Evolving Virtual Fireflies

In this paper, we present a study in the evolution of cooperative behavior, specifically synchronization, through digital evolution and multilevel selection. In digital evolution, a population of self-replicating computer programs exists in a user-defined computational environment and is subject to instructionlevel mutations and natural selection. Multilevel selection links the survival of the individual to the survival of its group, thus encouraging cooperation. Previous approaches to designing synchronization algorithms have taken inspiration from nature: In the well-known firefly model, the only form of communication between agents is in the form of “flash” messages among neighbors. Here we demonstrate that populations of digital organisms, provided with a similar mechanism and minimal information about their environment, are capable of evolving algorithms for synchronization, and that the evolved behaviors are robust to message loss. Moreover, analysis of the dominant genome reveals that the evolved solution utilizes an adaptive frequency strategy strikingly similar to that observed in fireflies.