Dynamic adjustment based traffic allocation algorithm over multipath network

The major issue related to the realization of multipath network is the design of suitable dynamic adjustment algorithm of traffic allocation. The design challenge is to guarantee the various qualities of service requirements for all traffic, special the loss rate while minimizing the cost function. In this paper a novel traffic allocation algorithm based on time series prediction and PSO (particle swarm optimization) is proposed. The paper describes how the algorithm dynamically predicts loss levels over all available network paths and adjusts the traffic allocation proportion of packets for next prediction period by the adaptive PSO. Simulation results show the superiority of our algorithm over equality splitting algorithm.