A novel vertical handover algorithm based on Adap-tive Neuro-Fuzzy Inference System (ANFIS)

摘要

Nowadays, there is an increased demand on an Internet connection anywhere at any time. Therefore, one has to exploit all available heterogeneous wireless networks where the target is achieving the Always Best Connected (ABC) among the different networks like UMTS, WiMAX, and WLAN. So, vertical handover techniques are used to ensure the best connectivity anywhere at any time. In this paper, novel ANFIS-based vertical handover is presented and compared with TOPSIS algorithm and other algorithms as a representative of Multi-criteria decision making (MCDM) algorithm's family. The simulation results show that the proposed handover technique provided better performance in terms of minimizing the time delay and improving the quality of service (QOS). This is because ANFIS requires iterations only in training phase otherwise, it has a much faster response. Our simulations considered the effect of many practical parameters on handover, such as subscriber speed, jitter, initial delay, bandwidth and received signal strength (RSS).According to these parameters, output values produced, which is utilized to choose the best candidate access network.