Evolutional-based network traffic flow behavior acquisition and fuzzy control system design

Abstract

A theory of IEEE 802.11e of the wireless channel service concept is applied in this paper. The integration of particle swarm optimization (PSO), fuzzy inference theory and Markov Chains model are proposed to design the evolutionary-based network traffic flow behavior acquisition and fuzzy control system. Markov Chains model simulates the flow saturation state of internal channel competitions to detect the channel busy probability with the regulation of backoff cycles. Wireless network flow model is developed to apply time serious service package based on the distribution of collided flow probability. The acquisition features of network traffic flow describe the package sending behavior to construct the fuzzy control architecture. Suitable fuzzy rules are selected by the PSO learning scheme to improve the flow congestion condition and shorten the service waiting time. Computer simulation results on two network station problems are derived to demonstrate the efficiency of the proposed methods.