING (ICCKE 2013)》 * |
| SEBASTIAN RUDER: "An overview of gradient descent optimization algorithms", 《HTTPS://RUDER.IO/OPTIMIZING-GRADIENT-DESCENT/》 * |
| 孙宇鹭: "基于地理位置的WSN节能分簇路由技术研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 左睿: "无线传感器网络分簇路由协议研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 张爱丽: "基于梯度的无线传感器网络路由算法的研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107949029B (en) | 2021-03-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107016175B (en) | It is applicable in the Automation Design method, apparatus and optimization method of neural network processor | |
| Li et al. | Data-driven evolutionary algorithm with perturbation-based ensemble surrogates | |
| CN107103113B (en) | The Automation Design method, apparatus and optimization method towards neural network processor | |
| Sudheer et al. | Radial basis function neural network for modeling rating curves | |
| Mora-Meliá et al. | Efficiency of evolutionary algorithms in water network pipe sizing | |
| CN105989408A (en) | A system and a method for mapping a neural network onto a neurosynaptic substrate | |
| CN103106535B (en) | Method for solving collaborative filtering recommendation data sparsity based on neural network | |
| CN107430704A (en) | Neural network algorithm is realized in nerve synapse substrate based on the metadata associated with neural network algorithm | |
| Fang et al. | Transfer learning across networks for collective classification | |
| Cui et al. | Big data over networks | |
| CN104798043A (en) | Data processing method and computer system | |
| CN109902203A (en) | The network representation learning method and device of random walk based on side | |
| CN102056183A (en) | Network flow prediction method and device based on cognitive network | |
| CN106327357A (en) | Load identification method based on improved probabilistic neural network | |
| CN108696453A (en) | The power telecom network SDN Business Streams cognitive method and system of lightweight | |
| CN111954209A (en) | Information processing method and device for improving security of wireless sensor node | |
| CN108985455A (en) | A kind of computer application neural net prediction method and system | |
| CN109299160A (en) | A kind of electric power CPS Safety Analysis Method excavated based on monitoring big data | |
| Ajitha et al. | Neural network model for software size estimation using use case point approach | |
| CN113642700A (en) | Cross-platform multi-modal public opinion analysis method based on federal learning and edge calculation | |
| CN107786958A (en) | A kind of data fusion method based on deep learning model | |
| Li et al. | FSB-EA: Fuzzy search bias guided constraint handling technique for evolutionary algorithm | |
| Muhsen et al. | Enhancing NoC-Based MPSoC performance: A predictive approach with ANN and guaranteed convergence arithmetic optimization algorithm | |
| Fu et al. | Nonlinear dynamic measurement method of software reliability based on data mining | |
| CN107949029A (en) | A kind of leader cluster node method for routing, device and medium based on wireless sense network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210316 Termination date: 20211222 |