E-ANDSF-Based Base Station Selection Scheme by Using MLP in Untrusted Environments

In this paper, we design an enhanced access network discovery and selection function (ANDSF), named E-ANDSF, which provides previous mobile access gateway (PMAG) with information for base station selection, and propose an improved IP-based handover/offload scheme, calledIP-basedHandover/offload system assisted byE-ANDSF (IHEAD), which invokes multilayer perceptron, a machine learning technique, to select a base station for user equipment to handover or offload network traffic to wireless local area network (WLAN). The purpose is improving network service quality with the assistance of stream control transmission protocol and E-ANDSF. Experimental results show that the IHEAD outperforms those tested schemes.