A dynamic data gathering in video based wireless sensor networks

Abstract

The theme of this thesis regards the domain of video based wireless sensors and proposes two algorithms fost real time field of view recovery with minimum of resources used. This implies maintaining the coverage at a high level while preserving and prolonging the lifetime of the network. The algorithms are performant and make use of novative concepts such as the one of redundancy, redundancy groups, events and prediction. These algorithms can be used to analyze both micro and macro traffic situations. Furthermore, in order to validate these algorithms, Monte Carlo simulation was performed and a simulator Simulo was implemented. The simulator’s unique characteristics represented the motivation for implementing a new simulator. Among these characteristics to be mentioned are the fact that Simulo is capable of simulating both micro and macro traffic situations, is capable of saving the previos test cases so that the same data sets to be used for other algorithms in order to obtain an objective comparison. The simulator was implemented considering a mathematical model that was developed for this purpose. All in all, the thesis proved the value of the proposed algorithms.