Time to Change: On Distributed Computing in Dynamic Networks (Keynote)

Abstract

In highly dynamic networks, topological changes are not anomalies but rather integral part of their nature. Such networks are becoming quite ubiquitous. They include systems where the entities are mobile and communicate without infrastructure (e.g. vehicles, satellites, robots, or pedestrian smartphones): the topology changes as the entities move. They also include systems, such as peer-to-peer networks, where the changes are caused by entities entering and leaving the system, They even include systems where there is no physical mobility at all, such as social networks. A vast literature on these dynamic networks has been produced in many different fields, including distributed computing. The several efforts to survey the status of the research and attempts to clarify and classify models and assumptions, have so far brought more valuable bibliographic data than order and clarity. Goal of this note is to ask questions that might bring author and readers to start to clarify some important research aspects and put some order in a sometimes confusing field. The focus here is entirely on distributed computing, specifically on its deterministic aspects.