Stochastic Geometry and Federated Learning in Computational Modeling of Communication Systems
Samuel Ibukun Olotu · 2024
A computational model refers to a mathematical description of a physical system. One important system that affects our daily lives is the communication system. It is an important system that enables communicating users to talk to each other and also share information and resources from any location around the world. It is reported that 86% of the world population uses smartphones to stay connected to the globe over different sophisticated communication networks. The description of the architecture, functionality and node movements of communication networks of these communication networks can assist network administrators and designers get better understandings on their functionalities and operations. One modeling approach that has been applied to communication systems and networks to describing their operations and the distribution of users properly is stochastic geometry. An advantage of the stochastic geometry technique is its ability to accurately model the random patterns of the nodes in the communication systems. A recent type of communication system is the federated learning (FL) that combines distributed computing and machine learning techniques to be trained locally and avoids the upload of data directly to the cloud server. The learning technique gives users an opportunity to benefit from the shared model that is trained from rich data without centrally storing it. The chapter presents a review of existing research works in stochastic geometry modeling of wireless communication systems and federated learning schemes. The review will assist researchers in communication network and system modeling based on stochastic geometry and federated learning.