Leading researcher: Dr. László Jereb
Collaborators: Dr. István Dőry, Dr. Károly Farkas, Dr. János Kalmár

The four-layer architecture is a system model which helps to collect, store, process data from the measurements of the sensors, as well as visualize these data on the internet. At the University of West Hungary, Insitute of Informatics and Economics, we created the demo version of the model that exemplifies the way the system functions.

1. The goal of the system:

The primary goal of the system is to collect environmental data with the help of decentralized or mobile sensors located separately or in network; the processing, forwarding, archiving of the great amount of data and visualizing them on the internet and return them to the volunteer users.

The secondary goal is to collect data on technological and habitat for timber industrial and silvicultural research with mobile devices. The  real-time establishment of the correlation between data with the informatical support of the research and decision-making.

2. The elements of the architecture:

The model of the four-layer architecture is shown below (Fig. 1). The lower layer consists of sensors that measure different physical features, such as carbon monoxide and noise level, and microclimatic components. The sensors communicate via a wireless interface with the devices of the ad hoc network that forward the data to the data storing and processing layer through a landline or a wireless interface.

The data storing and processing layer records the measured data in the database on which further investigations can be carried out with data mining tools. When new data arrive, the data storing and processing layer informs the web server in the visualization layer, which is responsible for displaying data on the internet. So, the most recent measurement results can be accessed by browsing the internet.

The four-layer architecture is flexible, and we can adjust the model to the given task. For example, a system measuring NOx concentration does not  necessarily have to publish the measurement data on the internet, if its purpose is to alert when a given limit is exceeded. So the visualization layer may be omitted in this kind of environment.

3. Applications:

The model is very useful in a number of areas. With its help, the material turnover, the environmental and labour safety parameters of extensive factories and warehouses can be monitored; security, property protection, fire alarm and forecast services can be run; data collection of field researches can be supported. Our model can be used for systemizing and processing of global climate data.

4. The existing demo version:

The demo version of the model described above has been created at our institute, in which the measurements of the virtual sensor layer are given periodically by a special web application from the stored and collected data, as if they had been measured in real time. There are mobile tools in the network layer, which collect and forward data when within the communication radius of the sensor. Similarly, if the mobile devices approach the computer of the data storing and processing layer, which receives the data, then they send over the data via wireless connections. The recording of the data in the database takes place also in this layer, as well as the informing of visualization layer about the arrival of the recent data. The visualization layer makes the data including GPS coordinates accessible on a web portal, and provides visualization in a map and a table form. The portal provides both web2 interactive access and information on climate protection.