Node Scheduling for All-Directional Intrusion Detection in SDR-Based 3D WSNs

For intrusion detection in 3D wireless sensor networks, the monitoring quality and the energy efficiency are both of great significance. In this paper, we develop a novel globoid model to ensure the all-directional detection quality while saving the network energy effectively, which divides the sensing area into outermost shell and interior region. We first propose an outermost shell coverage algorithm to guarantee the recognition quality of intruding events. Then, a Markov prediction model is designed to predict the motion probability in the adjacent area based on the historical trajectories of intruders. According to the predicted results, different working frequencies will be allocated to the covered nodes by using software defined radio technology. Moreover, a trajectory correction strategy is proposed to relocate the missing intruders during the operation. The performance evaluations show the efficiency of our scheme in terms of the network lifetime, trajectory prediction accuracy, and success rate of correction strategy.