A Robust Regression Model for Handover Prediction In Wireless Networks

Volume 17 , Issue 2 , June 2015 , Pages 93-104

Authors

¹ Mathematics Department, School of Science Education, University of Sulaimani, Sulaimani, Iraq.

² Faculty of Computer Science and Information Systems, Universiti Teknologi Malaysia, Malaysia.

³ Faculty of Engineering, School of Computing, Koya University, Koya, Iraq.

The recent applications of wireless networks demand high Quality of Service (QoS) and

mobility management between access networks. For this reason, emergent Advances in

IEEE 802.11 standards have been witnessed to achieve seamless communication

between mobile nodes. However, since existing handover mechanisms have

shortcomings in terms of mobility handover cost and latency, more robust handover

decision approach should be considered. To this end, in this research, we propose a

decision approach to accurately trigger handover process between current and next

access point. The proposed approach incorporates robust regression model is to predict

the received signal strength (RSS) of the wireless channel between the mobile node

(MN) and the access point. Thus, the MN makes an accurate handover decision as well

as layer 3 handover process occurs in advance to the one in layer 2. Our theoretical

analysis and simulation results show that the proposed handover decision approach is far

superior to the existing handover decision schemes in terms of handover latency and

cost.

Statistics