LAUSR

The design of wireless sensor networks (WSN) is still a hot topic either for industry or academia. Many research teams work on various issues as communication protocols, task scheduling, physical architecture, etc. Scheduling tasks in such environments is an important issue to consider. Indeed, tasks may have constraints (time execution, deadlines, execution cycle,...) which should be respected. In order to respect these conditions, adapted scheduling algorithms are proposed in this study. In this paper, we consider that a WSN is composed of clusters of sensors. And each cluster contains a set of sensors. We propose three different scheduling procedures: semi-dynamic algorithm (for each cluster we have a fixed planning but its sensors will have a dynamic slot repartition), dynamic algorithm (the sensor clusters will have a dynamic time slot repartition as all sensors) and high-priority dynamic (in addition to have dynamic scheduling, we consider tasks with different priorities). As a consequence of the use of scheduling methods, the communication between sensors is managed to be efficient and targets to avoid conflicting packets. To evaluate our proposals, we developed them over the OPNET simulator and run a large number of simulations. We have compared our scheduling approaches with a static scheduling method, we have analyzed some performance parameters. The simulation results show that our proposals reduce the latency and improve the packet delivery.