A Method Based on a Combination of Ontological Data Analysis and Cognitive Modeling Tools for Organizing the Computational Process in Peer-to-Peer Networks

Despite the rapid development of edge and fog computing technologies, including significant improvements in the characteristics of communication channels and computing devices themselves, the problems associated with the organization of the computing process in distributed systems remain relevant. This is due to the high dynamism of the environment and the large number of computing devices that operate in it. One of the mentioned problems in the organization of the computing process is the workload relocation problem. It includes the selection of placement options and requires significantly more time resources for the case of coupled tasks in comparison with uncoupled tasks. The need to solve coupled problems is associated with the development of such areas as loT, the Internet of Robotic Things, collaborative robotics, augmented reality, etc. One of the ways to reduce the total solution time of a coupled problem is to reduce the search space of candidate nodes, which can potentially be assigned to subtasks. In this paper, we propose to use a method for reducing the search space based on a combination of ontology analysis and cognitive modeling tools. The ontological model and a cognitive map of the computing infrastructure are developed. Computational experiments are conducted. Application of ontological modeling along with the use of cognitive analysis tools has shown its effectiveness in solving the workload relocation problems in highly dynamic environments. This is confirmed by the results of the conducted experiments. (Abstract)