Abstract
•A dynamic energy-efficient network model that can reduce the energy dissipation of sensor nodes while improving on WSN lifetime is proposed. The Harmony Search Optimization Algorithm is the core that the model is built on it.•The proposed model employs two approaches named a mobile base station and a cluster-based network technique to reduce the communicating distances between sensor nodes and their base station.•The cluster-based network technique seeks to build a dynamic network infrastructure where the optimal number of clusters and their members is predicted.•The mobile base station algorithm seeks to find the near optimal location (i.e. center) between the cluster heads that represent each cluster and the base station in a manner that can reduce the communicating distances and to be closer to all of them.•The proposed dynamic network model runs into rounds including the sensing and sending stages until the last node dies.
In wireless sensor networks (WSNs), the location of the base station (BS) relative to sensor nodes is an important consideration in conserving network lifetime. High energy consumption mainly occurs during data communication between sensor nodes and the BS, in both single and multi-hop infrastructures. A WSN design with dynamic BS relocation is therefore desirable as this may prolong the network operational lifetime. However, positioning the BS next to each sensor node may cause data gathering latency. The reorganization of sensor nodes into clusters and the choice of a delegate node from each cluster, known as a cluster head (CH), as a ‘communicator’ between each cluster and the moving BS appears to avoid this latency problem and enhance the energy utilization in WSNs. In this paper, we propose an energy-efficient network model that dynamically relocates a mobile BS within a cluster-based network infrastructure using a harmony search algorithm. First, this model allocates sensor nodes into an optimal number of clusters in which each sensor node belongs to the most appropriate cluster. Following this, the optimal CHs are chosen from the other clusters’ sensors in order to evenly distribute the role of the CHs among the sensors. This infrastructure changes dynamically based on the number of alive nodes, so that load balancing is achieved among sensor nodes. Subsequently, the optimal location of the moving BS is determined between the CHs and the BS, in order to reduce the distances for communication. Finally, sensing and data transmission takes place from each sensor node to their respective CH, and CHs in turn aggregate and send this sensed data to the BS. Simulation results show very high levels of improvements in network lifetime, data delivery and energy consumption compared to static and random mobile BS network models.