Congestion Control in Computer Networks with a New Hybrid Intelligent Algorithm

Abstract

With invention of computer networks, transferring data from one computer to another became possible, but as the number of computers that transfer data to each other increased and common communication channel bandwidth among them in a network limited, has led to a phenomenon called congestion, so that some of data packets would be dropped and never arrive to destination. Different algorithms have been proposed for overcoming congestion. These are divided into two general groups: 1- flow based algorithms and 2- class based algorithms. In present study, using class based algorithm with optimization of its control by fuzzy logic and new Cuckoo algorithm, we increased the number of packets that reach to destination and reduced the number of dropped packets considerably during congestion. Simulation results indicate a great improvement of efficiency.