International Journal of Computer Application
Issue 5, Volume 1 (Jan.- Feb. 2015)
Available online on http://www.rspublication.com/ijca/ijca_index.htm
ISSN: 2250-1797
A Wireless Sensor Networks with Scattered Transform of
Top-K Queries
N.Sasikala
(M.Tech),Department of computer science&Engineering,Madanapalli Institute of
Technology& Science,Madanapalle,Chittor,AP
ABSTRACT
Distributed networks (WSNs) consist of spatially dispersed identity of directed sensors to
observe the normal condition. Top-k new approach such as expected position technique is used
to get precise results in these Top-k results is exact possibility. The inter cluster processing are
used in three algorithms .it is used Sufficient-set based and necessary-set based, Boundary based
algorithm. For reducing transmission cost adaptive algorithm is used to give competent results in
restricted round of communications. It gives transmit cost and not exceed two round of
messages. Both two tier hierarchical and tree structured network topologies are wireless
networks.
.
Key words: Data reliability, reliable distributed space, fault localization, data vibrant.
INTRODUCTION
Data mining is the procedure of discover helpful information from a collection of
data. Data mining is withdrawal of secret predicatively information great databases. KDD in
database is non-trivial method of identify valid, earlier unknown and potentially helpful pattern
in data.
R S. Publication, rspublicationhouse@gmail.com
Page 129
International Journal of Computer Application
Issue 5, Volume 1 (Jan.- Feb. 2015)
Available online on http://www.rspublication.com/ijca/ijca_index.htm
ISSN: 2250-1797
The process of data mining is that mining of data or pattern from data in large
database. This process is known as knowledge discovery process. Data mining plays an essential
role in the KDD.
Knowledge discovery (KDD) process:
The following figures show in development of knowledge discovery (KDD).
Fig.1 Knowledge Discovery (KDD) process
Knowledge discovery (KDD) is the step by step process for extracting knowledge from
large databases.
Steps in KDD Process:
The following are the steps in Knowledge discovery (KDD) process.
1) Data Cleaning
2) Selection
3) Mining
4) Reduction
5) Transformation
6) Pattern evaluation
Data Cleaning is the process of detecting and correcting data records from the database.
Data Selection is the process selection of the data from the database for mining.
R S. Publication, rspublicationhouse@gmail.com
Page 130
International Journal of Computer Application
Issue 5, Volume 1 (Jan.- Feb. 2015)
Available online on http://www.rspublication.com/ijca/ijca_index.htm
ISSN: 2250-1797
Data Mining is process that selected data is to mine from the database.
Data reduction and transformation
Find positive features, invariant representation and dimensionality/variable reduction. In
Data transformation the process is choosing functions of data mining as classification,
summarization, regression, association and clustering.
Pattern evaluation is last step in KDD process, in this process Patterns and knowledge to be
used or stored into knowledge-base
Architecture of a Data Mining System
Fig 2: Data mining Architecture
Data mining system can separately mine out all of the expensive knowledge from a
known large database, without person intervention
REVIEW OF RELATED RESEARCH
A.) A cross pruning framework for top-k data collection in wireless sensor networks
Wireless sensor networks main problem is due to maintain of energy. In this work, we
can explore the methods of in-network collection for answering the queries A top-k request
recovers the k data items with the maximum results estimated by a recording role on concerned
features of antenna evaluations. Our study shows that existing techniques for handling top-k
request, e.g., Tiny Aggregation Service (TAG), are not energy effective due to deficits in their
direction-finding organizations and statistics accretion mechanisms. To address these deficits, we
propose to develop a new cross pruning (XP) aggregation framework for top-k data collection in
wireless sensor networks. The XP framework incorporates several novel ideas to facilitate
R S. Publication, rspublicationhouse@gmail.com
Page 131
International Journal of Computer Application
Issue 5, Volume 1 (Jan.- Feb. 2015)
Available online on http://www.rspublication.com/ijca/ijca_index.htm
ISSN: 2250-1797
efficient in-network aggregation and filtering, including 1) structure a cluster-tree routing
arrangement to aggregate more itemsclosey. 2) implementing a broadcast-then-filter
methodology for professionally defeating terminated information communications; and 3)
providing a cross pruning procedure to enhance in-network filtering effectiveness. An extensive
set of experiments based on simulation has been conducted to evaluate the performance of TAG
and the proposed XP framework. The experimental results validate our proposals and show that
XP significantly outperforms TAG in energy charge.
B.) Ranking with uncertain scores
Great data banks with provisional data are suitable in various present procedures contain
data combination, position tracking, and network search. In this presentation, we will handle
recent situation with uncertain items that are various from traditional classification. Specially,
improbability in documentations slash brings a limited instruction concluded proceedings, as
opposed to the entire instruction that is expected in the conventional dynamic surroundings. In
this work, we existent a new probabilistic design, based on limited orders, to summarize the
universe of probable positions initiates from score improbability. In this design we can develop
various ranking request kinds with various rules. We designate and evaluate a position of
proficient query estimation algorithms. We prove that our techniques can be used to solve the
problem of level aggregation in limited instructions. In calculation, we propose original case
techniques to divide near query answer. Our unsure estimate uses both actual and artificial
information. The provisional training validates the efficacy and usefulness of our skills in
unrelated settings.
EXISTING SYSTEM
Two schemes, additional evidence and crosscheck, were proposed in as solutions for securing
top-k query in tiered sensor networks. While the former generates hashes for each consecutive
pair of sensed data for verification purpose, the latter performs network-wide broadcast such that
the information about the readings is distributed over the entire network and therefore the query
result cannot be manipulated. In particular, the idea behind additional evidence is that if each
consecutive pair of sensed data is associated with a hash, once an unqualified sensor reading is
used to replace the genuine query result, the authority may know because it can find that there
are some missing sensor readings for hash verification. On the other hand, the idea behind
crosscheck is that the genuine top-k results are distributed to several sensor nodes. With certain
R S. Publication, rspublicationhouse@gmail.com
Page 132
International Journal of Computer Application
Issue 5, Volume 1 (Jan.- Feb. 2015)
Available online on http://www.rspublication.com/ijca/ijca_index.htm
ISSN: 2250-1797
probability, the authority will find query result incompleteness by checking the other sensor
nodes’ sensor readings. Hybrid method is a combined use of additional evidence and crosscheck,
attempting to balance the communication cost and the query result incompleteness detection
capability.
DISADVANTAGES OF EXISTING SYSTEM
1. There is no trusted central authority like proxy node in for such responsibility.
2. In real world deployment, these requirements are difficult to meet.
3. The methods do not handle the data privacy issue.
PROPOSED SYSTEM
The Verifiable top-k Query (VQ) schemes based on the novel dummy reading-based
anonymization framework are proposed for privacy preserving top-k query result integrity
verification in tiered sensor networks. A randomized and distributed version of Order Preserving
Encryption, rdOPE, is proposed to be the privacy foundation of our methods.AD-VQ-static
achieves the lower communication complexity at the cost of slight detection capability
degradation, which could be of both theoretical and practical interests. Analytical studies,
numerical simulations, and prototype implementation are conducted to demonstrate the
practicality of our proposed methods. A cell is a connected multihop network composed of a
storage node and a number of ordinary sensors. Storage nodes are storage-abundant, can
communicate with the authority via direct or multi-hop communications, and are assumed to
know their affiliated cells. Time on the nodes has been synchronized and is divided into epochs.
Note that time synchronization among nodes can be achieved by using algorithms.
ADVANTAGES OF PROPOSED SYSTEM
1. The message m to be communicated is associated with HMACK; the use of HMAC naturally
guarantees the data authenticity and integrity.
2. Hybrid Crosscheck incurs tremendous communication cost because it involves the data
broadcast over the cell.
3. SMQ achieves the data confidentiality through the use of bucket index
R S. Publication, rspublicationhouse@gmail.com
Page 133
International Journal of Computer Application
Issue 5, Volume 1 (Jan.- Feb. 2015)
Available online on http://www.rspublication.com/ijca/ijca_index.htm
ISSN: 2250-1797
4. IMPLEMENTATION
PT TOPK QUERY PROCESSING:
PTTOP K-queries provide with good conditions for the plagued discrete exclusion
problem. Query answer container exist attain by inquiring the tuples in sliding level organize
starting the sorted table it is still denoted plainness of a T. it can simply close to the top rank Ktuples completely in the react set as extended since their confidence are more to p as their
experience as PT-TOP K answer are not reliant on the subsistence of more than a few further
tuples.
SENSOR NETWORKS:
The partial energy resources accessible at sensors nodes, the main issues are enhanced
energy efficient technique to decrease infrastructure and force expenses in the network.
Estimated it base data aggregation technique .We can planned. The main scheme is to trade-off
some data feature for enhance force competence. That it base on numerical model technique a
form motivated advance be planned to balanced the assurance of faults respond adjacent to the
report charge in the networks.
R S. Publication, rspublicationhouse@gmail.com
Page 134
International Journal of Computer Application
Issue 5, Volume 1 (Jan.- Feb. 2015)
Available online on http://www.rspublication.com/ijca/ijca_index.htm
ISSN: 2250-1797
DATA PRUNING:
The Cluster head are reliable for produce insecure tuples of a data from the gather
unprocessed sensors reading inside their cluster. it react a uncertainty, it is normal for the clusters
head in the direction of cut down lay off undecided tuples of a data facing release to the
position. It organizes to decrease statement and force charge. The input substance here is to
obtain a compressed set of tuple necessary for the answer to the base station to the probabilistic
TOP K- queries.
STRUCTURAL NETWORK TOPOLOGY:
The prompt of a network query processing a routing tree is regularly created amongst
sensor nodes and the base station. A query is issued at the root of the routing tree and spread
down to the tree of a sufficient set of necessary and position introduce previous based on two-tier
hierarchical sensor networks (HSN). They are appropriate to tree-structured sensor network
(SSN).
Fig: Structural network topology
DATA TRANSMISSION:
Response time is one more significant metrics to asses query processing algorithms in
WSNs. All of those three algorithms like NSB, SSB, and BB in terms of query response time.
Thus we focus on the data transmission cost in the valuation. Finally, evaluate algorithms are
conducting the experiments. Tree structural network is a under the SSB-T, NSB-OPT and NSBT under.
R S. Publication, rspublicationhouse@gmail.com
Page 135
International Journal of Computer Application
Issue 5, Volume 1 (Jan.- Feb. 2015)
Available online on http://www.rspublication.com/ijca/ijca_index.htm
ISSN: 2250-1797
Fig : Data transmission
5. EXPERIMENTS AND RESULTS
Inside this assessment segment certain illustrated outcomes regarding referral excellence and
efficiency are envisioned.
Fig 5.1 denoting base station
Fig 5.2 welcome
R S. Publication, rspublicationhouse@gmail.com
page
Page 136
International Journal of Computer Application
Issue 5, Volume 1 (Jan.- Feb. 2015)
Available online on http://www.rspublication.com/ijca/ijca_index.htm
ISSN: 2250-1797
Fig 5.3 Updating page
Fig:5.4 Request for filespage
R S. Publication, rspublicationhouse@gmail.com
Page 137
International Journal of Computer Application
Issue 5, Volume 1 (Jan.- Feb. 2015)
Available online on http://www.rspublication.com/ijca/ijca_index.htm
ISSN: 2250-1797
Fig 5.5. MAIL receivers page
CONCLUSION:
We evaluate the performance of our newly proposed algorithm with that of existing approach
such as raw and iterative method. Both synthetic data and real data sets are used for performance
evaluation while estimating performance of iterative approach it takes more than 50-250 rounds
of communication at one round of communication only one data tuples is send to the base
station. So, it takes many rounds of communication to complete the process. Comparing our
newly proposed algorithms complete within the two rounds. Then the adaptive algorithm gives
the least transmission cost.
FUTURE WORK:
We proposed every web application has its own metrics and demerits. We increase a query loadbased spanning tree creation technique. It reduces the query response interruption as well as
energy spending in query execution and provides query response with the best possible
correctness.
REFERENCE
[1]. Y. Xu, W.-C. Lee, J. Xu, and G. Mitchell, “Processing Window Queries in Wireless Sensor
Networks,” Proc. IEEE 22nd Int’l Conf. Data Eng. (ICDE ’06), 2006.
[2]. M. Ye, X. Liu, W.-C. Lee, and D.L. Lee, “Probabilistic Top-k Query Processing in
Distributed Sensor Networks,” Proc. IEEE Int’l Conf. Data Eng. (ICDE ’10), 2010.
[3]. D. Zeinalipour-Yazti, Z. Vagena, D. Gunopulos, V. Kalogeraki, V.Tsotras, M. Vlachos, N.
Koudas, and D. Srivastava, “The ThresholdJoin Algorithm for Top-k Queries in Distributed
Sensor Networks,” Proc. Second Int’l Workshop Data Management for Sensor Networks
(DMSN ’05), pp. 61-66, 2005.
R S. Publication, rspublicationhouse@gmail.com
Page 138
International Journal of Computer Application
Issue 5, Volume 1 (Jan.- Feb. 2015)
Available online on http://www.rspublication.com/ijca/ijca_index.htm
ISSN: 2250-1797
[4]. A. Sharaf, J. Beaver, A. Labrinidis, and K. Chrysanthis, “Balancing Energy Efficiency and
Quality of Aggregate Data in Sensor Networks,” Int’l J. Very Large Data Bases, vol. 13, no. 4,
pp. 384-403,
2004.
[5]. A.S. Silberstein, R. Braynard, C. Ellis, K. Munagala, and J. Yang, “A Sampling-Based
Approach to Optimizing Top-k Queries in Sensor Networks,” Proc. 22nd Int’l Conf. Data Eng.
(ICDE ’06), p. 68, 2006.
[6] . Q. Han, S. Mehrotra, and N. Venkatasubramanian, “Energy Efficient Data Collection in
Distributed Sensor Environments,” Proc. 24th Int’l Conf. Distributed Computing Systems
(ICDCS ’04), pp. 590-597, 2004.
[7]. M. Wu, J. Xu, X. Tang, and W.-C. Lee, “Top-k Monitoring in Wireless Sensor Networks,”
IEEE Trans. Knowledge and Data Eng., vol. 19, no. 7, pp. 962-976, July 2007.
[8]. D. Wang, J. Xu, J. Liu, and F. Wang, “Mobile Filtering for Error-Bounded Data Collection
in Sensor Networks,” Proc. 28th Int’l Conf. Distributed Computing Systems (ICDCS ’08), pp.
530-537, 2008.
[9].K. Yi, F. Li, G. Kollios, and D. Srivastava, “Efficient Processing of Top-k Queries in
Uncertain Databases with X-Relations,” IEEE Trans. Knowledge and Data Eng., vol. 20, no. 12,
pp. 1669-1682, Dec. 2008.
[10] J. Li, B. Saha, and A. Deshpande, “A Unified Approach to Ranking in Probabilistic
Databases,” Proc. Int’l Conf. Very Large Data Bases (VLDB), vol. 2, no. 1, pp. 502-513, 2009.
[11] .X. Lian and L. Chen, “Probabilistic Ranked Queries in Uncertain Databases,” Proc. 11th
Int’l Conf. Extending Database Technology (EDBT ’08), pp. 511-522, 2008.
R S. Publication, rspublicationhouse@gmail.com
Page 139