CN101917730B - Scene adaptive energy balance-based sensor network vector quantization clustering method - Google Patents
Scene adaptive energy balance-based sensor network vector quantization clustering method Download PDFInfo
- Publication number
- CN101917730B CN101917730B CN2010102279336A CN201010227933A CN101917730B CN 101917730 B CN101917730 B CN 101917730B CN 2010102279336 A CN2010102279336 A CN 2010102279336A CN 201010227933 A CN201010227933 A CN 201010227933A CN 101917730 B CN101917730 B CN 101917730B
- Authority
- CN
- China
- Prior art keywords
- node
- zone
- energy
- representation
- network
- 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.)
- Expired - Fee Related
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses a scene adaptive energy balance-based wireless sensor network vector quantization clustering method. The method comprises a method for determining an optimum code book by an exclusive method when a convergent code is far away from an event source, a method for determining the optimum code book by a weighing method when the convergent code is near the event and a method for screening out the least representative node number by an iterative method. In the method, the least representative node number and the particular position thereof are screened out by taking a residual energy factor of the node into full consideration, utilizing the space correlation properties of the sensor network and adopting a residual energy-based vector quantization method, so that the nodes in an event source perceptual area can cooperate with one another and the aim of balancing the network energy is fulfilled; and the problem that because the energy is too little and the node is selected to be the representative node repeatedly, part of the nodes die untimely, so that the life cycle of the network is greatly prolonged and the resource utilization ratio is greatly improved.
Description
Technical field
The present invention relates to wireless communication technology field, is a kind of geographical position of not only having considered node in the overall network but also with respect to the vector quantization clustering method of individual node dump energy.This method is mainly used in the wireless sensor network, can prolong the life cycle of network effectively.
Background technology
The network system of monitoring, control and radio communication that sensor network has been integrated.Because most of sensor node volume is very little, and is distributed in the open air, is difficult to it and changes battery, so sensor node has all ten minutes features of limited of energy, disposal ability, storage capacity and communication capacity.How efficiently using finite energy to come maximization network life cycle is the overriding challenge that sensor network faces.
The typical sensor network all can adopt the space dense distribution in order to eliminate the cavity of overlay area.Therefore, a plurality of sensor nodes in the same Perception Area can perceive same incident simultaneously, node perceived to data have very big correlation and redundancy.In extensive, intensive wireless sensor network based on chance event; The data of transmission mostly have the spatial coherence of height; A lot of detected data similarity degree of sensor node are very big; Only need a part of node that information is sent to aggregation node, just can monitor the generation of incident.
Usually adopt the cluster structured network topology of optimizing in the moment sensor network,, improve energy efficiency and extension of network property with balance network load.In cluster structured; Node in the network is divided into several node set that is called bunch; Each bunch is made up of a leader cluster node and a plurality of member node usually, and the work of bunch member node is in charge of and is controlled to bunch head, in being responsible for simultaneously bunch data collection and bunch between data forwarding.Literature search through to prior art is found; Traditional classical sub-clustering algorithm all is on energy and time delay, to weigh when sub-clustering; Basically do not take the spatial correlation characteristic between the node into account, and only in follow-up leader cluster node is collected and merged bunch, considered spatial coherence during data.LEACH (low energyadaptive clustering hierarchy) the sub-clustering method that people such as Wendi B.Heinzelman propose is the most representative.This method is served as a bunch head equiprobably through each node, thereby reaches the equalizing network energy consumption, prolongs the purpose of network life.But bunch head selects to cause a bunch skewness at random among the LEACH, and the node energy factor is not considered in the selection of bunch head simultaneously, thereby causes the part of nodes premature dead; And bunch head adopts the single-hop mode to communicate by letter with aggregation node, and is very fast away from a bunch energy consumption of aggregation node.GAF (the geographical adaptive fidelity) algorithm that people such as Xu Y propose is to be the sub-clustering algorithm of foundation with the node geographical position, has proposed the thought by dummy unit lattice cluster dividing district.This algorithm is not considered the dump energy factor of node yet.Though there are a lot of scientific research personnel that above two kinds of classic algorithm are improved afterwards; Also introduced capacity factor; But these algorithms all do not make full use of the sensor network spatial correlation characteristic, work such as leader cluster node need accomplish data fusion, communicate by letter with aggregation node, and the energy of consumption is big.
In Sift MAC agreement, let the node in the relevant range utilize race mechanism to select leader cluster node, but it does not specify how to divide the relevant range.In Image Compression, the vector quantization coding technology is the position that is used under the distortion limit, finding out interdependent node.Vector quantization has effectively utilized the statistical redundancy degree that four kinds of correlations of each component (shape of linear dependence, non-linear dependence, probability density function and vector dimension etc.) in the vector are eliminated data.The relevant range is divided and the problem of leader cluster node screening so this method can be performed well in solving.
Because leader cluster node needs and bunch district in node manage and communicating by letter of controlling, also to carry out the collection and the fusion of data in bunch district, a lot of energy of requirement consumes.Therefore, people such as Mehmet C.Vuran proposes to filter out with iteration node selection algorithm (INS) number and the geographical position of representation node.Unique difference of representation node and leader cluster node is that representation node need not to carry out data fusion, and each bunch district only lets representation node transmission self data monitored to aggregation node.The selection of representation node has adopted the Vector Quantization algorithm in the image encoding to realize.
The INS algorithm has only been considered the spatial coherence between the node, does not take all factors into consideration the capacity factor of network.Because there is inhomogeneous defective in the network topology of random distribution, the INS algorithm possibly cause the less node of energy to repeat to elect as representation node, causes the phenomenon that occurs the part of nodes premature dead in the network, has shortened network life greatly.
Summary of the invention
The objective of the invention is deficiency, propose a kind of vector quantization clustering method based on scene adaptive energy balance to prior art.This method makes full use of the distinctive spatial correlation characteristic of sensor network, and takes all factors into consideration the status of energy consumption of network node, to the different scenes of event source apart from distance, adopts diverse ways to confirm the particular location of representation node according to aggregation node.Be intended to filter out minimum leader cluster node number and particular location thereof, reduce the offered load amount, the energy consumption of balanced each node improves the life cycle and the resource utilization of whole network.
The present invention mainly is the Vector Quantization algorithm that is applied to image encoding through improvement, in the code selection book, introduces the residue energy of node factor, makes it become the cluster-dividing method that is applicable to wireless sensor network.According to the distance near field scape of aggregation node, design the particular location that two kinds of methods are confirmed representation node from the event source Perception Area.Specifically realize through following technical scheme:
1. scene one: when aggregation node far away from the event source Perception Area local time, adopt exclusive method to carry out confirming of representation node number and particular location.Exclusive method mainly is meant gets rid of the node that individual node energy in bunch district is lower than mean cluster district energy in iteration screening representation node; A node that only lets node energy be higher than bunch district's average energy participates in the screening process of representation node, thereby got rid of the possibility that the less node of energy is chosen as representation node.The concrete steps of this method are following:
(1) adopt certain suitable method (normally picked at random) to generate an initial codebook that comprises M code word.
(2) according to the arest neighbors method, be the center with the code word, all vectors in the training set are classified, form M zone.
(3) node of dump energy in each zone less than the zone leveling energy foreclosed, recomputate new center, each zone, and the node that decentre is nearest calculates new average distortion rate as the new code word in this zone.
(4) jump back to (2) and carry out again, up to accomplishing predefined frequency of training or average distortion measure less than certain pre-set threshold.
2. scene two: when aggregation node from nearer place, event source Perception Area; Be included in the event source Perception Area; This moment, each bunch district interior nodes arrived the bigger apart from difference of event source; From the near node perceived of event source to contain much information, if the nearer node of event source just energy be slightly less than the average energy in bunch district, as adopt above-mentioned exclusive method can not make these key nodes participate in the screening process of representation nodes.Therefore, propose the another kind of weighing method that is fit to this kind scene and filter out minimum leader cluster node and particular location thereof.The concrete steps of the weighing method are following:
(1) adopt certain suitable method (normally picked at random) to generate an initial codebook that comprises M code word.
(2) according to the arest neighbors method, be the center with the code word, all vectors in the training set are classified, form M zone.
(3) set a cost function, be defined as
D (v wherein
i, s) be in each zone each node to the distance of event source,
It is the dump energy of each node in each zone.Parameter alpha and β mainly weigh two kinds of factors at screening shared proportion during representation node.The cost function value of sensor node is more little, shows that this nodal distance event source is near more, and dump energy is many more, the suitable more representation node that becomes.Recomputate the cost function value of each node in each zone, the node that cost function value is minimum calculates new average distortion rate as the new code word in this zone.
(4) jump back to (2) and carry out again, up to accomplishing predefined frequency of training or average distortion measure less than certain pre-set threshold.
The life cycle of wireless senser is defined as the information that node reports in the network and can satisfies the network lifetime under the maximum distortion degree threshold value in the distortion factor of aggregation node in the present invention, and wherein threshold value is decided by the concrete application on sensor network upper strata.Representation node is communicated by letter through the multi-hop mode with aggregation node.The energy of aggregation node is not limited in the applied network model of the present invention, and the fixed-site of each sensor node is constant.Aggregation node is known the particular location and the primary power of each sensor node in advance.The iterative process of whole algorithm all is to accomplish at aggregation node.Because it is a lot of that the energy of processing expenditure loss lacks than transport overhead, so the communication overhead among the present invention is compared basic not increase with the INS algorithm.
Compare with above-mentioned prior art; The present invention has following beneficial effect: the present invention utilizes Vector Quantization algorithm to compress the redundancy of sensor node reported data effectively; Made full use of the spatial coherence of sensor network; And taken all factors into consideration the dump energy factor of node, and guarantee that every representation node energy of selecting of taking turns is unlikely too small, solved the problem that is repeatedly elected as the part of nodes premature dead that representation node causes owing to the less node of energy.Thereby balanced network energy and load have improved the life cycle of network greatly.
Description of drawings
Fig. 1 is the rational representation node distribution map of spatial coherence that utilizes sensor network.
Fig. 2 is the sensor node of first embodiment and the network topological diagram of aggregation node.
Fig. 3 is the sensor node of second embodiment and the network topological diagram of aggregation node.
Fig. 4 is the particular flow sheet that the present invention is based on the sensor network cluster-dividing method of scene adaptive energy balance.
Fig. 5 is the particular flow sheet of the exclusive method used in the first embodiment of the invention.
Fig. 6 is the particular flow sheet of the weighing method used in the second embodiment of the invention.
Embodiment
In order to make the object of the invention, technical scheme and advantage clearer,, the present invention is further elaborated below in conjunction with accompanying drawing and embodiment.Should be appreciated that specific embodiment described herein only in order to explanation the present invention, and be not used in qualification the present invention.
Embodiment 1:
Present embodiment is based on network topology as shown in Figure 2 and describes.The cluster-dividing method of main explanation aggregation node under event source situation far away.The application network model is specific as follows:
100 sensor nodes are randomly dispersed in the zone of 100m * 100m, and this network has following character:
Sensor network nodes is a not mobile node of static state; Aggregation node is on (150,50) position; The energy of aggregation node is not limited, and can obtain the geographical location information and the primary power information of network node.For the sake of simplicity, event source is set in this regional center (50,50).The topology distribution of sensor node and aggregation node is as shown in Figure 2.
When supposing in the event source Perception Area all nodes all reporting information is to aggregation node, the information distortion degree of aggregation node is minimum.In the process of cluster dividing district and screening representation node, except the number of representation node, also to consider two factors: the sensor node n in (1) Perception Area
iAnd the coefficient correlation between the event source S, ρ (n
i, s), the distortion factor is with node n
iDistance to event source S increases and linear growth, so should select event source S node on every side as representation node as far as possible.(2) representation node n
iAnd n
jBetween coefficient correlation ρ (n
i, n
j), the distance between the representation node is far away more, and the redundancy between the reporting information is just more little, and then the distortion factor of aggregation node is also more little.So should select the less node of correlation as far as possible as representation node.As shown in Figure 1, should select the soft dot shown in Fig. 1 as representation node as far as possible.Based on above two principles, the distortion factor that aggregation node is received information is measured with following formula:
Wherein M is the number of representation node.
The network work flow process of present embodiment is following:
(1) netinit: 100 nodes are randomly dispersed in the zone of this 100m * 100m, and all nodes give self geographical location information and primary power information reporting from self nearest aggregation node.
(2) aggregation node just can carry out the vector quantization clustering method based on scene adaptive energy balance after getting access to above information.
(3) the node number M of initialization reporting information, all reporting information is to aggregation node to let in the Perception Area all nodes, and promptly M equals total node number.The distortion factor that this moment, aggregation node received information is for minimum.
(4) with the node number M of reporting information, promptly the representation node number deducts fixed step size k, M-k, and judge that aggregation node is to the distance between the event source.This moment, aggregation node was outside the Perception Area, and is and far away apart from event source, adopts exclusive method to confirm the particular location of representation node.
(5) particular location of representation node is confirmed back calculating relevant parameter ρ (n
i, s) and ρ (n
i, n
j), and then calculate this moment aggregation node according to above-mentioned formula and receive the distortion factor of information.
(6) if the distortion factor of information less than maximum distortion degree (threshold value), then jumps to step (4), otherwise the algorithm termination, and the M+k value of getting before the algorithm termination is minimum representation node number.
The practical implementation step of exclusive method is following:
(1) adopt random device to generate an initial codebook that comprises M code word.
(2) according to the arest neighbors method, be the center with the code word, all vectors in the training set are classified, form M zone.
(3) node of dump energy in each zone less than the zone leveling energy foreclosed, recomputate new center, each zone, and the node that decentre is nearest calculates new average distortion rate as the new code word in this zone.
(4) jump back to (2) and carry out again, the difference of the relative distortion rate that draws up to front and back training during less than a certain thresholding system ε algorithm stop, filter out the particular location of the representation node that meets the demands at this moment.
Exclusive method is on the basis of Vector Quantization algorithm, to consider residue energy of node, and goes out to meet the code book and the particular location thereof of distortion rate requirement through iterative computation.At the vector quantization clustering method of carrying out on the aggregation node based on scene adaptive energy balance is the representation node that under distortion factor restriction, filters out minimum number.
Embodiment 2:
Present embodiment is based on network topology as shown in Figure 3 and describes.The cluster-dividing method of main explanation aggregation node under the nearer situation of event source.The application network model is specific as follows:
100 sensor nodes are randomly dispersed in the zone of 100m * 100m, and this network has following character:
Sensor network nodes is a not mobile node of static state; Aggregation node is on (50,50) position; The energy of aggregation node is not limited, and can obtain the geographical location information and the primary power information of network node.For the sake of simplicity, event source is set in this regional center (50,50).The topology distribution of sensor node and aggregation node is as shown in Figure 3.
Aggregation node receives the distortion factor of information and measures with following formula:
Wherein M is the number of representation node.
The network work flow process of present embodiment is following:
(1) netinit: 100 nodes are randomly dispersed in the zone of this 100m * 100m, and all nodes give self geographical location information and primary power information reporting from self nearest aggregation node.
(2) aggregation node just can carry out the vector quantization clustering method based on scene adaptive energy balance after getting access to above information.
(3) all reporting information is to aggregation node at first to let in the Perception Area all nodes, and the distortion factor that this moment, aggregation node received information is for minimum.
(4) with the node number of reporting information, promptly the representation node number deducts fixed step size k, M-k, and judge that aggregation node is to the distance between the event source.This moment, aggregation node was in the Perception Area, and it is nearer to show that aggregation node leaves event source, adopts the weighing method to confirm the particular location of representation node.
(5) particular location of representation node is confirmed back calculating relevant parameter ρ (n
i, s) and ρ (n
i, n
j), and then calculate this moment aggregation node according to above-mentioned formula and receive the distortion factor of information.
(6) if the distortion factor of information less than maximum distortion degree (threshold value), then jumps to step (4), otherwise the algorithm termination, and get algorithm termination M+k value before as minimum representation node number.
The practical implementation step of the weighing method is following:
(1) adopt random device to generate an initial codebook that comprises M code word.
(2) according to the arest neighbors method, be the center with the code word, all vectors in the training set are classified, form M zone.
(3) set a cost function, be defined as with
D (v wherein
i, s) be in each zone each node to the distance of event source,
It is the dump energy of each node in each zone.Parameter alpha and β mainly weigh two kinds of factors at screening shared proportion during representation node.Recomputate the cost function value of each node in each zone, the node that cost function value is minimum calculates new average distortion rate as the new code word in this zone.
(4) jump back to (2) and carry out again, the difference of the distortion rate that draws up to front and back training during less than a certain threshold value ε algorithm stop, filter out the particular location of the representation node that meets the demands this moment.
The weighing method is on the basis of Vector Quantization algorithm, to consider residue energy of node, and goes out to meet the code book and the particular location thereof of distortion rate requirement through iterative computation.At the vector quantization clustering method of carrying out on the aggregation node based on scene adaptive energy balance is the representation node that under distortion factor restriction, filters out minimum number.
Claims (1)
1. sensor network vector quantization clustering method based on scene adaptive energy balance, its characteristic of on aggregation node, carrying out choosing bunch process comprises:
A. let in the event source Perception Area all nodes all reporting information to aggregation node;
B. the node number with reporting information reduces k, and k is a fixed step size, and judges the distance of aggregation node to event source, if distance is far away, then adopts exclusive method to confirm the particular location of representation node; If close together then adopts the weighing method to confirm the particular location of representation node;
C. the particular location according to representation node calculates the distortion factor that aggregation node receives information;
D. whether judge the distortion factor receive information less than the maximum distortion thresholding, if then algorithm stops, and with the representation node number under the last iteration as the minimum representation node number that satisfies under the maximum distortion degree thresholding; Otherwise, jump to step B;
Its exclusive method is characterised in that; In the iterative process of carrying out each new code book in zone; The node of dump energy in each zone less than this zone leveling energy foreclosed; The node that only lets dump energy surpass or equal this zone leveling energy participates in recomputating new center, each zone, and the node that decentre is nearest is as the new code word in this zone;
Its weighing method is characterised in that, in the iterative process of carrying out each new code book in zone, sets a cost function, be defined as with
D (v wherein
i, s) be in each zone each node to the distance of event source,
Be the dump energy of each node in each zone, parameter alpha and β mainly weigh two kinds of factors at screening shared proportion during representation node; Recomputate the cost function value of each node in each zone, the node that cost function value is minimum is as the new code word in this zone.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010102279336A CN101917730B (en) | 2010-07-16 | 2010-07-16 | Scene adaptive energy balance-based sensor network vector quantization clustering method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010102279336A CN101917730B (en) | 2010-07-16 | 2010-07-16 | Scene adaptive energy balance-based sensor network vector quantization clustering method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101917730A CN101917730A (en) | 2010-12-15 |
| CN101917730B true CN101917730B (en) | 2012-11-21 |
Family
ID=43325089
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010102279336A Expired - Fee Related CN101917730B (en) | 2010-07-16 | 2010-07-16 | Scene adaptive energy balance-based sensor network vector quantization clustering method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101917730B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104618947B (en) * | 2015-02-03 | 2018-07-03 | 中国人民解放军信息工程大学 | Dynamic clustering wireless sense network method of data capture and device based on compressed sensing |
| CN108447279A (en) * | 2018-05-18 | 2018-08-24 | 上海会为智能技术有限公司 | A kind of signal management method, apparatus, equipment and medium |
| CN115473781B (en) * | 2022-11-01 | 2023-02-28 | 国网湖北省电力有限公司信息通信公司 | Distributed SNMP data center equipment monitoring method |
Citations (2)
| 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 |
| US20090285140A1 (en) * | 2005-12-01 | 2009-11-19 | Yong-Sun Kim | Method for Managing Dormant Nodes in Wireless Sensor Network |
-
2010
- 2010-07-16 CN CN2010102279336A patent/CN101917730B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090285140A1 (en) * | 2005-12-01 | 2009-11-19 | Yong-Sun Kim | Method for Managing Dormant Nodes in Wireless Sensor Network |
| 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 |
Non-Patent Citations (1)
| Title |
|---|
| 传感网分簇算法研究及其进展.胡静等.《中兴通讯技术》.2009, * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101917730A (en) | 2010-12-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108966239B (en) | Method for deploying relay nodes based on energy voidage | |
| US9730153B2 (en) | System and methods for multi-objective cell switch-off in wireless networks | |
| CN103327513A (en) | Intelligent method for collecting data | |
| CN107182063A (en) | A kind of Cluster-Based Routing Protocols for Wireless Sensor based on natural energy | |
| CN106102075A (en) | The cluster-dividing method divided based on hierarchical region in radio sensing network and system | |
| CN101917730B (en) | Scene adaptive energy balance-based sensor network vector quantization clustering method | |
| CN104284386A (en) | Vertex-betweenness-based cluster head selection method in wireless sensor networks | |
| CN105050095A (en) | Topology construction method for heterogeneous wireless sensor networks based on energy prediction | |
| Bhakare et al. | An energy-efficient grid based clustering topology for a wireless sensor network | |
| Pandana et al. | Robust connectivity-aware energy-efficient routing for wireless sensor networks | |
| Xenakis et al. | Topology control with coverage and lifetime optimization of wireless sensor networks with unequal energy distribution | |
| CN111083724B (en) | OCWG-TOPSIS-based user association method in fog computing network | |
| CN112996076B (en) | Mobile charging and data collection method in wireless sensor network | |
| Farahani | Energy Consumption Reduction in Wireless Sensor Network Based on Clustering | |
| Choi et al. | Regional clustering scheme in densely deployed wireless sensor networks for weather monitoring systems | |
| Wang et al. | An energy efficient fuzzy cluster head selection algorithm for WSNs | |
| Kim et al. | Estimation of the optimal number of cluster-heads in sensor network | |
| Adetona et al. | A cluster head assisted routing (CHAR) scheme for improved energy load balancing in wireless sensor networks | |
| Abdulsalam et al. | Data aggregation for evolving wireless sensor networks using W-LEACH | |
| Roshan et al. | Improved LEACH protocol with cache nodes to increase lifetime of wireless sensor networks | |
| Chang et al. | Not every bit counts: A resource allocation problem for data gathering in machine-to-machine communications | |
| KR101219052B1 (en) | Formation method for hierarchical cluster architure | |
| CN108934054A (en) | A kind of Ad-hoc network cluster dividing algorithm based on node data value | |
| KR101244648B1 (en) | Maintenance method for hierarchical cluster architure | |
| Enam | Energy efficient differential data aggregation in a dynamic cluster based WSN |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20121121 Termination date: 20190716 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |