Design and Analysis of Wireless Sensor Networks for Animal Tracking in Large Monitoring Polar Regions Using Phase-Type Distributions and Single Sensor Model

Data gathering through wireless sensor networks (WSNs) has been used for the monitoring of endangered species. However, when it comes to animals that live in difficult access environments with large areas, where human access is extremely difficult, such as in polar regions, remote monitoring by traditional methods becomes complicated and even inefficient. This paper proposes the characterization of the animal random trajectories by means of the random walk model in order to select the appropriate detection range and number of nodes to guarantee a target detection probability. The animals are detected by static gatherer sensor nodes placed on land either by mobile sensor nodes attached to the animal or by land sensor nodes that detect them through movement, sound, or temperature among other methods. Due to their natural movement, the animal may be outside or inside the sensor nodes coverage radius. In order to reduce energy consumption, it is proposed that nodes be active and inactive, effectively increasing the system lifetime. As such, an inherent compromise between energy consumption and reporting efficiency is present in the design of the network. Building on this, careful network design is required in order to calculate the probability of successful detection and system lifetime. To this end, a mathematical model based on a Markov chain is proposed and developed. The model suitability is assessed via numerical simulations. The obtained results allow concluding that their trajectories can be modeled using specific phase type distributions. Finally, instead of considering that in polar regions, it is not feasible to have a conventional WSNs, where many nodes are placed together, single nodes are placed in strategic locations isolated among them, and communication between nodes is not possible. As such, it proposes a novel and simplified model, where a single sensor is used to analyze the performance of the complete multi-sensor network.