Optimization of selection strategies for P2P streaming network based on daily users’ behavior and users’ distribution over time zones

In this paper, an optimization problem of selection strategies for peer-to-peer (P2P) live streaming network is discussed. To solve the problem, the simulation model of P2P live streaming network is developed. The model considers daily peers behavior, their distribution over time zones, collisions, time lags between the server and a peer, lags between peers, and three types of selection strategies: neighbor selection strategy, peer selection strategy, and chunk selection strategy. Daily peers’ behavior is defined as the distribution of the number of online users by the time of day. Initial data for the peers distribution over time zones and their daily behavior are taken from the known Internet sources. The aim of the research is to find an appropriate solution of the proposed optimization problem and to show how the choice of a certain set of selection strategies affects the key characteristics of P2P streaming networks. The results of the conducted numerical analysis show the increase of the network performance up to 16,25 %. © Springer International Publishing AG 2016