Estimation of Link Quality and Residual Time in Vehicular Ad Hoc Networks

High node mobility and transient connectivity in vehicular ad hoc networks have introduced numerous challenges in the design of efficient communication protocols for these networks. Toward the goal of efficient design of such networks, the focus of this work is to develop methods that will allow the estimation of a wireless link's future quality and the remaining (residual) time for which this will remain efficiently active and useful for data transmission. Such information can serve as key input to the routing algorithms and data dissemination for VANETs. Specifically, we propose a cross-layer approach to the link estimations by introducing a method that utilizes the received signal strength of data packets. Small scale fading induced by the relative movement of the nodes, limited knowledge of the signal, caused by the fact that the received energy metric will be numerically available only when packets are sent, and the small number of available samples of the metric make the problem challenging. To overcome these, a signal processing technique, empirical mode decomposition is used along with robust regression for the prediction of link quality and residual time. The validity of the proposed methods are tested and validated via extensive simulations.