Computational methods for analyzing dynamic spanning tree games on time-varying networks

Abstract

Dynamic spaning tree games have gained popularity as a robust approach to investigating resources allocation and interactions of strategies over time-varying networks. The commoblative networks are dynamic, and it is focused on devising novel computational methods for their analysis, including algorithms for the correction of Nash equilibria as well as establishing the structural properties of optimal strategies. Thus we design a set of scalable algorithms capable of addressing large instances of dynamic spanning tree problems by borrowing advanced techniques from graph theory, optimization as well as game theory. Our methods are novel in the sense that they address the network dynamic in question, owing to the use of sophisticated data structures, avoid existing methods. In this way they enable greater computational speed up. Such theory is deeply buried under real world settings. The suggested computation framework allows to expand the use and applications of dynamic spanning tree games to transportation and logistics, social networks and infrastructure systems. We demonstrate the effectiveness of our algorithms through extensive numerical experiments on both synthetic and real-world datasets, showcasing their ability to uncover key insights into the strategic behavior of players in dynamic network settings.