A fast and efficient end-to-end QoS measurement scheme for real seamless handover

We propose a fast and efficient end-to-end QoS measurement scheme. It collects distributed QoS information which is autonomously measured at each network segment in the backbone networks, and estimates the end-to-end QoS parameters from the collected data. It does not require sending the probe packets and statistical analysis for some duration, the proposed method provides the end-to-end QoS information much faster than the conventional methods. The measurement time of our method depends on round-trip time of the end-to-end path which is typically on the order of 10(1) or 10(2) milliseconds. In the user-centric heterogeneous wireless networking, each mobile terminal has to autonomously select an optimal PoAs (Point of Attachment; access point and base station) from available network candidates according to such an end-to-end QoS information to continue the services (VoIP, video conference, web browse, etc.) since their QoS requirements are different. Therefore, the fast end-to-end QoS estimation is crucial for seamless handover among various networks with a right decision. In this paper, we have shown that our proposed method can estimate end-to-end delay and jitter with only a few milliseconds error by simulation.