作者: Kamran Ali , Ijaz Haider Naqvi
DOI: 10.1109/WCNC.2016.7564802
关键词: Identification (information) 、 Tracking (particle physics) 、 Event (probability theory) 、 Computer science 、 Artificial intelligence 、 Scheme (programming language) 、 Cluster analysis 、 Pattern recognition
摘要: This paper introduces EveTrack, an online and distributed method for localization tracking of global composite events. Based on hyper-ellipsoid clustering model, we compute the percentage contributions individual attributes in multi-attribute correlated In addition, EveTrack utilizes spatio-temporal correlations between multiple events during its event identification phase. Finally, estimates location using iterative Linear Least Square (LLS) approach based intensities estimated at different nodes. The results our algorithm show 4–10 fold improvement accuracy with significantly less computational complexity when compared to previously proposed algorithms.