期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2016
卷号：84
期号：3
出版社：Journal of Theoretical and Applied
摘要：Widespread applications of location based services for indoor environments have created an opportunity for researchers to develop new techniques for accurate position estimation with less complexity. In this paper, we present a heuristic approach to localization by employing clustering for an indoor environment. Depending upon the number of light emitting diodes (LEDs) used as transmitter, level 1 clustering is achieved by simply comparing the signal strength for each combination of transmitter�s light intensities at the receiver. For level 2 clustering, a new technique of clustering is proposed, named portion clustering, is applied to further partition the area where the object of interest can be located. From the simulation results, it can be observed that the location estimation up to 16 centimeters can be achieved for an indoor environment with the dimensions of 3 � 3 � 3 using LEDs.
关键词：Localization; Light Emitting Diode; Machine Learning; Clustering; Received Signal Strength
