Orchestration and analysis of decentralized workflows within heterogeneous networking infrastructures

This paper applies the Network Edge Workflow Tool (Newt) to decentralized causal workflows in dynamic heterogeneous wireless networks. We build on previous work applying a Newt prototype to an emergency response scenario and extend in two ways: we apply Newt to a far more advanced use case which models distributed complex causal processing and interactions; and second, we apply time-windowed conversational graphs to analyze the distributed workflow. In addition to the prior disaster recovery use case, we apply Newt to the orchestration of William Shakespeare's Hamlet by distributing the actors across a wireless network and have actors converse their lines as decentralized messages of communication between one actor and another as the play progresses. We construct and analyze conversational graph models for the Hamlet workflow by comparing an a priori baseline model and a post facto estimate resulting from measurements of network conversational adjacencies during the execution of the workflow.