Fuzzy decision fusion for single target classification in wireless sensor networks

Thesis Type: Postgraduate

Institution Of The Thesis: Orta Doğu Teknik Üniversitesi, Faculty of Engineering, Department of Computer Engineering, Turkey

Approval Date: 2009


Consultant: ADNAN YAZICI


Nowadays, low-cost and tiny sensors are started to be commonly used due to developing technology. Wireless sensor networks become the solution for a variety of applications such as military applications. For military applications, classification of a target in a battlefield plays an important role. Target classification can be done effectively by using wireless sensor networks. A wireless sensor node has the ability to sense the raw signal data in battlefield, extract the feature vectors from sensed signal and produce a local classification result using a classifier. Although only one sensor is enough to produce a classification result, decision fusion of the local classification results for the sensor nodes improves classification accuracy and loads lower computational burden on the sensor nodes. Decision fusion performance can also be improved by picking optimum sensor nodes for target classification. In this thesis, we propose fuzzy decision fusion methods for single target classification in wireless sensor networks. Our proposed fusion algorithms use fuzzy logic for selecting the appropriate sensor nodes to be used for classification. Our solutions provide better classification accuracy over some popular decision fusion algorithms. In addition to fusion algorithms, we present some techniques for feature vector size reduction on sensor nodes, and training set formation for classifiers.