Multi-channel polling in multi-hop clusters of hybrid sensor networks

In this paper we propose a multi-channel polling algorithm in multi-hop clusters of hybrid sensor networks. The hybrid sensor network consists of two types of nodes: basic sensor nodes and cluster head nodes. Basic sensor nodes have limited communication capacity and mainly focus on sensing the environment, while cluster head nodes are equipped with more powerful transceivers but simpler sensing modules. The cluster head node organizes basic sensor nodes around it into a cluster, and collects sensing data from sensors and forwards data to the outsider observer. This type of network has better energy-efficiency since traffic within cluster is scheduled by cluster head and packet collisions can be avoided. In addition, idle listening time can be shortened by turning off the transceivers of sensor nodes after data has been forwarded to cluster head. We focus on finding energy efficient and collision-free polling schedules in the multi-hop cluster with multiple frequency channels. Due to its energy efficiency and scalability, the proposed algorithm is very suitable for applications such as large scale environment monitoring. We also implement the proposed algorithm on NS-2. Simulation results show that multi-channel polling algorithm shortens the active time of sensor nodes by a significant amount compared to single channel polling. In the case that the total frequency bandwidth allocated to the cluster is fixed, the optimal number of channels can be obtained.