|
|
System for Processing Data from Smart Devices
Bureš, Michal ; Burget, Radek (referee) ; Hynek, Jiří (advisor)
This thesis deals with the design, implementation, and testing of a new information system for processing significant volumes of data from smart devices using user criteria – terminologically Key Performance Indicators. The thesis describes the domain of the Internet of Things and Smart Cities and discusses Key Performance Indicators and their applicability for monitoring smart devices and transforming data from these devices into practically applicable knowledge. Another significant topic of the thesis is information systems, their architecture, and performance – here the thesis describes various concepts and aspects through which it is possible to define and evaluate the performance of software, especially information systems and web services. Subsequently, the thesis deals with the system itself, first from the perspective of requirement specification, determined by analysis of potential users and their needs, and also analysis of existing solutions and their shortcomings. Significant requirements for the new system include stream processing of data from smart devices in real-time, high system capacity, system stability, and expandability. Later, the design of the system is described with an emphasis on the data model of the system, the process of evaluating Key Performance Indicators, and the architecture of the system, followed by implementation with an emphasis on the implementation of the client side of the system, the implementation of individual services forming the backend of the system, communication between these services using RabbitMQ technology, etc. The final part of the thesis, dedicated to testing the system, focuses on system performance, i.e., the throughput of individual services in processing messages from smart devices, optimization of throughput, benefits of horizontal scaling, and system-wide latency.
|
guest ::
