ABSTRACT
This chapter provides a comprehensive review of contemporary data fusion methodologies, and overviews the most recent developments and emerging trends in this field. It discusses the recent advances in mobile and ubiquitous sensing, cloud storage and computing, and prevalence of social networks, the new and emerging directions in data fusion research, such as social data fusion, cloud-enabled and big data fusion, and fusion of streaming data. Multisensor data fusion aims to overcome the limitations of individual sensors and produce accurate, robust, and reliable estimates of the world state based on multisensory information. The most common and popular conceptualization of fusion systems is the Joint Directors of Laboratories (JDL) model. Performance evaluation aims at studying the behavior of a data fusion system operated by various algorithms and comparing their pros and cons based on a set of measures or metrics. The outcome is typically a mapping of different algorithms into different real values or partial orders for ranking.
