Document Details

Document Type : Thesis 
Document Title :
ENHANCING DATA STREAM MINING IN WIRELESS SENSOR NETWORIKS USING CLUSTERING ALGORITHMS
تعزيز تنقيب البيانات المتدفقة في شبكات الاستشعار اللاسلكية باستخدام خوارزميات التجميع
 
Subject : Faculty of Computing and Information Technology 
Document Language : Arabic 
Abstract : The past few years have witnessed an increased interest in the potential use of Wireless Sensor Networks (WSNs) in a wide range of applications in the field of military surveillance, fire detection, habitat monitoring, industry, health monitoring and many more. WSNs consist of individual nodes that are able to interact with their environment by sensing and controlling physical parameters. Sensor nodes tend to generate a large amount of sequential small and tuple-oriented data that is considered as Data Streams. Data streams usually are huge data sets that arrive in an online fashion, flowing rapidly in a very high speed, where they are unlimited and there is no control on the arrival processing order. Due to sensor network limitations, some challenges are faced and urgently need to be solved. Such challenges include long lasting the WSN lifetime and reducing nodes energy consumption. Data mining could deal with the WSN challenges. Clustering is one of mining techniques and plays an important role in organizing WSNs. It has proven its efficiency on network performance by extending network lifetime, saving energy of sensor nodes, reducing delay and delivering more data packets. This research develops an algorithm called the Density Grid-base Clustering algorithm (DeGiCA) that enhances the clustering mining technique in WSNs by combining density and grid techniques. The deployment density variation technique can find arbitrary shaped clusters while the grid technique is used to avoid clustering quality problems by discarding the boundary points of grids. DeGiCA helps to face the limitations found in WSNs that carry data streams. By using a MATLAB-based simulator, DeGiCA is compared with other clustering algorithms in WSNs that manipulate data streams. Fuzzy Clustering Means algorithm (FCM) and K-means algorithm are two selected algorithms used to be compared with DeGiCA performance metrics results. The simulation results indicate that the performance of DeGiCA outperforms K-Means in terms of network lifetime by 15%, energy consumption by 13% and packet delivery ratio by 40%. DeGiCA also outperforms FCM in terms of network lifetime by 17%, energy consumption by 11% and packet delivery ratio by 70%. 
Supervisor : Dr. Manal Abdulaziz Ali Abdullah 
Thesis Type : Master Thesis 
Publishing Year : 1438 AH
2017 AD 
Added Date : Tuesday, September 26, 2017 

Researchers

Researcher Name (Arabic)Researcher Name (English)Researcher TypeDr GradeEmail
ياسمين سند الغامديAlghamdi, Yassmeen SanadResearcherMaster 

Files

File NameTypeDescription
 42816.pdf pdf 

Back To Researches Page