Consumo energético de animais em pastagem baseado em um modelo de mobilidade híbrido para redes de sensores sem fio

Abstract

Animal monitoring has emerged for livestock as an important factor for quality control and productivity. The idea is to monitor each animal to provide the producer with a daily summary of cattle activities. In most cases, monitoring is accomplished through a collar with a built-in GPS sensor. The collar stores the animal’s positioning in a given environment, allowing real-time monitoring by sending information to a base station. However, testing and deploying this equipment in the real world is costly, complex and takes time to get up and running. In this scenario, the use of wireless sensor network simulators becomes essential for the development of this type of technology. Wireless sensor networks are made up of small sensor nodes with limitations in processing, memory, and power. Each sensor node is powered by an internal battery, making network life an important parameter. In recent years, the routing problem in wireless sensor networks has been highlighted in the literature and many routing protocols have been proposed. Most of these protocols focus on transmission reliability and the energy efficiency of networks. In this work, we present a configurable model capable of calculating the energy consumption of pasture animals based on a hybrid mobility model used in the individual mobility models (Markovian Random Walk and Gauss-Markov) and groups (with Reference Point) in ad hoc networks. The mobility model was implemented in the Castalia 3.2 simulator and the algorithm that makes the energy consumption calculations was implemented in Python. The solution was evaluated as proof of concept in 3 area distributions, with 3 relief matrices, to observe the expected behavior. The results showed that the model is fully configurable and has a good functioning, similar to the actual mobility of the animals, being able to make the expected energy consumption calculations. Also, this work presents the performance evaluation of the routing protocols: Multipath Rings (Proactive), Bypass and AODV (Reactive) and LEACH (Hierarchical). As a result, the protocol that generated the lowest power consumption was LEACH and the highest power consumption was Multipath Rings. About the number of packets received by the sink node, the lowest loss was for Multipath Rings and the highest loss was for LEACH.