The overall performance of a WLAN 802.11e is determined, among other things, by the Enhanced Distributed Channel Access (EDCA) parameters, typically configured with default and static values. The same does not vary according to the dynamic requirements of Quality of Service (QoS) of the network. To achieve significant improvements in the performance of the network, and simultaneously comply with the specifications of QoS, a new algorithm of tuning is proposed to make the adjustment of these parameters dynamically, which we call Algorithm for the Differentiation of Traffic Multiple (MTDA). The algorithm has been evaluated in realistic scenarios with noise, reaching a initialization effective, rapid convergence, and the differentiation of effective multiple wireless traffic (voice, video and best effort with Pareto distribution). It adopted a model of wireless station implemented in Hierarchical Stochastic Activities Networks (HSANs), which runs on the simulator M??bius.