A novel MAD-based network selection algorithm for heterogeneous networks

Next generation wireless network (NGWN) is expected to integrate different radio access technologies (RATs) and to support user services with different quality-of-service (QoS) requirements. In the overlapped area of various RATs, the problem of optimally selecting the access network has to be considered. However, the heterogeneity and incompatibility of different wireless access technologies, and the variety of user services pose challenges to guarantee the users´ QoS. In this paper, we stress the problem of network selection in heterogeneously integrated networks, and propose a novel multiple attribute decision (MAD)-based network selection algorithm. In addition to network current states and user requirements, user long-term handoff information between various access networks, which can be modeled based on hidden Markov model (HMM), is taken into account in designing the optimal network selection scheme. Numerical results show that the proposed algorithm can offers the better call block rate and achieves the balanced load status compared to the original MAD-based network selection algorithm.