Data fusion architectures: A survey and comparison

Abstract

The data fusion architecture has an important role in the efficiency of the processed information and the significance of the made decision at the output level. For this reason, we choose to concentrate this paper on architectures of systems working on information fusion. This will allow scientists to have an idea about which architecture is more suitable to their application, and in which step one can work in order to modifY the information fusion process or to ameliorate a given framework.

Dissimilarity‐based test case prioritization through data fusion

“…In addition to DTCP, many other criteria can be used to support TCP, including coverage-based TCP (CTCP): 86 We look forward to investigating how to enhance CTCP through data fusion, and to fuse DTCP with CTCP. Apart from the Dasarathy architecture, 40 it will be interesting to apply other data fusion architectures 50 to TCP: This is something that we also look forward to doing in our future work.…”

“…Data fusion has been widely used in many applications, including: motor fault detection; 42 radar and image measurements; 43 rank review; 44 distributed tracking; 45 security enhancements; 46 IOT; 47 robust estimation; 48 and machine learning. 49 As discussed by Ayed et al, 50 many architectures have been proposed for data fusion. 40,41,51 In this article, we focus on the Dasarathy architecture.…”

An enhanced vehicle control model for assessing highly automated driving safety

“…Initially we modified the MISRA VCM by 'splitting' the Driver Control element between the human and machine actors, and using data-informationknowledge-wisdom pyramids [33] to represent the different information used for human and machine perception. However, having attempted to modify the model in this way, two things became apparent: Firstly, this approach simply added complexity to the diagram, which intuitively would never lead to an elegant solution.…”