CN104469875B - Method for tracking target and its system based on prediction in radio sensing network - Google Patents

Method for tracking target and its system based on prediction in radio sensing network Download PDF

Info

Publication number
CN104469875B
CN104469875B CN201410693468.3A CN201410693468A CN104469875B CN 104469875 B CN104469875 B CN 104469875B CN 201410693468 A CN201410693468 A CN 201410693468A CN 104469875 B CN104469875 B CN 104469875B
Authority
CN
China
Prior art keywords
node
sensor node
tracking
sequence
prediction
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.)
Active
Application number
CN201410693468.3A
Other languages
Chinese (zh)
Other versions
CN104469875A (en
Inventor
高志鹏
程伟静
芮兰兰
王颖
刘会永
熊翱
亓峰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing University of Posts and Telecommunications
Original Assignee
Beijing University of Posts and Telecommunications
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing University of Posts and Telecommunications filed Critical Beijing University of Posts and Telecommunications
Priority to CN201410693468.3A priority Critical patent/CN104469875B/en
Publication of CN104469875A publication Critical patent/CN104469875A/en
Application granted granted Critical
Publication of CN104469875B publication Critical patent/CN104469875B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W40/00Communication routing or communication path finding
    • H04W40/02Communication route or path selection, e.g. power-based or shortest path routing
    • H04W40/18Communication route or path selection, e.g. power-based or shortest path routing based on predicted events
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W40/00Communication routing or communication path finding
    • H04W40/24Connectivity information management, e.g. connectivity discovery or connectivity update
    • H04W40/248Connectivity information update
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W84/00Network topologies
    • H04W84/18Self-organising networks, e.g. ad-hoc networks or sensor networks

Abstract

The present invention relates to method for tracking target and system based on prediction, methods described in a kind of radio sensing network to include:S1, sub-clustering is carried out to all the sensors node in radio sensing network;S2, all the sensors node after sub-clustering is placed in active state, collects the history mobile route information of mobile object;S3, the backward dependence sequence for establishing according to the history mobile route information all sensor nodes, and by described backward sequence storage is relied on to corresponding leader cluster node;S4, in each cluster according to the tracking for carrying out the mobile object after described to the state for relying on sequence control sensor node.This method reduces the long haul communication between sensor node and base station by way of by sensor node sub-clustering, realizes the local updating of information of forecasting.And during tracking, it is predicted according to backward dependence mode, greatly reduces time complexity, and improve tracking precision, supported more type objects while track.

Description

Method for tracking target and its system based on prediction in radio sensing network
Technical field
The present invention relates to the mesh based on prediction in technology of wireless sensing network field, more particularly to a kind of radio sensing network Mark tracking and its system.
Background technology
In numerous applications of radio sensing network, target following Sensor Network (OTSN) is one of application most consumed energy, OTSN is mainly used to track mobile object and reports latest position, and this dynamic process expends many Internet resources, and senses The power supply energy of device is extremely limited.There are two kinds of main approach to solve sensor energy consumption problem at present, first, being sensed by improving The hardware design of device, optimize physical arrangement to reduce the energy consumption of sensing node;Second, by controlling sensor states to be saved, Idle sensor is set to keep resting state as far as possible, this kind of method is divided into two kinds again, and a kind of is the method for being not based on prediction, all Sensor is set to active and resting state by phase property, and another kind of is the method based on prediction, by entering to path log information Row law-analysing, sensor node is selectively allowed to keep the resting state of maximum duration.
Scheme 1:Paper《An Energy Efficient Technique for Object Tracking in Wireless Sensor Networks》(ICC,2011:1-5) propose PTSP methods and carry out articles path tracking.This method Mainly according to history mobile route, the front and rear dependence sequence of all sensor nodes is generated, then according to sequence-dependent frequency Rate size selects the stronger sequence of dependence, and the mobile route of object is predicted based on these sequences.This method relies in generation It is in sequence process, it is necessary to higher to node before and after each sensor node traversal on path, complexity.In addition, pick out Dependence sequence can not be fully described object moving direction, even some sensor nodes without stronger front and rear dependence sequence, Will result in can not judge next-hop sensor node, and Loss Rate is higher.
Scheme 2:Paper《An object tracking scheme for wireless sensor networks using data mining mechanism》(NOMS IEEE,2012:Application Apriori methods are based on going through in 526-529) History routing information Mining Association Rules, this method sensor node that progressively Mining Frequent accesses, each step increase a sensing Device node, the sensor sequence that supporting rate is less than threshold value is deleted, terminated until sensor node can not be increased.Finally give sensing The most long dependence sequence of device, path trace is carried out according to the dependence sequence control sensor node state.This method is excavated in every step During the sensor node frequently accessed, it is both needed to travel through All Paths information once.The mobile route information content of object is larger, therefore should Method complexity is higher.In addition, the final most long dependence sequence excavated cannot be guaranteed to include all the sensors node, for not In the most long sensor node relied in sequence, next hop information can not be judged.
Scheme 3:A kind of wireless sensor network target track sides based on prediction of patent No. ZL 200710164468.4 Method, this method comprise the following steps:A. the current measurement data or historical measurement data moved according to target determines target Motion feature;B. the Future Positions of the information prediction target such as the current location of combining target, speed, direction of motion and next prison Control the wake-up moment of node;C. when target prodiction fails, network is recorded according to the motion history of target and priori Start prediction of failure recovery process step by step.The statistics that the present invention moves according to target determines the motion feature of target, and according to The Future movement of this prediction target.The invention is at the prediction Future Positions of target, wake-up moment of next monitor node and pre- Dendrometry loses when being recovered, and is required to target maximal rate, peak acceleration, maximum angular rate, maximum angular acceleration limit fortune The prioris such as dynamic parameter so that sensor node need to consume the calculating that more energy carries out priori.Though the invention exists There is some superiority in prediction accuracy, but it is larger to consume energy.
Scheme 4:A kind of method for tracking target of wireless sensor network of patent No. ZL200810103125.1, including it is following Step:Step A, data are observed using history target status information and current time, carry out importance sampling, obtain particle state Estimated information, track survival index and remaining measured value is calculated;Step B, decide whether to terminate according to track survival index The track, and update track set;Step C, using the particle after resampling, the current state for obtaining target complete track is estimated Meter, that is, move current location and the movement velocity of target, realize target locating.The invention uses the particle after resampling, Obtain the current state estimation of target complete track, that is, move current location and the movement velocity of target, realize target positioning with Track, this requires that all the sensors node necessarily be in time synchronized state, and complexity is higher.
Scheme 5:A kind of 200810225565.4 wireless sensor network targets of patent No. ZL position and tracking, institute Stating method mainly includes:At the arbitrarily positioning moment, according to sensor node metrical information pre-estimation target location, foundation includes mesh Mark the learning region of pre-estimation position, any number of location point chosen in learning region, using Polynomial kernel function and Ε- The mapping relations that support vector regression approaches location point to sensor node distance vector and position point coordinates obtain decision-making letter Number, sensor node is obtained into target location estimate to object ranging vector input decision function, by target location estimate Send to base station, base station and renewal target trajectory is fitted to object location history data, realizes target following.The invention By the way that sensor node is obtained into target location estimate, target location estimate quilt to object ranging vector input decision function Base station is sent to, base station is fitted renewal target trajectory to object location history data, realizes target following, for mesh The situation that movement locus does not have any curve law is marked, larger prediction error occurs in fit procedure, enters for the invention And cause Loss Rate higher.
The content of the invention
Based on above mentioned problem, the present invention provides the method for tracking target in a kind of radio sensing network based on prediction and its is System, by way of by sensor node sub-clustering, reduce the long haul communication between sensor node and base station, realize prediction Information local updating.And during tracking, it is predicted according to backward dependence mode, greatly reduces time complexity, And improve tracking precision.
According to above-mentioned purpose, the present invention provides the method for tracking target based on prediction in a kind of radio sensing network, described Method includes:
S1, sub-clustering is carried out to all the sensors node in radio sensing network;
S2, all the sensors node after sub-clustering is placed in active state, collects the history mobile route letter of mobile object Breath;
S3, the backward dependence sequence for establishing according to the history mobile route information all sensor nodes, and will The backward sequence that relies on is stored to corresponding leader cluster node;
S4, carry out the motive objects to the state for relying on sequence control sensor node according to after described in each cluster The tracking of body.
Wherein, the step S1 is specifically included:Sub-clustering is carried out to the sensor node using K-means algorithms.
Wherein, the step S2 also includes:By the history mobile route information classification and store.
Wherein, the step S4 is specifically included:
S41, calculate each sensor node in cluster destination node confidence level, and be arranged according to confidence level descending One-dimension array;
S42, stored the content of the one-dimension array as information of forecasting into corresponding sensor node;
S43, when moving object tracking, the information of forecasting that current sensor node stores according to the node swashs Next sensor node living, realize that mobile object tracks.
Wherein, the step S43 is specifically included:
S431, after the current sensor node is activated, start monitor tracking area in whether there is mobile object;
S432, when tracking the period in trace into mobile object, then tracking result is reported to a upper sensor section Point;
S433, next sensor node is predicted according to the information of forecasting of storage in preset time period, and activation is believed Breath is sent to next sensor node of prediction;
S434, the next sensor node perform step S431, the current sensor node receive it is described under Enter resting state after the tracking result of one sensor node, otherwise perform next step;
When S435, the current sensor are not received by the tracking result of next sensor node, then the movement Object tracking fails, and starts Restoration Mechanism, recovers the tracking of the mobile object.
Wherein, the step S435 is specifically included:
The Loss Rate during mobile object tracking is calculated, it is described current when the Loss Rate is less than threshold value The prediction of failure information is sent to corresponding leader cluster node by sensor, after then activating the current sensor node To objective sensor node all in sequence is relied on, subsequently into resting state;
If the objective sensor node traces into the mobile object, the sensor section of the mobile object is traced into Point sends confirmation to the leader cluster node;
The leader cluster node updates the rear to dependence of the sensor node of tracking failure after the confirmation is received Sequence;
According to the backward dependence sequence after renewal, using the sensor node for tracing into the object as working as prosthomere Point, perform step S431;
Wherein, the step S435 also includes:
If all objective sensor nodes all do not trace into the mobile object, using the leader cluster node as Source node activates all sensor nodes, moves the tracking of object;
The sensor node for tracing into the object is sent into confirmation to the leader cluster node;
The leader cluster node updates the rear to dependence of the sensor node of tracking failure after the confirmation is received Sequence;
According to the backward dependence sequence after renewal, using the sensor node for tracing into the object as working as prosthomere Point, perform step S431;
Wherein, the step S435 also includes:
When the Loss Rate is more than threshold value, the backward dependence sequence of renewal is sent to corresponding biography by the leader cluster node Sensor node, each corresponding sensor node update the backward dependence sequence of oneself according to the backward dependence sequence of reception Row, then perform step S41.
According to another aspect of the present invention, there is provided the Target Tracking System based on prediction in a kind of radio sensing network, institute The system of stating includes:
Sub-clustering unit, for carrying out sub-clustering to all the sensors node in radio sensing network;
History mobile message collector unit, for all the sensors node after sub-clustering to be placed in into active state, collect and move The history mobile route information of moving-target;
The backward sequence that relies on establishes unit, for establishing all sensors according to the history mobile route information The backward dependence sequence of node, and the backward sequence that relies on is stored to corresponding leader cluster node;
Object tracking unit, in each cluster according to after described to the shape for relying on sequence control sensor node State carries out the tracking of the mobile object.
Method for tracking target and its system based on prediction in the radio sensing network of the present invention, it need to only be based on backward rely on Object tracking can be carried out, greatly reduces time complexity, and improved in terms of accuracy;Meanwhile by dividing Cluster reduces the long haul communication between sensor node and base station, and prediction result is stored in into sensor node, reduces Interacting between sensor and cluster head, in path during rule changes locally, sub-clustering mechanism can not influence other parts On the premise of fast and convenient realization predict again;In addition, when tracking failure, restoration methods of the invention are not simple Flooding mechanism, but recovered first according to the path rule excavated, if still failing, then carry out global flooding. By this method, realization that can be quick and easy recovers, and significantly reduces tracking Loss Rate;In addition, this method supports multiclass Object tracks simultaneously.
Brief description of the drawings
The features and advantages of the present invention can be more clearly understood by reference to accompanying drawing, accompanying drawing is schematically without that should manage Solve to carry out any restrictions to the present invention, in the accompanying drawings:
Fig. 1 shows the flow chart of the method for tracking target based on prediction in radio sensing network of the invention.
Fig. 2 shows the detailed process signal of the method for tracking target based on prediction in radio sensing network of the invention Figure.
Fig. 3 shows the structured flowchart of the Target Tracking System based on prediction in radio sensing network of the invention.
Fig. 4 shows in the radio sensing network of the present invention method for tracking target based on prediction with existing method in total energy Comparison figure in terms of the consumption of source;
Fig. 5 is shown in the radio sensing network of the present invention to be lost based on the method for tracking target of prediction and existing method Comparison figure in terms of rate;
Fig. 6 is shown in the radio sensing network of the present invention and participated in based on the method for tracking target of prediction and existing method The comparison figure of the quantitative aspects of the sensor node of recovery;
Embodiment
Below in conjunction with accompanying drawing, embodiments of the present invention is described in detail.
Fig. 1 shows the flow chart of the method for tracking target based on prediction in radio sensing network of the invention.
Fig. 2 shows the detailed process signal of the method for tracking target based on prediction in radio sensing network of the invention Figure.
Referring to Figures 1 and 2, should the invention provides the method for tracking target based on prediction in a kind of radio sensing network Method includes:
S1, sub-clustering is carried out to all the sensors node in radio sensing network;
In the present embodiment, sub-clustering is carried out to the sensor node using K-means algorithms, after sub-clustering, in object tracking During, sensor node only need to carry out short-range communication with leader cluster node.
S2, all the sensors node after sub-clustering is placed in active state, collects the history mobile route letter of mobile target Breath;
In the present embodiment, OTSN surrounding, it can arrange that some are used to distinguishing the sensor of kind of object or a variety of The combination of sensor, then it is all placed in living by the sensor node for distinguishing kind of object and for the sensor node of mobile tracking Jump state, collection move the routing information of target and distinguish classification and stored.
This implementation is a large amount of log informations collected based on sink nodes, by the regularity of analytical behavior data, On the premise of allowing certain Loss Rate, the information of forecasting of next sensor node is formed.
S3, the backward dependence sequence seq for establishing according to the history mobile route information all sensor nodes [now, des], wherein, now represents the information of current sensor node, and des represents the information of objective sensor node, i.e., next The information of individual sensor node.
Rear to during relying on sequence foundation, tracking information is sent to leader cluster node, cluster by the sensor node in cluster Head node is responsible for tracking data being sent to base station, and object Move Mode is excavated in base station, is established according to history mobile route information Dependence sequence backward, and the rear of foundation is sent into meeting leader cluster node to dependence sequence, leader cluster node is by after to dependence sequence transmission To corresponding sensor node.Meanwhile during tracking, when the movement law of mobile object in cluster changes, only need Want current cluster head node to excavate the Move Mode of mobile object again, update backward dependence sequence, and other clusters are unaffected.By This understands, only needs each leader cluster node to carry out long haul communication with base station after sub-clustering, and when one of cluster is updated, its His cluster is unaffected, so as to reduce the complexity of calculating, saves the time.
S4, move object to the state for relying on sequence control sensor node according to after described in each cluster Tracking.
Step S4 is specifically included:
S41, calculate each sensor node in cluster destination node confidence level, and be arranged according to confidence level descending One-dimension array SN (SN_NUM) (type)={ des }, wherein, SN_NUM is the quantity of array, and respective sensor number, type is The type of array;
S42, by the content (the objective sensor node collection (des1, des2, des3 ...) of descending arrangement) of one-dimension array As information of forecasting storage into corresponding sensor node;
S43, when moving object tracking, the information of forecasting that current sensor node stores according to the node swashs Next sensor node living, realize that mobile object tracks.
Reference picture 2, step S43 are specifically included:
After S431, current sensor node are activated, start to whether there is object in monitoring tracking area;
S432, when tracking the period in trace into object, then tracking result is reported to a upper sensor node;
S433, next sensor node is predicted according to the information of forecasting of storage in preset time period, and activation is believed Breath is sent to next sensor node of prediction;
S434, next sensor node perform step S431, the current sensor node receive it is described next Enter resting state after the tracking result of sensor node, otherwise perform next step;
When S435, the current sensor are not received by the tracking result of next sensor node, then the movement Object tracking fails, and starts Restoration Mechanism, recovers the tracking of the mobile object.
Specifically, above-mentioned steps S43 is described in detail by specific examples below.
As illustrated, current sensor in a dormant state, after the T-X moment is activated, is opened within first T-X time Begin to whether there is object in monitoring tracking area, and report tracking result ACK to a upper sensor node at the X moment;Then exist In second T-X period, current sensor node predicts next sensor node according to this information of forecasting, in T-X Active information is sent after moment to next sensor node;And within second X period, current sensor node is with One sensor node is in active state, next sensor node after object is traced into, send tracking result ACK to Current sensor node, now indicate the correctness of prediction.Current sensor node is after ACK is received, by enlivening shape State is transformed into resting state, and next sensor continues to track as current sensor.
If current sensor is not received by ACK within second X period, illustrate prediction of failure, start and recover Mechanism, while prediction of failure information is sent to leader cluster node by current sensor node.
The process for starting Restoration Mechanism comprises the following steps:
The Loss Rate during mobile object tracking is calculated, it is described current when the Loss Rate is less than threshold value The prediction of failure information is sent to corresponding leader cluster node by sensor, then activates the backward of the current sensor node Objective sensor node all in sequence is relied on, subsequently into resting state;
If the objective sensor node traces into the mobile object, the sensor section of the mobile object is traced into Point sends confirmation to the leader cluster node;
The leader cluster node updates the rear to dependence of the sensor node of tracking failure after the confirmation is received Sequence;
According to the backward dependence sequence after renewal, using the sensor node for tracing into the object as working as prosthomere Point, perform step S431;
If all objective sensor nodes all do not trace into the mobile object, start flooding mechanism, i.e., All sensor nodes are activated as source node using the leader cluster node, move the tracking of object;
The sensor node for tracing into the object is sent into confirmation to the leader cluster node;
The leader cluster node updates the rear to dependence of the sensor node of tracking failure after the confirmation is received Sequence;
According to the backward dependence sequence after renewal, using the sensor node for tracing into the object as working as prosthomere Point, perform step S431;
In addition, when the Loss Rate is more than threshold value, i.e., when Loss Rate is higher, the leader cluster node is by the backward dependence of renewal Sequence is sent to corresponding sensor node, and each corresponding sensor node updates according to the backward dependence sequence of reception The backward dependence sequence of oneself, then perform step S41.
Fig. 3 shows the structured flowchart of the Target Tracking System based on prediction in radio sensing network of the invention.
Reference picture 3, according to another aspect of the present invention, there is provided in a kind of radio sensing network based on the target of prediction with Track system, this, which states system, includes:
Sub-clustering unit 10, for carrying out sub-clustering to all the sensors node in radio sensing network;
History mobile message collector unit 20, for all the sensors node after sub-clustering to be placed in into active state, collect The history mobile route information of mobile target;
The backward sequence that relies on establishes unit 30, for establishing all sensors according to the history mobile route information The backward dependence sequence of node, and the backward sequence that relies on is stored to corresponding leader cluster node;
Object tracking unit 40, in each cluster according to after described to relying on sequence control sensor node State moves the tracking of object.
Following table is to be tracked the comparison information table with existing method to object using the method for the present invention.
Method name Complexity Whether multiclass is supported to track Whether time correlation
PTSP O(3N) It is no It is no
Apriori O(N2) It is no It is no
MLS O(N2) It is no It is no
MOTA Influence factor is more, and complexity is high It is It is no
TMP-Mine O(N2) It is It is
ASM O(N2) It is It is
The method of the present invention O(N+n) It is It is no
In upper table, PSPT (Prediction-based Tracking Technique using Sequential Patterns) it is the sequence pattern tracking technique based on prediction;
MLS (multi-level structure) is sandwich construction tracking;
((Multi-model based Object Tracking Architecture) is based on tracking structure to MOTA Multi-mode tracking method;
TMP-Mine (Temporal Movement Patterns-Mine) is real-time mobile patterns mining method;
ASM(adaptive schedule monitoring):It is adaptive tracking method;
By finding out in table, compared to other congenic methods, method overall energy of the invention consumption is less, tracks Loss Rate It is smaller, and complexity is relatively low, supports multiclass tracking, and the time being traced to independent of object.
In order to verify the performance of institute's method of the present invention, recovered by emulation experiment from total energy consumption, tracking Loss Rate, participation The aspect of number of sensors three compares of the invention with existing method.
Existing method is as follows:
IMP (improved-Mining Patterns) is the method for tracking target based on prediction of the present invention.
OTDM (object tracking using data mining mechanism) is the object based on data mining Tracking.
The major experimental environment of the present embodiment is Visual Studio and MATLAB, wherein sensing node number quantity Level is 100-700, is uniformly distributed in bulk portion, is divided into Cn cluster, the communication between sensor node and leader cluster node Based on multi-hop shortest-path method.The communication range of sensor node is 20m, and what object stopped on each sensing node is averaged Time T=1s.Energy resource consumption is emulated using Rockwell WINS.Design parameter sets as follows:
Fig. 4 shows in the radio sensing network of the present invention method for tracking target based on prediction with existing method in total energy Comparison figure in terms of the consumption of source;
As shown in figure 4, the method IMP energy consumptions of the present invention are smaller compared with PTSP, OTDM, MLS.Wherein PTSP and OTDM are Forecasting Methodology, without Restoration Mechanism, acquiescence uses flooding mechanism.PTSP method prediction of failure rates are higher, start flooding method Often, so PTSP method energy consumptions are higher.Restoration methods use flooding mechanism in OTDM methods, therefore energy consumption is compared with IMP height A lot.Sensor is layered by MLS methods, is predicted and is recovered according to hierarchical relationship, and energy consumption is relatively low, but also above IMP. The inventive method IMP reduces communication energy consumption by sub-clustering mechanism, is recovered first according to the dependence excavated, its It is secondary to ensure that zero loses using flooding method, therefore energy consumption is relatively low.
Fig. 5 is shown in the radio sensing network of the present invention to be lost based on the method for tracking target of prediction and existing method Comparison figure in terms of rate;
As shown in figure 5, the method IMP Loss Rates of the present invention are suitable with OTDM, other two methods are respectively less than.PTSP methods Mainly according to the front and rear prediction for relying on sequence and carrying out next-hop, but some sensors may rely on identical but forward direction backward Difference, prediction of failure is now will result in, so PTSP method Loss Rates are higher.The final Result of OTDM methods and the present invention It is identical, therefore Loss Rate is almost identical with IMP.Sensor is layered by MLS methods, is predicted according to hierarchical relationship, is lost Rate is relatively low, but also above IMP.IMP of the present invention quick and precisely realizes prediction using backward rely on, and Loss Rate is relatively low.
Fig. 6 is shown in the radio sensing network of the present invention and participated in based on the method for tracking target of prediction and existing method The comparison figure of the quantitative aspects of the sensor node of recovery;
As shown in fig. 6, the sensor node number participated in recovery process with other three kinds of methods from the present invention can be with Find out, the number of sensors that the present invention participates in is minimum, and PTSP and OTDM are required for whole biographies every time because directly using flooding mechanism Sensor participates in recovering, therefore the sensor node number participated in is most.
Method for tracking target and its system based on prediction in the radio sensing network of the present invention, it need to only be based on backward rely on Object tracking can be carried out, greatly reduces time complexity, and improved in terms of accuracy;Meanwhile by dividing Cluster reduces the long haul communication between sensor node and base station, and prediction result is stored in into sensor node, reduces Interacting between sensor and cluster head, in path during rule changes locally, sub-clustering mechanism can not influence other parts On the premise of fast and convenient realization predict again;In addition, when tracking failure, restoration methods of the invention are not simple Flooding mechanism, but recovered first according to the path rule excavated, if still failing, then carry out global flooding. By this method, realization that can be quick and easy recovers, and significantly reduces tracking Loss Rate;In addition, this method supports multiclass Object tracks simultaneously.
Although being described in conjunction with the accompanying embodiments of the present invention, those skilled in the art can not depart from this hair Various modifications and variations are made in the case of bright spirit and scope, such modifications and variations are each fallen within by appended claims Within limited range.

Claims (9)

1. the method for tracking target based on prediction in a kind of radio sensing network, it is characterised in that methods described includes:
S1, sub-clustering is carried out to all the sensors node in radio sensing network;
S2, all the sensors node after sub-clustering is placed in active state, collects the history mobile route information of mobile object;
S3, the backward dependence sequence for establishing according to the history mobile route information all sensor nodes, and will described in Backward to rely on sequence storage leader cluster node corresponding to, the backward sequence that relies on is seq [now, des], wherein, now is represented The information of current sensor node, des represent the information of objective sensor node;
S4, in each cluster according to after described to the state for relying on sequence control sensor node carry out the mobile object with Track.
2. the method for tracking target based on prediction in a kind of radio sensing network according to claim 1, it is characterised in that The step S1 is specifically included:Sub-clustering is carried out to the sensor node using K-means algorithms.
3. the method for tracking target based on prediction in a kind of radio sensing network according to claim 2, it is characterised in that The step S2 also includes:By the history mobile route information classification and store.
4. the method for tracking target based on prediction in a kind of radio sensing network according to claim 1, it is characterised in that The step S4 is specifically included:
S41, calculate cluster in each sensor node backward dependence sequence in objective sensor node confidence level, and will All objective sensor nodes are arranged in one-dimension array according to confidence level descending;
S42, stored the content of the one-dimension array as information of forecasting into corresponding sensor node;
S43, when moving object tracking, under the information of forecasting that current sensor node is stored according to the node activates One sensor node, realize that mobile object tracks.
5. the method for tracking target based on prediction in a kind of radio sensing network according to claim 4, it is characterised in that The step S43 is specifically included:
S431, after the current sensor node is activated, start monitor tracking area in whether there is mobile object;
S432, when tracking the period in trace into mobile object, then tracking result is reported to a upper sensor node;
S433, next sensor node is predicted according to the information of forecasting of storage in preset time period, and active information is sent out It is sent to next sensor node of prediction;
S434, the next sensor node perform step S431, the current sensor node receive it is described next Enter resting state after the tracking result of sensor node, otherwise perform next step;
When S435, the current sensor are not received by the tracking result of next sensor node, then the mobile object Tracking failure, starts Restoration Mechanism, recovers the tracking of the mobile object.
6. the method for tracking target based on prediction in a kind of radio sensing network according to claim 5, it is characterised in that The step S435 is specifically included:
Calculate the Loss Rate during mobile object tracking, when the Loss Rate is less than threshold value, the current sensor The prediction of failure information is sent to corresponding leader cluster node by device, then activates the backward dependence of the current sensor node All objective sensor nodes in sequence, subsequently into resting state;
If the objective sensor node traces into the mobile object, trace into the sensor node of the mobile object to The leader cluster node sends confirmation;
The leader cluster node updates the rear to dependence sequence of the sensor node of tracking failure after the confirmation is received;
According to the backward dependence sequence after renewal, using the sensor node for tracing into the object as present node, hold Row step S431.
7. the method for tracking target based on prediction in a kind of radio sensing network according to claim 6, its feature exist In the step S435 also includes:
If all objective sensor nodes all do not trace into the mobile object, saved by source of the leader cluster node Point activates all sensor nodes, moves the tracking of object;
The sensor node for tracing into the object is sent into confirmation to the leader cluster node;
The leader cluster node updates the rear to dependence sequence of the sensor node of tracking failure after the confirmation is received;
According to the backward dependence sequence after renewal, using the sensor node for tracing into the object as present node, hold Row step S431.
8. the method for tracking target based on prediction in a kind of radio sensing network according to claim 7, it is characterised in that The step S435 also includes:
When the Loss Rate is more than threshold value, the backward dependence sequence of renewal is sent to corresponding sensor by the leader cluster node Node, each corresponding sensor node update the backward dependence sequence of oneself according to the backward dependence sequence of reception, so Step S41 is performed afterwards.
9. the Target Tracking System based on prediction in a kind of radio sensing network, it is characterised in that the system includes:
Sub-clustering unit, for carrying out sub-clustering to all the sensors node in radio sensing network;
History mobile message collector unit, for all the sensors node after sub-clustering to be placed in into active state, collect motive objects The history mobile route information of body;
The backward sequence that relies on establishes unit, for establishing all sensor nodes according to the history mobile route information It is backward to rely on sequence, and the backward sequence that relies on is stored to corresponding leader cluster node, the backward sequence that relies on is seq [now, des], wherein, now represents the information of current sensor node, and des represents the information of objective sensor node;
Object tracking unit, for entering in each cluster according to after described to the state for relying on sequence control sensor node The tracking of the row mobile object.
CN201410693468.3A 2014-11-26 2014-11-26 Method for tracking target and its system based on prediction in radio sensing network Active CN104469875B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410693468.3A CN104469875B (en) 2014-11-26 2014-11-26 Method for tracking target and its system based on prediction in radio sensing network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410693468.3A CN104469875B (en) 2014-11-26 2014-11-26 Method for tracking target and its system based on prediction in radio sensing network

Publications (2)

Publication Number Publication Date
CN104469875A CN104469875A (en) 2015-03-25
CN104469875B true CN104469875B (en) 2018-01-12

Family

ID=52915063

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410693468.3A Active CN104469875B (en) 2014-11-26 2014-11-26 Method for tracking target and its system based on prediction in radio sensing network

Country Status (1)

Country Link
CN (1) CN104469875B (en)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105629225B (en) * 2015-12-30 2018-05-11 中国人民解放军信息工程大学 A kind of more hypothesis method for tracking target based on improvement K shortest paths
CN105451182B (en) * 2016-01-04 2018-11-27 北京科技大学 A kind of target identification tracking system and method
CN107205255A (en) * 2017-05-15 2017-09-26 中国科学院上海微系统与信息技术研究所 Towards the method for tracking target of the wireless sensor network based on imaging sensor
CN108152790A (en) * 2018-01-05 2018-06-12 燕山大学 A kind of non-cooperation multi-target traces projectional technique based on distributed structure/architecture
CN109219118B (en) * 2018-11-27 2021-08-24 上海交通大学 Distributed consistent state estimation method based on sequence awakening mechanism
CN109890002B (en) * 2019-01-04 2022-08-16 中国电子科技集团公司电子科学研究院 Method and device for predicting environmental risk based on knowledge reasoning

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101251593A (en) * 2008-03-31 2008-08-27 中国科学院计算技术研究所 Method for tracking target of wireless sensor network
CN101644758A (en) * 2009-02-24 2010-02-10 中国科学院声学研究所 Target localization and tracking system and method
US8805008B1 (en) * 2011-11-02 2014-08-12 The Boeing Company Tracking closely spaced objects in images

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101251593A (en) * 2008-03-31 2008-08-27 中国科学院计算技术研究所 Method for tracking target of wireless sensor network
CN101644758A (en) * 2009-02-24 2010-02-10 中国科学院声学研究所 Target localization and tracking system and method
US8805008B1 (en) * 2011-11-02 2014-08-12 The Boeing Company Tracking closely spaced objects in images

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
An Energy Efficient Technique for Object Tracking in Wireless Sensor Networks;Thangarajan et. al.;《Communication Systems and Network Technologies(CSNT)》;20130408;全文 *
基于无线传感器网络的定位跟踪技术研究;周伟;《中国博士学位论文全文数据库信息科技辑》;20130515(第5期);全文 *

Also Published As

Publication number Publication date
CN104469875A (en) 2015-03-25

Similar Documents

Publication Publication Date Title
CN104469875B (en) Method for tracking target and its system based on prediction in radio sensing network
CN103716867B (en) Based on event driven wireless sensor network multi-target real-time tracking system
Tashtarian et al. On maximizing the lifetime of wireless sensor networks in event-driven applications with mobile sinks
Bhuiyan et al. Local area prediction-based mobile target tracking in wireless sensor networks
CN101505532B (en) Wireless sensor network target tracking method based on distributed processing
CN101534470B (en) System and method for tracking moving target based on wireless sensor network
CN103354642B (en) A kind of method improving mobile sensor network coverage rate
Yeow et al. Energy efficient multiple target tracking in wireless sensor networks
CN110989352B (en) Group robot collaborative search method based on Monte Carlo tree search algorithm
Keally et al. Watchdog: Confident event detection in heterogeneous sensor networks
Ma et al. Energy efficiency on location based applications in mobile cloud computing: a survey
Altman et al. Combined optimal control of activation and transmission in delay-tolerant networks
CN102883429A (en) Method and device for tracking move object in sensor network based on directional antenna
CN108513250A (en) A kind of trajectory track method and system based on pattern-recognition
Hsu et al. POOT: An efficient object tracking strategy based on short-term optimistic predictions for face-structured sensor networks
Lin et al. Hybrid charging scheduling schemes for three-dimensional underwater wireless rechargeable sensor networks
Deldar et al. Designing a prediction-based clustering algorithm for target tracking in wireless sensor networks
CN107071848A (en) A kind of target tracking algorism based on cluster structured reduction energy consumption
CN106209261B (en) The mobile data collection method of three-dimensional UASNs based on probability neighborhood grid
CN103260238A (en) Mobile node positioning method based on speed sampling zone in wireless sensor network
Guo et al. A clustering-based improved Grey-Markov target tracking algorithm in wireless sensor networks
Kiani et al. Target Tracking Based on base Station Node Using Prediction Method and Cluster Structure in Wireless Sensor Networks
Xu et al. Dynamic mobile agent migration in Wireless Sensor Networks
Sakthivel et al. An energy efficient technique for object tracking in wireless sensor networks
Mansur et al. Survey of prediction algorithms for object tracking in wireless sensor networks

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant