Minimising peer on-time for energy efficient peer-to-peer file distribution

Peer to peer (P2P) techniques are an effective means of content distribution. By exploiting the upload bandwidth of peers, P2P can reduce the time required for file distribution, yet requires the peers to stay active to participate in the distribution process (for downloading, uploading or both). So far little attention has been paid to its energy consumption. This paper studies lower bounds on the possible total time that peers need to be active in order to distribute a file from a single server to a given set of peers. We show that a P2P system optimized for energy efficiency can consume half the energy of one optimized purely to minimize download time, while imposing minimal additional delay. To achieve this, peers should finish in increasing order of upload capacity, which is the reverse of what is optimal for delay. Moreover, peers should turn off as soon as they finish downloading. For networks of up to three peers, optimal strategies are derived and proven to be optimal, and larger systems are studied by simulation.