A wireless sensor node (WSN) system, capable of sensing animal motion and transmitting motion data wirelessly, is an effective and efficient way to monitor pigs' activity. However, the raw sensor data sampling and transmission consumes lots of power such that WSNs' battery have to be frequently charged or replaced. The proposed work solves this issue through WSN edge computing solution, in which a Random Forest Classifier (RFC) is trained and implemented into WSNs. The implementation of RFC on WSNs does not save power, but the RFC predicts animal behavior such that WSNs can adaptively adjust the data sampling frequency to reduce power consumption. In addition, WSNs can transmit less data by sending RFC predictions instead of raw sensor data to save power. The proposed RFC classifies common animal activities: eating, drinking, laying, standing, and walking with a F-1 score of 93%. The WSN power consumption is reduced by 25% with edge computing intelligence, compare to WSN power that samples and transmits raw sensor data periodically at 10 Hz. / Master of Science / A wireless sensor node (WSN) system that detects animal movement and wirelessly transmits this data is a valuable tool for monitoring pigs' activity. However, the process of sampling and transmitting raw sensor data consumes a significant amount of power, leading to frequent recharging or replacement of WSN batteries. To address this issue, our proposed solution integrates edge computing into WSNs, utilizing a Random Forest Classifier (RFC). The RFC is trained and deployed within the WSNs to predict animal behavior, allowing for adaptive adjustment of data sampling frequency to reduce power consumption. Additionally, by transmitting RFC predictions instead of raw sensor data, WSNs can conserve power by transmitting less data. Our RFC can accurately classify common animal activities, such as eating, drinking, laying, standing, and walking, achieving an F-1 score of 93%. With the integration of edge computing intelligence, WSN power consumption is reduced by 25% compared to traditional WSNs that periodically sample and transmit raw sensor data at 10 Hz.

This study is devoted to investigating and developing WSN based localisation approaches with high position accuracies indoors. The study initially summarises the design and implementation of localisation systems and WSN architecture together with the characteristics of LQI and RSSI values. A fingerprint localisation approach is utilised for indoor positioning applications. A k-nearest neighbourhood algorithm (k-NN) is deployed, using Euclidean distances between the fingerprint database and the object fingerprints, to estimate unknown object positions. Weighted LQI and RSSI values are calculated and the k-NN algorithm with different weights is utilised to improve the position detection accuracy. Different weight functions are investigated with the fingerprint localisation technique. A novel weight function which produced the maximum position accuracy is determined and employed in calculations. The study covered designing and developing the centroid localisation (CL) and weighted centroid localisation (WCL) approaches by using LQI values. A reference node localisation approach is proposed. A star topology of reference nodes are to be utilized and a 3-NN algorithm is employed to determine the nearest reference nodes to the object location. The closest reference nodes are employed to each nearest reference nodes and the object locations are calculated by using the differences between the closest and nearest reference nodes. A neighbourhood weighted localisation approach is proposed between the nearest reference nodes in star topology. Weights between nearest reference nodes are calculated by using Euclidean and physical distances. The physical distances between the object and the nearest reference nodes are calculated and the trigonometric techniques are employed to derive the object coordinates. An environmentally adaptive centroid localisation approach is proposed. Weighted standard deviation (STD) techniques are employed adaptively to estimate the unknown object positions. WSNs with minimum RSSI mean values are considered as reference nodes across the sensing area. The object localisation is carried out in two phases with respect to these reference nodes. Calculated object coordinates are later translated into the universal coordinate system to determine the actual object coordinates. Virtual fingerprint localisation technique is introduced to determine the object locations by using virtual fingerprint database. A physical fingerprint database is organised in the form of virtual database by using LQI distribution functions. Virtual database elements are generated among the physical database elements with linear and exponential distribution functions between the fingerprint points. Localisation procedures are repeated with virtual database and localisation accuracies are improved compared to the basic fingerprint approach. In order to reduce the computation time and effort, segmentation of the sensing area is introduced. Static and dynamic segmentation techniques are deployed. Segments are defined by RSS ranges and the unknown object is localised in one of these segments. Fingerprint techniques are applied only in the relevant segment to find the object location. Finally, graphical user interfaces (GUI) are utilised with application program interfaces (API), in all calculations to visualise unknown object locations indoors.