Correlation-aware cross-layer design for network management of wireless sensor networks

The observations of the nodes of a wireless sensor network are spatiotemporally correlated. Sensor nodes can exploit the correlation for enhancing network efficiency. However, an energy-efficient collaboration is required for better network management. For saving energy, sensor nodes schedule between Active and Sleep states. Nodes extract information from medium access control layer, and use that information along with the correlation of observations as a means of energy-efficient collaboration and proper scheduling of their Active and Sleep states. Furthermore, sensor nodes use non-deterministic reinforcement learning-based approach for reducing energy consumption and end-to-end delay by regulating the duration of their Sleep states. Extensive simulations have shown that the proposed cross-layer approach outperforms existing benchmark schemes in terms of end-to-end delay, data accuracy and energy efficiency.