No abstract

No abstract

No abstract

Effective and timely link layer trigger mechanisms can significantly influence the handover performance. In this paper, a predictive link trigger mechanism for seamless horizontal and vertical handovers in heterogeneous wireless networks is proposed. Firstly, the time required to perform a handover is estimated based on the neighboring network conditions. Secondly, the time to trigger a Link_Going_Down to initiate a handover is determined using a Least Mean Square (LMS) linear prediction in which the prediction interval (k h ) is dynamically determined based on the estimated handover time. Simulation results of the proposed predictive link triggering mechanism show that it provides a timely proactive handover. The packet loss rate observed in a Gaussian fading channel remains low during a handover. ⋅

No abstract

Next-generation wireless networks should be able to coordinate and integrate different communication systems. It has been a challenging problem to support a seamless handover in these diverse wireless network environments. Link level triggers can provide information about events which can help handover decision and layer 3 entities better streamline their handover related activities. In most conventional layer 2 triggering approaches, a pre-defined threshold for a specific perspective such as the received signal strength is used. This may cause too late or too early handover executions. In this paper we propose a new predictive handover framework that uses the neighbor network information to generate timely the link triggers so that the required handover procedures can appropriately finish before the current link goes down. First we estimate a required handover time for the given neighbor network conditions, then using a predictive link triggering mechanism the handover start time is dynamically determined to minimize handover costs. The handover costs are analyzed in terms of the total required handover time and the service disruption time. The numerical analysis and simulation results show that the proposed method significantly enhances the handover performance in heterogeneous wireless networks.

Effective and timely link‐layer trigger mechanisms can significantly influence the handover performance. The handover process will not perform the correct decision and execution unless adequate and timely link‐layer trigger information is delivered. In this paper, a predictive link trigger mechanism for seamless horizontal and vertical handovers in heterogeneous wireless networks is proposed. Unlike previous link trigger algorithms based on pre‐defined signal level thresholds, the link layer triggers in this study are adaptively and timely fired in accordance with the network conditions. Firstly, the time required to perform a handover is estimated based on the neighboring network conditions. Secondly, the time to trigger a Link_Going_Down to initiate a handover is determined using a least mean square linear prediction in which the prediction interval (kh) is dynamically determined based on the estimated handover time. An upper bound for the packet loss rate during a handover is derived for a Gaussian shadowing channel. A manner in which this approach can be applied to IEEE 802.21 is shown in media independent handover scenarios. Simulation results of the proposed predictive link triggering mechanism show that it provides a timely proactive handover. The packet loss rate observed in a Gaussian shadowing channel remains low during a handover. Copyright © 2008 John Wiley & Sons, Ltd.

This paper presents a methodology for assessing the suitability of various wireless technologies for meeting the communication requirements of Smart Grid applications. It describes an approach for translating application requirements to link traffic characteristics, determining the transmission range or coverage area of a wireless technology, and modeling the link layer to obtain performance measures such as message reliability, delay, and throughput. To illustrate the use of this approach, we analyze the performance of three representative application use cases over an IEEE 802.11 link.