Next generation wireless network (NGWN) is expected to integrate different access technologies, including cellular network, WLAN and mobile Ad Hoc network, etc., and to support users' seamless communication among different access networks. However, the heterogeneity and the diversity of access networks propose challenges on user handoff performance enhancement and system performance guarantee. In this paper, we address the problem of user QoS enhancement and system performance improvement in heterogeneous integrated system of WLAN and UMTS. Neural network (NN) based access network modeling method is proposed and an adaptive parameter adjustment algorithm based on NN model is presented. Applying the algorithm in handoff source and destination networks, user parameters can be determined optimally. The simulation results show that the proposed scheme enables the handoff user to adapt the destination network environment quickly and the variation of the throughput can be avoided efficiently. Furthermore, the network performance of the handoff source network can also be guaranteed.