Aysegul Alaybeyoglu

A distributed fuzzy logic-based root selection algorithm for wireless sensor networks

In this study, a distributed fuzzy logic (DFL) method with five input parameters namely, energy, centrality, distance to base station, number of hops and node density is proposed for efficient root election system. In order to prevent high energy consumption during message transmission, we made each node run fuzzy logic engine distributedly. We decrease the number of message transmissions from member nodes to root node by running DFL on intermediate nodes and by eliminating the messages of nodes that have less probability to be selected as a new root. The proposed system also includes fault tolerance, load balance, timeliness and the scalability mechanisms. To prove the efficiency of our algorithm, we compared it with the algorithms namely; Low Energy Adaptive Clustering Hierarchy (LEACH), Adaptive Clustering Algorithm via Waiting Timer (ACAWT), Cluster Head Election mechanism using Fuzzy logic (CHEF) and Guptas Algorithm.

Transmission of image data using fuzzy logic based clustering infrastructure in mobile multimedia sensor networks

In this study, by considering nodes mobility, image data is transmitted distributedly to the base station via fuzzy logic based clustering infrastructure. This study is composed of two phases. In the first phase, fuzzy logic based clustering infrastructure is constructed with six input parameters namely, nodes speed, energy, centrality, distance to base station, number of hops and node density. In the second phase, huge amount of image data is distributed among the nodes to compress and send it to the base station. The proposed system is compared with the centralized approach and ICGA for energy consumption, network lifetime and image quality parameters on different mobility models of the sensor nodes.

Performance evaluation of learning styles based on fuzzy logic inference system

Determining best convenient learning style in accordance with the individuals capabilities and personalities is very important for learning rapidly, easily, and in high quality. When it is thought that each individual has different personality and ability, it can be recognized that each individuals best convenient learning style will be different. Because of the importance of lifelong learning, many methods and approaches have been developed to determine learning styles of the individuals. In this study, a rule based fuzzy logic inference system is developed to determine best convenient learning styles of the engineering faculty stuffs and the students. During studies, two different learning style models namely Honey&Mumford and McCarthy are used in implementations. This study is carried out with a total number of 60 and 26 engineering faculty students and stuffs, respectively. The personal information form and Learning Style Preference Survey of Honey&Mumford and McCarthy are used to collect the data which are analyzed using the techniques of frequency, percentage, mean, standard deviation, and t‐test. While Honey&Mumford learning style classifies engineering faculty students and stuffs as Activist, Reflector, Theorist, and Pragmatist; McCarthy learning style classifies them as Innovative, Analytic, Common Sense, and Dynamic. Gender, age, and department are selected as the metrics for evaluation of the system performance.

A mobile application for safe driving

In this study, a safe driving system is developed to reduce traffic accidents that cause social losses considerably. The ability of the driver is reduced to perceive environmental incidents during high speed driving. Speeding warnings over mobile applications which creating with comparing the speed of the vehicle with the speed limits of the roads and calculating the maximum speed before sharp bends will increase the concentration level of the drivers against accidental probabilities.

A design of fuzzy logic based android application for safe driving

The aim of this study is to design safe driving system which is constantly being worked on every year in the world to reduce traffic accidents that cause economic and social loss in very large sizes. The developed safe driving system works on mobile phones that are not part of other secure driving systems. By this way, regardless of the technological equipment of the vehicle, old or new all means of vehicles via mobile phones have become safe driving possibilities. The second goal of this study is to integrate knowledge of mechanical engineering with knowledge of software engineering so that these two professions will work together in future projects.

Designing of an expert system based on firefly algorithm for diagnosis of Heart Disease

Expert systems based on artificial intelligence techniques help to experts by using in various areas. Especially recently they contribute to a fast and accurate diagnosis in regions that lack doctors by using in medical areas and they can provide confort for both doctors and patients, saving on time and labour-saving by reducing the time for talks between the patient and physician. Heart disease is a difficult disease to notice because its symptoms are ignored by patients and it is generally understood in the first recognition of a first heart attack. In some cases it can be late for early treatment because of the first heart attack is severe. In this study, an expert system based on Firefly Algorithm is designed for diagnosis of heart disease and the results were evaluated using the evaluation criteria accuracy, positive predictive value, negative predictive value, sensitivity and specificity by making applications with data set. Accuracy value is obtained ninety percent and it is shown that the system is useful and beneficial in this area.

A NEW DISTRIBUTED LOCATION-BASED ROUTING ALGORITHM FOR WIRELESS SENSOR NETWORKS

Searching and routing procedures are important in order to ensure communication in wireless sensor networks (WSN). Although naive flooding-based searching is simple to implement, it costs a high number of message transmissions and results in high energy consumption. In this study, we propose a new distributed location-based routing algorithm for WSN. Our goal was to decrease the number of message transmissions and to increase coverage by constructing relay zones. Directed broadcast, relay zone, and broadcast suppression constitute the backbone of our algorithm. We compared our algorithm with a flooding-based approach, and saw that our algorithm performs much better for several parameters.