An efficient solution for max-min fair rate allocation in P2P simulation

With the rapid development of peer-to-peer networks, simulation is considered as a useful tool to demonstrate initial ideas before building prototypes with real network transmissions. In flow-based simulation models of peer-to-peer networks, the algorithm for rate allocation which computes the per-flow rate for several competing flows at each peer in a given network topology is a fundamental component. Approaching to large scale peer-to-peer systems, we develop a new approach to compute max-min fair rate allocation with an objective of reducing computational cost compared to previous work. Our results show that the proposed solution outperforms the benchmark: the CPU consumption time is saved up to 78-82%.