CN105246109A - Intra-cluster data fusion method for vehicle Ad-Hoc Network - Google Patents
Intra-cluster data fusion method for vehicle Ad-Hoc Network Download PDFInfo
- Publication number
- CN105246109A CN105246109A CN201510595921.1A CN201510595921A CN105246109A CN 105246109 A CN105246109 A CN 105246109A CN 201510595921 A CN201510595921 A CN 201510595921A CN 105246109 A CN105246109 A CN 105246109A
- Authority
- CN
- China
- Prior art keywords
- node
- bunch
- sensor node
- represent
- neighborhood
- 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
- H04W28/00—Network traffic management; Network resource management
- H04W28/02—Traffic management, e.g. flow control or congestion control
- H04W28/08—Load balancing or load distribution
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W28/00—Network traffic management; Network resource management
- H04W28/16—Central resource management; Negotiation of resources or communication parameters, e.g. negotiating bandwidth or QoS [Quality of Service]
- H04W28/24—Negotiating SLA [Service Level Agreement]; Negotiating QoS [Quality of Service]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W84/00—Network topologies
- H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Quality & Reliability (AREA)
- Mobile Radio Communication Systems (AREA)
- Arrangements For Transmission Of Measured Signals (AREA)
Abstract
The invention relates to an intra-cluster data fusion method for a vehicle Ad-Hoc Network, and the method comprises the following steps that sensor nodes in the vehicle Ad-Hoc Network form a plurality of clusters; intra-cluster sensor nodes collect real-time data, updates sampling data in a sliding window, then calculates node credible values, and enables the node credible values to be broadcast to neighborhood nodes; all sensor nodes calculate a credible gain and neighborhood credible feedback quality according to the node credible values transmitted by the neighborhood nodes, then calculate separation vector gain with the neighborhood nodes, enable the obtained neighborhood credible feedback quality and separation vector gain to be transmitted to a cluster head node; the cluster head node calculates a benefit function comprising a cluster redundancy degree and a cluster structure change degree, searches an optimal transmission strategy according to a rule of benefit function maximization; and finally a part of intra-cluster sensor nodes are selected to transmit the sampling data to the cluster head node according to the obtained optimal transmission strategy. The method facilitates the improvement of the high efficiency and precision of data fusion.
Description
Technical field
The present invention relates to car networking data integration technology field, particularly a kind of towards car self-organization network bunch in data fusion method, be mainly used in carrying out data fusion to the sample information of the vehicle node of high-speed motion.
Background technology
Vehicle self-organizing network (VehicularAdHocNetwork is called for short VANET) is the important component part of intelligent transportation field.Data fusion is a vital task in car networking, is that all tool is of great significance in theory or in practical application.Data fusion can analyze road traffic condition from a large amount of vehicle sensory information, and the transport information that logistics monitoring information etc. is important, contains huge using value in the vehicle sensory information of magnanimity.
For the problem of data fusion of mobile ad hoc network, Mitra etc. use mobile agent optional m igration path adaptively in a particular area, and periodically collect the data message of nearby vehicle.The method utilizes flexibility and the adaptivity of mobile agent, effectively can reduce vehicle self-organizing network topological structure mutability to the impact of data fusion accuracy, but the uncertainty of the quantity of mobile agent and Mobile routing can affect the stability of algorithm.M.Zhao and Yang utilizes the information collection node of movement to collect data.Vehicle node is divided into poll node and leg gusset by the method, and leg gusset can send data to poll node by the intermediate node of multi-hop, the data that poll node processing is collected, and fused data is regularly sent to the information collection node of movement.The problem finding poll node in the method is np hard problem, and along with the increase of car self-organization network scale, the processing time will significantly increase, and greatly reduces fusion efficiencies.Wang etc. propose a kind of data fusion method based on distance, and integration region is divided into two sub regions of formed objects by the method, and every sub regions has specific cluster head node.At the edge of integration region, there is the aggregation node of a movement.Member node in bunch, by the syncretizing mechanism based on distance, utilizes the transmission means of multi-hop, and the aggregation node to cluster head node or edge sends data.The data that cluster head node process receives, and aggregation node sampling data transmitting being given edge.But the transmission of mass data will cause offered load overweight, cannot adapt to large-scale vehicle self-organizing network.Korteweg etc. propose a kind of fusion method considering data syn-chronization and propagation delay, effectively can reduce the time delay of fusion, but the method do not consider the reliability of initial data, merge accuracy and have much room for improvement.Wischhof etc. propose a kind of non-hierarchical fusion method, the method is to concrete under distributed environment, up-to-date vehicle traffic data carries out modeling, apply didactic adaptive algorithm to carry out data fusion, limit the propagation quantity with the data of translation specifications, thus adapt to large-scale sparse car self-organization network.But the control information that the method is used is too much, ageing needing is optimized further.For all kinds of problems existed in said method, the present invention design a kind of towards car self-organization network bunch in data fusion method to reduce the redundancy of data, improve the accuracy of data fusion.
Summary of the invention
The object of the present invention is to provide a kind of towards car self-organization network bunch in data fusion method, the method is conducive to improving the high efficiency of data fusion and accuracy.
For achieving the above object, technical scheme of the present invention is: a kind of towards car self-organization network bunch in data fusion method, comprise the following steps:
Step S1: the sensor node in car self-organization network forms several sensor node set to be called bunch, and each bunch is made up of a cluster head node and multiple member node;
Step S2: each sensor node in bunch gathers real time data, and the sampled data in the sliding window of more new sensor node;
Step S3: each sensor node in bunch according to the sampled data in sliding window, computing node confidence values, and node confidence values is broadcast to neighborhood node;
Step S4: the node confidence values that each sensor node in bunch sends according to the neighborhood node received, calculates credible gain and neighborhood trusted feedback quality;
Step S5: each sensor node in bunch, according to the situation of change of the relative displacement between itself and neighborhood node, calculates the separating vector gain for estimating wireless communication link stability between itself and neighborhood node;
Step S6: the neighborhood trusted feedback quality that oneself calculates by each sensor node in bunch, separating vector gain send to cluster head node;
Step S7: the neighborhood trusted feedback quality that sends of each sensor node in cluster head node receives bunch and separating vector gain, calculate the benefit function comprising bunch redundancy and clustering architecture change degree;
Step S8: cluster head node finds optimal transmission strategy according to the maximized criterion of benefit function, namely theory of games is utilized to solve the Nash Equilibrium optimal solution of benefit function, if Nash Equilibrium optimal solution does not meet round requirement, then disruption and recovery is adopted to solve locally optimal solution;
Step S9: cluster head node according to obtained optimal transmission strategy, select bunch in operative sensor node send sampled data to cluster head node, all the other sensor nodes do not send sampled data.
Further, in described step S2, the update method of the sampled data in sliding window is: if sliding window less than, then the real time data of up-to-date collection is put into sliding window; If fill sampled data in sliding window, then adopt queue first in first out, abandon the sampled data on first position in sliding window, and the real time data of up-to-date collection to be filled in sliding window on last position.
Further, in described step S3, sensor node according to the sampled data in sliding window, computing node confidence values, computing formula is:
Wherein,
krepresent the
kindividual sampling instant,
qrepresent the sampling instant of the sampled data that first position of sliding window stores,
rV x (
k)
represent sensor node
x?
kthe node confidence values of individual sampling instant,
wrepresent sliding window size,
arepresent regulatory factor,
dP x (
k) represent sensor node
x?
kthe sampled value of individual sampling instant,
with
be illustrated respectively in
kindividual sampling instant sensor node
xsliding window in the average of sampled data set and variance,
std() expression asks variance to the data in bracket.
Further, in described step S4, calculate credible gain and neighborhood trusted feedback quality, comprise the following steps:
Step S401: the credible gain between calculating sensor node and neighborhood node, for comparing the node confidence values between neighborhood node, computing formula is:
Wherein,
rG xy (
k)
represent the
kindividual sampling instant sensor node
yto sensor node
xcredible gain;
rV x (
k)
with
rV y (
k)
represent sensor node respectively
xand sensor node
y?
kthe node confidence values of individual sampling instant;
Step S402: calculating sensor joint neighborhood of a point trusted feedback quality, weigh the ability in sampling of sensor node, computing formula is:
Wherein,
rQ x (
k)
represent sensor node
x?
kthe neighborhood trusted feedback quality of individual sampling instant;
n x (
k)
represent sensor node
x?
kthe neighborhood node set of individual sampling instant, sensor node
yit is sensor node
xsensor node in neighborhood; The neighborhood trusted feedback quality of sensor node is higher, and its sampled data more can represent the attributive character in this region, and the ability in sampling of sensor node is stronger.
Further, in described step S5, the computing formula for the separating vector gain of wireless communication link stability between estimated sensor node and neighborhood node is:
Wherein,
represent the
kindividual sampling instant, sensor node
v j with sensor node
v i between separating vector gain;
represent sensor node
v i and sensor node
v j ?
kthe relative displacement size of individual sampling instant, sensor node
v i receiving node, sensor node
v j be sending node, separating vector gain analyzes wireless communication link variation tendency according to the relative displacement situation of change of adjacent node within the unit interval, and separating vector gain is larger, and link is more unstable.
Further, in described step S7, calculate benefit function and comprise the following steps:
Step S701: compute cluster redundancy, computing formula is:
Wherein,
cRD x (
k)
represent the
kindividual sampling instant, bunch employing transmission policy
xthe redundancy obtained;
xrepresent the
kthe transmission policy of individual sampling instant bunch interior nodes,
x=
p 1 (
k)
,
p 2 (
k)
...,
p n (
k)
,
p i (
k)
represent node
i?
kthe transmission policy of individual sampling instant,
p i (
k)
=1 represents that cluster head node selects node
i?
kindividual sampling instant transmits sampled data to cluster head node,
p i (
k)
=0 represents that cluster head node selects node
i?
kindividual sampling instant is not to cluster head node transmission sampled data;
nrepresent bunch interior nodes number,
c(
h) represent cluster head node
hthe node set that all nodes at place bunch are formed;
rQ i (
k)
represent node
i?
kthe neighborhood trusted feedback quality of individual sampling instant, by being incorporated in bunch Redundancy Analysis by neighborhood trusted feedback quality, reduces the necessity that the lower node of ability in sampling sends data, reaches the object of sampled data redundancy in reducing bunch;
Step S702: compute cluster structural change degree, computing formula is:
Wherein,
cVD h (
k)
represent cluster head node
hthe clustering architecture change degree at place bunch;
kbe
kindividual sampling instant,
represent the
iindividual sampling instant node
yto cluster head node
hseparating vector gain;
qrepresent the sampling instant corresponding to first position of sliding window; By clustering architecture change degree, can the passing in time of analytic manifold structure and the trend changed, clustering architecture change degree is less, and degree of stability is larger;
Step S703: calculate benefit function, computing formula is:
Wherein,
f(
x) represent the benefit function of game subject; Sensor node in bunch is mapped as the main body participating in game, and in bunch, all the sensors node constitutes the game subject set participated in the competition
i=1,2 ...,
n, wherein
nfor a bunch interior nodes quantity,
p i (
t)
{ 0,1} represents game subject to ∈
i?
tthe strategy in moment,
xrepresent the overall transfer strategy of bunch interior nodes,
x=
p 1 (
k)
,
p 2 (
k)
...,
p n (
k)
,
τrepresent benefit function regulatory factor, span is (0,1), for representing the balanced relation of bunch redundancy and clustering architecture change degree.
Further, in described step S8, round requirement when Nash Equilibrium optimal solution does not meet, when namely optimal policy contains decimal solution, adopt disruption and recovery, solve the local optimum transmission policy satisfied condition; Perturbation process comprises the following steps:
Step S801: main body
ipass through function
rand(0,1) generates random number
a i ;
Step S802: compare
a i the disturbance probability set with algorithm
γif,
a i <
γ, then Stochastic choice main body
iset of strategies in an element, replace current strategies; Otherwise, maintain former strategy constant;
Step S803: all main bodys obtain new meromixis transmission policy combination after the disturbance of step S801 and step S802
x ', calculate according to benefit function
x 'benefit value, if the benefit value of this strategy is greater than current strategies
xbenefit value, then replace current optimal policy, complete a Static disturbance; Otherwise maintain former strategy constant, game situation is in stable state;
Step S804: through Static disturbance and the stable state recovery of several times, finally form the stable optimal transmission strategy of local data's emerging system.
The invention has the beneficial effects as follows propose a kind of towards car self-organization network bunch in data fusion method, the method is applicable to the data fusion application scenarios under car self-organization network environment, effectively can reduce redundant data transmissions, obtain high-precision data fusion effect, there is stronger practicality and wide application prospect.
Accompanying drawing explanation
Fig. 1 is the realization flow figure of the inventive method.
Fig. 2 is the realization flow figure of step S4 in the inventive method.
Fig. 3 is the realization flow figure of step S7 in the inventive method.
Fig. 4 is the realization flow figure of step S8 in the inventive method.
Embodiment
Below in conjunction with drawings and the specific embodiments, the present invention is described in further detail.
Traditional data fusion method combines with car networked environment by the present invention, realizes the data fusion application under car networked environment.Few by designing a kind of bandwidth consumption, fusion accuracy is high, can adapt to the data fusion method of the car self-organization network of change at a high speed to reduce communication overhead and network configuration maintenance costs, improve fusion mass.
Fig. 1 be of the present invention towards car self-organization network bunch in the realization flow figure of data fusion method.As shown in Figure 1, the method comprises the following steps:
Step S1: the sensor node in car self-organization network forms several sensor node set to be called bunch, and each bunch is made up of a cluster head node and multiple member node.
Step S2: each sensor node in bunch gathers real time data, and the sampled data in the sliding window of more new sensor node.
Concrete, the update method of the sampled data in sliding window is: if sliding window less than, then the real time data of up-to-date collection is put into sliding window; If fill sampled data in sliding window, then adopt queue first in first out, abandon the sampled data on first position in sliding window, and the real time data of up-to-date collection to be filled in sliding window on last position.
Step S3: each sensor node in bunch according to the sampled data in sliding window, computing node confidence values, and node confidence values is broadcast to neighborhood node.
Concrete, the computing formula of node confidence values is:
Wherein,
krepresent the
kindividual sampling instant,
qrepresent the sampling instant of the sampled data that first position of sliding window stores,
rV x (
k)
represent sensor node
x?
kthe node confidence values of individual sampling instant,
wrepresent sliding window size,
arepresent regulatory factor,
dP x (
k) represent sensor node
x?
kthe sampled value of individual sampling instant,
with
be illustrated respectively in
kindividual sampling instant sensor node
xsliding window in the average of sampled data set and variance,
std() expression asks variance to the data in bracket.
Step S4: the node confidence values that each sensor node in bunch sends according to the neighborhood node received, calculating credible gain and neighborhood trusted feedback quality, by comparing the confidence values size between neighborhood node, again weighing the ability in sampling of sensor node.
Fig. 2 is the realization flow figure of step S4 in the inventive method, specifically comprises the following steps:
Step S401: the credible gain between calculating sensor node and neighborhood node, to compare the node confidence values between neighborhood node, computing formula is:
Wherein,
rG xy (
k)
represent the
kindividual sampling instant sensor node
yto sensor node
xcredible gain;
rV x (
k)
with
rV y (
k)
represent sensor node respectively
xand sensor node
y?
kthe node confidence values of individual sampling instant;
Step S402: calculating sensor joint neighborhood of a point trusted feedback quality, to weigh the ability in sampling of sensor node.
Concrete, neighborhood trusted feedback quality calculation formula is:
Wherein,
rQ x (
k)
represent sensor node
x?
kthe neighborhood trusted feedback quality of individual sampling instant;
n x (
k)
represent sensor node
x?
kthe neighborhood node set of individual sampling instant, sensor node
yit is sensor node
xsensor node in neighborhood;
rG xy (
k)
represent the
kindividual sampling instant sensor node
yto sensor node
xcredible gain; The neighborhood trusted feedback quality of sensor node is higher, and its sampled data more can represent the attributive character in this region, and the ability in sampling of sensor node is stronger.
Step S5: each sensor node in bunch, according to the situation of change of the relative displacement between itself and neighborhood node, calculates the separating vector gain for estimating wireless communication link stability between itself and neighborhood node.
Concrete, the computing formula of separating vector gain is:
Wherein,
represent the
kindividual sampling instant, sensor node
v j with sensor node
v i between separating vector gain;
represent sensor node
v i and sensor node
v j ?
kthe relative displacement size of individual sampling instant, sensor node
v i receiving node, sensor node
v j be sending node, separating vector gain analyzes wireless communication link variation tendency according to the relative displacement situation of change of adjacent node within the unit interval, and separating vector gain is larger, and link is more unstable.
Step S6: the neighborhood trusted feedback quality that oneself calculates by each sensor node in bunch, separating vector gain send to cluster head node.
Step S7: the neighborhood trusted feedback quality that sends of each sensor node in cluster head node receives bunch and separating vector gain, calculate the benefit function comprising bunch redundancy and clustering architecture change degree.
Fig. 3 is the realization flow figure of step S7 in the inventive method, specifically comprises the following steps:
Step S701: compute cluster redundancy, computing formula is:
Wherein,
cRD x (
k)
represent the
kindividual sampling instant, bunch employing transmission policy
xthe redundancy obtained;
xrepresent the
kthe transmission policy of individual sampling instant bunch interior nodes,
x=
p 1 (
k)
,
p 2 (
k)
...,
p n (
k)
,
p i (
k)
represent node
i?
kthe transmission policy of individual sampling instant,
p i (
k)
=1 represents that cluster head node selects node
i?
kindividual sampling instant transmits sampled data to cluster head node,
p i (
k)
=0 represents that cluster head node selects node
i?
kindividual sampling instant is not to cluster head node transmission sampled data;
nrepresent bunch interior nodes number,
c(
h) represent cluster head node
hthe node set that all nodes at place bunch are formed;
rQ i (
k)
represent node
i?
kthe neighborhood trusted feedback quality of individual sampling instant, by being incorporated in bunch Redundancy Analysis by neighborhood trusted feedback quality, reduces the necessity that the lower node of ability in sampling sends data, reaches the object of sampled data redundancy in reducing bunch;
Step S702: compute cluster structural change degree, computing formula is:
Wherein,
cVD h (
k)
represent cluster head node
hthe clustering architecture change degree at place bunch;
kbe
kindividual sampling instant,
represent the
iindividual sampling instant node
yto cluster head node
hseparating vector gain;
qrepresent the sampling instant corresponding to first position of sliding window; By clustering architecture change degree, can the passing in time of analytic manifold structure and the trend changed, clustering architecture change degree is less, and degree of stability is larger;
Step S703: calculate benefit function, computing formula is:
Wherein,
f(
x) represent the benefit function of game subject; Sensor node in bunch is mapped as the main body participating in game, and in bunch, all the sensors node constitutes the game subject set participated in the competition
i=1,2 ...,
n, wherein
nfor a bunch interior nodes quantity,
p i (
t)
{ 0,1} represents game subject to ∈
i?
tthe strategy in moment,
xrepresent the overall transfer strategy of bunch interior nodes,
x=
p 1 (
k)
,
p 2 (
k)
...,
p n (
k)
,
τrepresent benefit function regulatory factor, span is (0,1), for representing the balanced relation of bunch redundancy and clustering architecture change degree.
Step S8: cluster head node finds optimal transmission strategy according to the maximized criterion of benefit function, namely theory of games is utilized to solve the Nash Equilibrium optimal solution of benefit function, if Nash Equilibrium optimal solution does not meet round requirement, then disruption and recovery is adopted to solve locally optimal solution.
By obtaining the analysis of game target function, the Nash Equilibrium optimal solution of method existence anduniquess that the present invention proposes, when Nash Equilibrium optimal policy meet round require time, algorithm terminates.Round requirement when Nash Equilibrium optimal solution does not meet, when namely optimal policy contains decimal solution, adopt disruption and recovery, solve the local optimum transmission policy satisfied condition.
Fig. 4 is the realization flow figure of step S8 in the inventive method, specifically comprises the following steps:
Step S801: main body
ipass through function
rand(0,1) generates random number
a i ;
Step S802: compare
a i the disturbance probability set with algorithm
γif,
a i <
γ, then Stochastic choice main body
iset of strategies in an element, replace current strategies; Otherwise, maintain former strategy constant;
Step S803: all main bodys obtain new transmission policy combination after the disturbance of step S801 and step S802
x ', calculate according to benefit function
x 'benefit value, if the benefit value of this strategy is greater than current strategies
xbenefit value, then replace current optimal policy, complete a Static disturbance; Otherwise maintain former strategy constant, game situation is in stable state;
Step S804: through Static disturbance and the stable state recovery of several times, finally form the stable optimal transmission strategy of local data's emerging system.
Step S9: cluster head node according to obtained optimal transmission strategy, select bunch in operative sensor node send sampled data to cluster head node, all the other sensor nodes do not send sampled data.
Be more than preferred embodiment of the present invention, all changes done according to technical solution of the present invention, when the function produced does not exceed the scope of technical solution of the present invention, all belong to protection scope of the present invention.
Claims (7)
1. towards car self-organization network bunch in a data fusion method, it is characterized in that, comprise the following steps:
Step S1: the sensor node in car self-organization network forms several sensor node set to be called bunch, and each bunch is made up of a cluster head node and multiple member node;
Step S2: each sensor node in bunch gathers real time data, and the sampled data in the sliding window of more new sensor node;
Step S3: each sensor node in bunch according to the sampled data in sliding window, computing node confidence values, and node confidence values is broadcast to neighborhood node;
Step S4: the node confidence values that each sensor node in bunch sends according to the neighborhood node received, calculates credible gain and neighborhood trusted feedback quality;
Step S5: each sensor node in bunch, according to the situation of change of the relative displacement between itself and neighborhood node, calculates the separating vector gain for estimating wireless communication link stability between itself and neighborhood node;
Step S6: the neighborhood trusted feedback quality that oneself calculates by each sensor node in bunch, separating vector gain send to cluster head node;
Step S7: the neighborhood trusted feedback quality that sends of each sensor node in cluster head node receives bunch and separating vector gain, calculate the benefit function comprising bunch redundancy and clustering architecture change degree;
Step S8: cluster head node finds optimal transmission strategy according to the maximized criterion of benefit function, namely theory of games is utilized to solve the Nash Equilibrium optimal solution of benefit function, if Nash Equilibrium optimal solution does not meet round requirement, then disruption and recovery is adopted to solve locally optimal solution;
Step S9: cluster head node according to obtained optimal transmission strategy, select bunch in operative sensor node send sampled data to cluster head node, all the other sensor nodes do not send sampled data.
2. according to claim 1 a kind of towards car self-organization network bunch in data fusion method, it is characterized in that, in described step S2, the update method of the sampled data in sliding window is: if sliding window less than, then the real time data of up-to-date collection is put into sliding window; If fill sampled data in sliding window, then adopt queue first in first out, abandon the sampled data on first position in sliding window, and the real time data of up-to-date collection to be filled in sliding window on last position.
3. according to claim 1 a kind of towards car self-organization network bunch in data fusion method, it is characterized in that, in described step S3, sensor node according to the sampled data in sliding window, computing node confidence values, computing formula is:
Wherein,
krepresent the
kindividual sampling instant,
qrepresent the sampling instant of the sampled data that first position of sliding window stores,
rV x (
k)
represent sensor node
x?
kthe node confidence values of individual sampling instant,
wrepresent sliding window size,
arepresent regulatory factor,
dP x (
k) represent sensor node
x?
kthe sampled value of individual sampling instant,
with
be illustrated respectively in
kindividual sampling instant sensor node
xsliding window in the average of sampled data set and variance,
std() expression asks variance to the data in bracket.
4. according to claim 1 a kind of towards car self-organization network bunch in data fusion method, it is characterized in that, in described step S4, calculate credible gain and neighborhood trusted feedback quality, comprise the following steps:
Step S401: the credible gain between calculating sensor node and neighborhood node, for comparing the node confidence values between neighborhood node, computing formula is:
Wherein,
rG xy (
k)
represent the
kindividual sampling instant sensor node
yto sensor node
xcredible gain;
rV x (
k)
with
rV y (
k)
represent sensor node respectively
xand sensor node
y?
kthe node confidence values of individual sampling instant;
Step S402: calculating sensor joint neighborhood of a point trusted feedback quality, weigh the ability in sampling of sensor node, computing formula is:
Wherein,
rQ x (
k)
represent sensor node
x?
kthe neighborhood trusted feedback quality of individual sampling instant;
n x (
k)
represent sensor node
x?
kthe neighborhood node set of individual sampling instant, sensor node
yit is sensor node
xsensor node in neighborhood; The neighborhood trusted feedback quality of sensor node is higher, and its sampled data more can represent the attributive character in this region, and the ability in sampling of sensor node is stronger.
5. according to claim 1 a kind of towards car self-organization network bunch in data fusion method, it is characterized in that, in described step S5, the computing formula for the separating vector gain of wireless communication link stability between estimated sensor node and neighborhood node is:
Wherein,
represent the
kindividual sampling instant, sensor node
v j with sensor node
v i between separating vector gain;
represent sensor node
v i and sensor node
v j ?
kthe relative displacement size of individual sampling instant, sensor node
v i receiving node, sensor node
v j be sending node, separating vector gain analyzes wireless communication link variation tendency according to the relative displacement situation of change of adjacent node within the unit interval, and separating vector gain is larger, and link is more unstable.
6. according to claim 1 a kind of towards car self-organization network bunch in data fusion method, it is characterized in that, in described step S7, calculate benefit function comprise the following steps:
Step S701: compute cluster redundancy, computing formula is:
Wherein,
cRD x (
k)
represent the
kindividual sampling instant, bunch employing transmission policy
xthe redundancy obtained;
xrepresent the
kthe transmission policy of individual sampling instant bunch interior nodes,
x=
p 1 (
k)
,
p 2 (
k)
...,
p n (
k)
,
p i (
k)
represent node
i?
kthe transmission policy of individual sampling instant,
p i (
k)
=1 represents that cluster head node selects node
i?
kindividual sampling instant transmits sampled data to cluster head node,
p i (
k)
=0 represents that cluster head node selects node
i?
kindividual sampling instant is not to cluster head node transmission sampled data;
nrepresent bunch interior nodes number,
ithe set that in representing bunch, all nodes is formed;
c(
h) represent cluster head node
hthe node set that all nodes at place bunch are formed;
rQ i (
k)
represent node
i?
kthe neighborhood trusted feedback quality of individual sampling instant, by being incorporated in bunch Redundancy Analysis by neighborhood trusted feedback quality, reduces the necessity that the lower node of ability in sampling sends data, reaches the object of sampled data redundancy in reducing bunch;
Step S702: compute cluster structural change degree, computing formula is:
Wherein,
cVD h (
k)
represent cluster head node
hthe clustering architecture change degree at place bunch;
kbe
kindividual sampling instant,
represent the
iindividual sampling instant node
yto cluster head node
hseparating vector gain;
qrepresent the sampling instant corresponding to first position of sliding window; By clustering architecture change degree, can the passing in time of analytic manifold structure and the trend changed, clustering architecture change degree is less, and degree of stability is larger;
Step S703: calculate benefit function, computing formula is:
Wherein,
f(
x) represent the benefit function of game subject; Sensor node in bunch is mapped as the main body participating in game, and in bunch, all the sensors node constitutes the game subject set participated in the competition
i=1,2 ...,
n, wherein
nfor a bunch interior nodes quantity,
p i (
t)
{ 0,1} represents game subject to ∈
i?
tthe strategy in moment,
xrepresent the overall transfer strategy of bunch interior nodes,
x=
p 1 (
k)
,
p 2 (
k)
...,
p n (
k)
,
τrepresent benefit function regulatory factor, span is (0,1), for representing the balanced relation of bunch redundancy and clustering architecture change degree.
7. according to claim 1 a kind of towards car self-organization network bunch in data fusion method, it is characterized in that, in described step S8, requirement is rounded when Nash Equilibrium optimal solution does not meet, namely when optimal policy contains decimal solution, adopt disruption and recovery, solve the local optimum transmission policy satisfied condition; Perturbation process comprises the following steps:
Step S801: main body
ipass through function
rand(0,1) generates random number
a i ;
Step S802: compare
a i the disturbance probability set with algorithm
γif,
a i <
γ, then Stochastic choice main body
iset of strategies in an element, replace current strategies; Otherwise, maintain former strategy constant;
Step S803: all main bodys obtain new transmission policy combination after the disturbance of step S801 and step S802
x ', calculate according to benefit function
x 'benefit value, if the benefit value of this strategy is greater than current strategies
xbenefit value, then replace current optimal policy, complete a Static disturbance; Otherwise maintain former strategy constant, game situation is in stable state;
Step S804: through Static disturbance and the stable state recovery of several times, finally form the stable optimal transmission strategy of local data's emerging system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510595921.1A CN105246109B (en) | 2015-09-18 | 2015-09-18 | Data fusion method in a kind of cluster towards car self-organization network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510595921.1A CN105246109B (en) | 2015-09-18 | 2015-09-18 | Data fusion method in a kind of cluster towards car self-organization network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105246109A true CN105246109A (en) | 2016-01-13 |
| CN105246109B CN105246109B (en) | 2018-10-30 |
Family
ID=55043530
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510595921.1A Expired - Fee Related CN105246109B (en) | 2015-09-18 | 2015-09-18 | Data fusion method in a kind of cluster towards car self-organization network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105246109B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111194018A (en) * | 2020-01-06 | 2020-05-22 | 北京邮电大学 | Vehicle ad hoc network-oriented mobile edge data uploading method and system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110280154A1 (en) * | 2006-12-07 | 2011-11-17 | Silverstrim James E | Method and Apparatus for Management of a Global Wireless Sensor Network |
| KR20120031569A (en) * | 2010-09-27 | 2012-04-04 | 아주대학교산학협력단 | Content-based routing method in mobile ad-hoc networks and apparatus thereof |
| CN103580806A (en) * | 2013-10-31 | 2014-02-12 | 西安电子科技大学 | Method for energy-efficient transmission of robustness in cognitive network |
| CN104038997A (en) * | 2014-06-25 | 2014-09-10 | 中国电子科技集团公司第七研究所 | Method and system for activating energy efficiency of wireless sensor network nodes |
| CN104053163A (en) * | 2014-06-24 | 2014-09-17 | 无锡北邮感知技术产业研究院有限公司 | Heterogeneous wireless network networking method based on calculation and processing |
-
2015
- 2015-09-18 CN CN201510595921.1A patent/CN105246109B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110280154A1 (en) * | 2006-12-07 | 2011-11-17 | Silverstrim James E | Method and Apparatus for Management of a Global Wireless Sensor Network |
| KR20120031569A (en) * | 2010-09-27 | 2012-04-04 | 아주대학교산학협력단 | Content-based routing method in mobile ad-hoc networks and apparatus thereof |
| CN103580806A (en) * | 2013-10-31 | 2014-02-12 | 西安电子科技大学 | Method for energy-efficient transmission of robustness in cognitive network |
| CN104053163A (en) * | 2014-06-24 | 2014-09-17 | 无锡北邮感知技术产业研究院有限公司 | Heterogeneous wireless network networking method based on calculation and processing |
| CN104038997A (en) * | 2014-06-25 | 2014-09-10 | 中国电子科技集团公司第七研究所 | Method and system for activating energy efficiency of wireless sensor network nodes |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111194018A (en) * | 2020-01-06 | 2020-05-22 | 北京邮电大学 | Vehicle ad hoc network-oriented mobile edge data uploading method and system |
| CN111194018B (en) * | 2020-01-06 | 2021-03-02 | 北京邮电大学 | Vehicle ad hoc network-oriented mobile edge data uploading method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105246109B (en) | 2018-10-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Luo et al. | Intersection-based V2X routing via reinforcement learning in vehicular ad hoc networks | |
| Wang et al. | An energy-efficient distributed adaptive cooperative routing based on reinforcement learning in wireless multimedia sensor networks | |
| CN101188535B (en) | Method for determining clustering energy balance route of wireless sensor network based on binary tree | |
| Kumar et al. | Energy-efficient mobile data collection in wireless sensor networks with delay reduction using wireless communication | |
| CN110336751B (en) | Low-orbit satellite network routing strategy based on membership function | |
| Karim et al. | Energy efficient and fault tolerant routing protocol for mobile sensor network | |
| CN102316496A (en) | Data merging method based on Kalman filtering in wireless sensor network | |
| CN103200598A (en) | Wireless sensor network clustering method based on particle swarm optimization | |
| CN104469879A (en) | Dynamic k value clustering routing method | |
| Zhang et al. | Routing clustering protocol for 3D wireless sensor networks based on fragile collection ant colony algorithm | |
| Madhumathy et al. | Enabling energy efficient sensory data collection using multiple mobile sink | |
| CN104284386A (en) | Vertex-betweenness-based cluster head selection method in wireless sensor networks | |
| Kour | An energy efficient routing algorithm for WBAN | |
| CN106161118A (en) | In a kind of wireless sense network, the k of balancing energy jumps cluster routing method | |
| CN103906245A (en) | Distributed rate control method for wireless sensor network | |
| CN105246109A (en) | Intra-cluster data fusion method for vehicle Ad-Hoc Network | |
| Peng et al. | A new clustering algorithm based on ACO and K-medoids optimization methods | |
| Robinson et al. | Tree based data fusion clustering routing algorithm for illimitable network administration in wireless sensor network | |
| Raghunandan et al. | Hierarchical agglomerative clustering based routing algorithm for overall efficiency of wireless sensor network | |
| Chen et al. | Region segmentation model for wireless sensor networks considering optimal energy conservation constraints | |
| CN103179631B (en) | A kind of choose the method that optimal angle carries out sub-clustering | |
| Wang et al. | Advanced MDS-MAP localization algorithm with weighted clustering and heuristic ranging | |
| Zhu et al. | Multiple ant colony routing optimization based on cloud model for WSN with long-chain structure | |
| CN104602255A (en) | Delay-tolerant mobile sensor network node transmitting probability dynamic calculation method under RWP (Random Way-Point) model | |
| CN111010704A (en) | Underwater wireless sensor network data prediction optimization method based on exponential smoothing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into 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: 20181030 Termination date: 20210918 |