Event-driven hybrid MAC protocol for a two-tier cognitive wireless sensor network: design and implementation

Abstract

In this paper, an integrated hardware and software system for a cognitive-radio wireless sensor network CR-WSN is designed and implemented for event detection and monitoring applications. The proposed CR-WSN is a pre-deployed two-level cluster-based network. The lower level consists of several clusters of sensor nodes, each with its own cognitive-enabled cluster-head CR-CH. The upper-level consists of a network of CR-CHs. The data collected by each sensor node is sent to the associated CR-CH. The CR-CH opportunistically transmits the sensed information over one of the available primary radio PR channels to a monitoring node sink to take the appropriate action. To connect between the various elements and nodes in the system, we develop efficient communication protocols for the CR-WSN. Specifically, we present a novel event-driven hybrid MAC design called CRS-MAC for the CR-WSN such that reliable data delivery is achieved. CRS-MAC divides the MAC-layer functionality into two types: CSMA-based single-channel intra-cluster communications and cognitive multi-channel CSMA-based inter-cluster communications. The developed protocol is implemented and tested to provide a flexible software platform that enables efficient and reliable operation for the CR-WSN. Experimental and simulation results demonstrate a robust and scalable performance of the proposed CR-WSN system.