Predictive duty cycle adaptation for wireless camera networks

Wireless sensor networks (WSN) typically employ dynamic duty cycle schemes to efficiently handle different patterns of communication traffic in the network. However, existing duty cycling approaches are not suitable for event-driven WSN, in particular, camera-based networks designed to track humans and objects. A characteristic feature of such networks is the spatially-correlated bursty traffic that occurs in the vicinity of potentially highly mobile objects. In this paper, we propose a concept of indirect sensing in the MAC layer of a wireless camera network and an active duty cycle adaptation scheme based on Kalman filter that continuously predicts and updates the location of the object that triggers bursty communication traffic in the network. This prediction allows the camera nodes to alter their communication protocol parameters prior to the actual increase in the communication traffic. Our simulations demonstrate that our active adaptation strategy outperforms TMAC not only in terms of energy efficiency and communication latency, but also in terms of TIBPEA, a QoS metric for event-driven WSN.