An application of genetic fuzzy system to improve the random exponential marking algorithm in active queue management

Active Queue Management algorithms (AQM) are setup in routers on telecommunication networks in order to provide a mechanism to mark or discard packets in an appropriate way so as not to cause congestion on network and maintain the best quality of services with different traffic flows. Previous methods of AQM are designed with fixed parameters so It should not really online adapt to the network dynamic. To overcome this drawback, this paper proposed a new algorithm to improve Random Exponential Marking (REM) by combining fuzzy system and genetic algorithm. This new method allows automatic adjustment of parameters with the network dynamic. Results are verified through software simulation on NS2.