RLS for Link Trigger in Handover across Heterogeneous Wireless Networks

The wireless communication system is evolving towards a ubiquitous wireless network in which multiple-radio devices can roam anytime and anywhere across different access networks. Minimal latency and low packet loss ratio during handover (HO) are the indispensable requirements to reduce service disruption and support real-time services. However, minimizing service disruption is a challenging issue to multiple access networks since HO latency is much higher than that of homogeneous networks. Link layer trigger or layer 2 trigger has been proposed to address this issue. Nevertheless, the precise timing and definitive criteria for initiating L2 trigger which can greatly affect the handover performance is not specified. This paper considers the issue of how to timely initiate handover decision from the link layer by using adaptive filtering algorithm to predict forthcoming received signal. Particularly, we propose Recursive Least Squares (RLS) as the predictive method for layer 2 triggering and compare its performance with various conventional filtering techniques. A simulation in which the proposed predictive algorithm is implemented to a handover scenario from Wi-Fi to WIMAX is also presented. Our numerical analysis and simulation results show that the proposed algorithm can achieve better performance than that of the other algorithms and can be used to enhance the handover performance in heterogeneous wireless networks.