Susanto, Ferry ORCID: 0000-0002-7484-8253 (2017) High Resolution Environmental Modelling Application Using a Swarm of Sensor Nodes. PhD thesis, Victoria University.

Abstract

The advancement of sensing technology has successfully reduced the physical size of a sensor node and stimulated the application of swarm sensing (millimetre scale sensors). Such a system has been envisioned to provide novel applications. For example, CSIRO has commenced the application of swarm sensing technology to track insect that aims to understand how the environmental situation influences bee behaviour. While the micro-sensor is still under development, it is crucial to have a baseline data set for initial data analysis purposes so that reasoning with the rich data is possible once the hardware has been developed and deployed. This work will propose a field simulation to address this issue. A hybrid environmental sensor network will be deployed, for the purpose of making highly dense observations, that consists of: (i) fixed sensor nodes, acting as weather stations, that collect data in a regularly-spaced time interval; and (ii) mobile nodes that sense the environmental parameters while insects move within the region with extremely high frequency – i.e. seconds. The proposed spatio-temporal interpolation algorithm in this dissertation (i.e. for environmental modelling) has achieved a computational efficiency factor from highly-dense sample data with an acceptable statistical error. The method also reconstructs the environmental situation in reality – e.g., produce a smooth surface in space and over the time. Combination of a successfully developed field simulation and the interpolation algorithm has opened up a wide range of applications. For instance, researchers could infer bee behavioural dynamics based on the environmental changes that they are experiencing. Such activities could assist entomologists to deepen their understanding of bee behaviour, with a view to advance our knowledge of the decline in bee populations worldwide.

Search Google Scholar

Repository staff login