Predicting the link stability based on link connectivity changes in mobile ad hoc networks

Link stability issues are significant for the route selection process in mobile ad hoc networks (MANETs). A link can be instable due to the random characteristics of the wireless channel, and/or the mobility of nodes. This paper considers instability due to mobility, and proposes a scheme to predict the link stability in mobile scenarios of MANETs. The prediction scheme is based on link connectivity changes and can be performed on the network layer, without the need of low layer data. We assume that the link connectivity follows the continuous-time Markov chain model, and consider the case of non-stationary movements. We propose a method to estimate the transition rates of the link connectivity model. The stability of the link is evaluated based on the transition rates. The prediction scheme is derived analytically and requires no prior information about parameters of the mobility model. Simulation results show that the proposed scheme can provide correct prediction in both stationary and non-stationary scenarios.