Deterministic Algorithms in Dynamic Networks: Formal Models and Metrics

The number of telecommunication networks deployed in a dynamic environment is quickly growing. This trend exists both in everyday life (e.g., smartphones, vehicles, and commercial satellites) and in a military context (e.g., dismounted soldiers or swarms of UAVs). Unfortunately, few theoretical tools to date have enabled the study of dynamic networks in a formal and rigorous way. As a result, it is hard and sometimes impossible to guarantee, mathematically, that a given algorithm will reach its objectives once deployed in real conditions. In this report, we identify a collection of recent theoretical tools whose purpose is to model, describe, and leverage dynamic networks in a formal way. These tools include a dynamic graph formalism, various computational models, and communication models for distributed networks. We extend many graph theoretical concepts towards a dynamic variant and show how these new variants impact the solution of classical distributed problems. The report also presents a hierarchy of dynamic networks based on dynamic graph properties, thereby offering a combinatorial alternative to the well-known mobility models typically used in simulations.