Proposed Architecture to manage critical states predictions in IoT applications
Catalin Constantin Cerbulescu, Claudia Monica Cerbulescu · 2017
In IoT applications, data from various sensors are processed and delivered in different structures and formats (JSON, XML, CSV etc) so a noSQL database is a good approach for a persistence layer. Analyzing sensors data evolutions could predict critical states in a system (infarction, stoke, hyperglycemic coma). The data analysis in noSQL databases is time-consuming and not effective in real time decision. This paper present a new layer as a persistence layer wrapper which will separate sensor data into two DBMS focused on different performances and purposes. All data will be stored in a noSQL database. Data used for critical states predictions will be selected and stored in a fast, SQL DBMS and adequate algorithms will be used to manage them.