Design, Preparation, Implementation, and Evaluation of a Field Study for Wireless Sensor Nodes

Einleitung

The Wireless Sensor Node INGA will be used in a field study in the fifth year of the GAL project. The original intention of this field study is the collection of acceleration data, gathered by a worn accelerometer. Several probands are expected to wear a sensor node, which stores and/or transmits the recorded data. The analysis of the raw data is supposed to be performed offline – after the study.

Task Desciption

The tasks of this thesis can be divided into 4 phases: design, preparation, implementation, and evaluation, whereas the continuous evaluation of the constantly improved GAL system in “the real world” shall be the overall guiding principle.

The work shall focus on the recent developments of the Institute of Operating Systems and Computer Networks (IBR), and especially the interaction of the INGA Wireless Sensor Node and the OSGi based Multi Services Home Platform (MSHP). For evaluation purposes the gathered data from other areas (e.g. Home Automation sensors) can also be used.

The availability of probands will interfere with a strict serial handling of the 4 phases. Therefore, an iterative process should be established, so that a proband – when available – is equipped with a system running at the latest (stable) software release. In several iterations the findings of the current iteration shall be used to improve the next software release for the coming iteration.

The first phase is the design of the field study. In contradiction to the obvious goal of the study (“collecting accelerometer data”), the collection of other parameters is of significant value. During the design phase, relevant parameters from a computer networking perspective should be defined. These parameters should be recorded by all participating systems (MSHP and INGA) and should contain data that allows to draw conclusions about the condition of the wireless links, the system status (INGA and MSHP), as well as the sensor readings.

In the second phase, methods for the recording and transmission of the previously defined parameters should be implemented in the systems (MSHP and INGA). The data should be stored in a persistent way and should be easy to extract and easy to handle for the later evaluation. It is necessary to assign timestamps to all values (including the ones recorded on INGA), to allow a precise and synchronous assessment. Other work improving INGA and the communication protocols will probably be finished during the study; a contemporary integration in the field study could be helpful.

In the third phase the probands will be equipped with the devices and data will be gatherd. To achieve a satisfactory study, the system should be easy to deploy by a third party.

During the evaluation, all recorded data should be collected and processed. Especially the beforehand defined parameters should be taken into account, but nevertheless, also the recorded raw data should be presented. Due to the expectably iterative and dynamic character of the study, additional evaluations in the lab will be necessary to predetermine the behavior of different protocols under several conditions.