Примена на облачиња и пресметување со магла

Abstract

Mobile devices are an important part of people’s everyday life and represent the most important and relevant communication tool, providing them with different functionalities. The number of new applications designed for the needs of mobile users is growing every day. Due to the limitations of their own resources, mobile devices need to use additional resources. An alternative method is to use external processing units, which allow end users to use their resources to run mobile applications. The current methodologies for the development of mobile applications that use external resources for execution (cloud, cloudlet, fog computing, dew computing) are not yet perfected to the full extent used in the development of today’s mobile applications. There are a large number of shortcomings and uncertainties which are the main reason that the implementation of such methodologies is still at an early stage. There are various architectural solutions to speed up the processing of mobile devices that involve offloading to server computing, such as, cloud computing, cloudlet computing, fog computing or dew computing. The main goal of the doctoral dissertation is to model the behavior of a mobile application using an algorithm for optical character recognition of words and phrases, using a variety of architectural solutions for computing in cloud, cloudlet, fog or dew, and define an optimal solution. Through a series of theoretical and experimental research, as well as through modeling the performance of various architectural solutions, the answer to the question of the best architectural approach is the cloudlet. For practical implementation of the conclusions of this research, an appropriate business model was developed.