A distributed indexing mechanism for the Internet of Things

Abstract

Over the years, the growing innovations in information and communication technologies (ICT) in the computer and communications industry have allowed distributed systems researchers to design new systems and models to meet the needs of an increasingly diverse user community. In this project, we studied a set of indexing techniques in spatial metrics with a strict subset. We had neither the time nor the space to make an exhaustive overview. As a result, we were able to highlight only the metric space indices. Based on what appear to be the best recent approaches, we proposed a distributed approach to indexing data in a metric space, essentially inspired by the binary tree. In summary, our proposal can be considered as a paginated and parameterized version of the latter. Nevertheless, the problem encountered each time by these techniques is that of the "curse of the dimension". In a second step, we developed a distributed version to improve performance, which, in our opinion, should stagnate, or at least progress in this family of indices.