Optimization in task--completion networks

TL;DR

This paper analyzes how individual optimization strategies in task-completion networks lead to complex collective behaviors like multiple equilibria and hysteresis, highlighting the importance of local decision-making on overall network dynamics.

Contribution

It introduces a model linking individual buffer management strategies to emergent network phenomena such as coexistence of equilibria and hysteresis.

Findings

Network exhibits multiple stable equilibria

Hysteresis effects are observed in collective behavior

Individual strategies significantly influence system dynamics

Abstract

We discuss the collective behavior of a network of individuals that receive, process and forward to each other tasks. Given costs they store those tasks in buffers, choosing optimally the frequency at which to check and process the buffer. The individual optimizing strategy of each node determines the aggregate behavior of the network. We find that, under general assumptions, the whole system exhibits coexistence of equilibria and hysteresis.