Wireless Sensor Networks: The greatest invention since the Internet?

Although a relatively young technology, the potential of wireless sensor networks is encouraging intense research focus. Future systems are likely to require both small nodes and a high density of deployment, making efficiency and adaptability crucial to further development, says Professor Anders Rydberg.

The potential of wireless sensor networks (WSNs) – thousands of tiny monitoring devices which interconnect with physically remote environments – is of great excitement amongst the research community. A research team lead by Professor Anders Ryberg at the VINNOVA VINNExcellence Centre for Wireless Sensor Networks (WISENET), part of Uppsala University in Sweden, aims to design wireless interfaces for antennas and radios which also include stretchable methods of wireless interface integration. This goal involves finding technologies that not only improve antenna efficiency and adaptability, but also reduce the cost of antenna–sensor integration and enable fast, versatile antenna characterisation. “The number of potential applications of WSN is growing rapidly. It has even been predicted that in future its importance will match that of the Internet,” says Professor Anders Rydberg.

“Efficiency and adaptability are crucial to energy-optimised sensor networks, and are thus addressed in our proposal,” outlines Rydberg. “The objective of the sensor is to provide an energy-optimised, low power means of communication over a relatively short distance – between one and 100 metres – using very small volume sensors. The communication bit-rate is not expected to be very high – less than 100kbit – thus the bandwidth required for the antenna can be small. We are, together with colleagues, designing a stretchable, 3-D Conforming Wireless Interface Integration for Medical and Wearable Body Area Network Sensor Applications, and aim to improve the operating characteristics of the novel structures which act as the node platform in wireless body area networks (WBAN) applications, as well as other applications where stretchability and 3D conformity are important.”

Wireless sensors need to be intelligent, capable of operating in harsh environments and consume only a low amount of power, if they are to be successful. Many specialists in the field predict that future wireless sensor systems will require very small nodes and a high density of deployment, so size is another important factor. As sensors become smaller so the area available for the antenna is reduced, and while numerous innovative integrated antenna designs have been presented, they have been in cases where the specification was unrestricted, a shortcoming which is now being addressed.

“Recently the first steps have been taken towards complete, very small, low-power ad-hoc network sensor nodes with integrated microcontroller and compact radio with a very closely integrated small antenna. The antenna beam pattern for these sensors will be influenced by the integration technique, this is due to the small size of the antenna groundplane,” explains Professor Rydberg. Variations in the electromagnetic (EM) environment act as another limitation on system performance, so proper antenna design will ensure that not only is the system robust, but that signal reliability and transmission quality can be maintained. “Some applications will also require a reconfigurable beam pattern in order not to waste the energy of the sensor and lessen the need to scavenge for power. For small terminals at high frequencies much of our research is focused on antenna adaptability as a means to adapt to a changing environment,” continues Professor Rydberg. “As antennae grow more complex, and product costs are reduced, a need for new characterisation methods emerges. These should be fast, while also allowing for advanced tests of multiple antenna interaction and antenna adaptability. Performance tests in various types of environment bring yet another dimension to the problem of antenna characterisation. While some of the envisioned applications can be approached with existing WSN technology, others rely on further research on sensor size, longevity, energy consumption, low-cost building practice, and characterisation methods.”

Prime among these potential application areas is e-Health: the application of Information and Communications Technologies (ICT) in the health sector. With demographic changes increasing the burden on European health services, administrators are keen to use technology to ensure that specific patient needs are met and process efficiency is improved. Versatile, reliable systems that make patient- and therapy-specific data available to clinicians in real time are particularly important. “Remote patient supervision uses wireless biosensors, bio-data analysis and communication technologies and offers a very important opportunity for cost saving. Research into the integration of different types of sensors with textiles, or directly on the body in connection with a wireless body area network (WBAN), has recently been undertaken in many countries. This wireless network is a significant breakthrough. It enables the first-time monitoring of body fluids and biochemical measurements via sensors distributed on a textile substrate, either in cloths or on the body,” explains Professor Rydberg.

“We are working from a body-centric perspective, and together with both our academic and commercial partners in WISENET are developing a complete WBAN,” continues Professor Rydberg. “This involves using a number of tools to measure movement (displayed by an avatar), including GPRS for the transfer of data, GPS to monitor the position, and also a mini-Linux server in the belt to control the data stream. We are also developing the wireless interface for health and fitness sensors. We collaborate closely with Philips in this work, and are working on wave propagation along the body to improve communication between sensors.” This research holds real potential, but is only one of the strands being pursued: the team are also working to develop WSNs for hostile environments.

This interest from the commercial sector further underlines the wide potential of microwave technology, as illustrated by the development in WISENET of wireless sensor networks for the Swedish Railroad Authorities, for use in measuring the temperature and vibrations of ball bearings on train wagons. This kind of hostile environment places significant strain on WSNs, and with reliability a key issue, networks must undergo rigorous testing and validation procedures before they can achieve practical application. This involves using a test-bed belonging to the Swedish Railroad Authorities to analyse and evaluate not only improvements in the wireless interface (antenna and radio), but also wave-propagation between sensors and overall reliability. This potential for continued improvement of WSNs is encouraging Rydberg to pursue further research in this area.

“However, there are also a number of general constraints in WSN development,” explains Professor Rydberg. “The lifetime of a sensor network depends heavily on battery capacity, energy consumption and the ability of the node to generate electrical energy from the environment. All parts of a sensor, as well as the system as a whole, must cooperate to minimise energy consumption, which will demand research across a number of disciplines. In future there will be a spectrum of sizes and capabilities of WSN nodes, and we will work on all sizes. The smallest are likely to be very simple, cheap and smaller than a sugar lump. We believe that the integration of all parts of a WSN node is the important issue, not the size per se. Integration is attractive for two main reasons; firstly that an integrated node can be mass-produced and hence made inexpensive; and secondly, that an integrated unit is likely to be more reliable.”

For more information on the research, contact Professor Anders Rydberg at anders.rydberg@angstrom.uu.se

(Below: An illustration of the wireless sensor system and the potential applications) 

Published: Thursday, 4th March 2010 by Tom Freeman

Comment on this article

Comments

No comments for this article, be the first to comment