Summary |
The behavior of distributed resource sharing and scheduling algorithms in communication networks can commonly be described in terms of a hard-core interaction model on a conflict graph. The underlying conflict graph is typically assumed to be a given static structure, but in actual dynamic regimes both the nodes and edges tend to be subject to significant random variation over time. For example, edges in the conflict graph may come and go over time as interference dynamically varies due to fading and user mobility in wireless communication networks. Moreover, nodes in the graph may emerge when users request service and disappear again once the service request has been completed, which could require a measure-valued description of the conflict graph. In some scenarios the random variations in the graph may be governed by an exogenous stochastic process, while in other instances the graph dynamics may be strongly affected by the state evolution of the hard-core interaction model, giving rise to a complex two-way interplay. The goal of this project is to examine the fundamental characteristics of hard-core interaction models on time-varying conflict graphs |