ABSTRACT The qualities of telecommunication system play an important role in modern logistics management. Many public places have begun the construction of wireless networks around hot spots. Generally, their will be more than one access point (AP) is installed between floors. However, the number and location of access points are decided by engineers apply the Taguchi’s approach to resolve the problems with dynamic characteristics. The Taguchi’s dynamic approach accommodates a range of input signals that have an ideal relationship with the output response; a dynamic case has multiple values for the output optimization goals, where necessary for the two-step optimization. However, this method can only find the best one of the specified parameter level combinations. This study illustrates that integrating neural networks (NNs), genetic algorithm(GA), and desirability function (or exponential desirability function) can obtain a better result than the analysis by the Taguchi’s approach. A case study of the propagation system evaluation in wireless network demonstrates the effectiveness of the proposed approach.