Sensors, which are increasingly present in our environment, are a very useful tool for achieving scientific progress. They are used to collect and monitor environmental data such as humidity, light, temperature, vibrations, movement, pollution, and sound. Sensors have limitations, particularly in terms of memory, computing power, and energy storage, but they also offer a series of advantages, including their small size, light weight, and low cost.
Despite their simplicity, these autonomous devices can play an important computational role in scientific research when used in a distributed and cooperative way. Together, using local radio-frequency communication between sensors, they can perform complex computational calculations.
This is the rationale behind the European research project known as Wisebed, funded under the European Union's Seventh Framework Programme. The pioneering project in Europe, which began in 2008, has led to the creation of a sensor network that will serve as a laboratory or testbed that scientists can use to test the computational power of their algorithms.
Up until now, the possibility of executing algorithms on sensor networks was subject to the availability of this technology. As a result, theoretical results and resulting errors could frequently be studied only in simulations. "Wisebed will provide the scientific community with a real testing laboratory where researchers can carry out their experiments without travelling to a specific physical location or investing in hardware", stated Maria J. Blesa, a researcher with the UPC's ALBCOM group.
Though the team directed by Josep Díaz has participated in developing the hardware for the project, the main focus of their work has been on creating algorithms for time synchronisation between sensors, hierarchical organisation of the network (to speed up communication), and self-regulation of the energy consumed by sensors to maximise their lifetime while maintaining the required connectivity.
The main advantage of this technology is that it can be used to make calculations in a distributed manner. Communication does not depend on a central unit that sends commands to the sensors; instead, they interact with each other within defined areas and the network is self-managing. "This way if part of the network goes down, we know that the rest of the system is intact and operational", stated Maria J. Blesa.
The project, led by the University of Lübeck, Germany, has been carried out with the collaboration of universities and research centres in Germany, Greece, Switzerland, the Netherlands, and the United Kingdom. The platform, which during the initial stage will be available only to project members, is based on local connectivity. It links sensor networks located in each of the participating countries to create a global network of over a thousand sensors. In the case of the UPC, the sensors are located inside and outside the Omega building.
European scientists will soon be able to use the laboratory via a website that will allow them to access the various sensor networks of which it is composed, choose a configuration, and control the computation process. Beyond the area of informatics, it is anticipated that the platform will be used in other fields such as biology, physics and environmental science, by academic users, companies and institutions.
One of the key software elements developed for the project is WISELIB, a library that sets out the guidelines and basic rules for use of the laboratory and contains 50 specialised algorithms for sensor networks. The library allows users to execute their algorithms without concerning themselves with the specific technological features of the sensors, provided that the rules established in the library are followed.
Although initially designed to contain algorithms used in computer science, the library can be extended to include algorithms from other branches of science and technology.
Other elements developed for the project include sensor network visualisers, a web-based reservation system, and WISEML, an XML-derived language for standardised description of data collected by sensors.