A Decentralized Algorithm for Coordinating Independent Peers: An Initial Examination

Peer-to-peer (P2P) applications are composed of a distributed collection of peers that cooperate in order to perform some common task. Though P2P applications have attracted the attention of researchers, there has been little exploration of the deep issues; rather initial attention has been on widely known but technically shallow applications such as found in Napster and Gnutella. One particularly rich domain for examining the utility of P2P applications is distributed, decentralized crisis response. This paper studies the applicability of a peerto-peer approach in such an emergency response situation. We have developed a prototype peer-to-peer infrastructure that models a group of firefighters who communicate with each other while fighting fires. Each firefighter (peer) runs a novel distributed "k-server" algorithm that makes local autonomous decisions based on the information received from other firefighters. While this study was limited in that it used a simulation to study the algorithm, the emergent behavior observed suggests that further, more detailed investigations are warranted.