Abstract: These days, heterogeneous networks are one of the most used networks. Customers are using these networks to access various services. However, Handover/Handoff (HO) is one of the major problems which are being faced in every network. Different criteria are used by the researchers in order to execute effective handoff in order to avoid disconnection and interruption in the call. When the ongoing call reaches beyond its base station, the call is needed to be transferred to other BS to maintain its continuity. Thus, HO is initiated and the call is transferred. It is important to perform this process seamlessly. Thus, neural networks are also introduced in this field to take HO decisions. HO decision relies on data rate; monetary cost, RSSI and speed of MS. Neural networks are proficient for this process as it involves deep analysis of the data HO parameters. In this paper, a novel technique is involved to make neural networks efficacious to take HO decision. The proposed model is the amalgamation of artificial neural network and Fuzzy logic which in turn resulted in ANFIS model. This model utilized four parameters namely, RSSI, cost, data rate and velocity as imperative factors to determine the supremacy of the designed model. Further, MATLAB is used to perform the simulation analysis. Eventually, obtained results are compared with existing techniques which ensures the efficacy of proposed model in terms of velocity and HO probability.
Keywords: Handover. fuzzy logic, Artificial neural network, ANFIS
| DOI: 10.17148/IJARCCE.2019.81213