Highly integrated wireless sensing for body area network applications

The combination of highly integrated radio systems-on-chips, ultra-low-power processing, and smart textiles enables body area networks and novel human-to-machine applications.
23 December 2013
Vincent Peiris

Novel concepts in the fields of human machine interfaces (HMIs) and body area networks (BANs)sensors placed around the body—are paving the way toward tomorrow's Internet of Things (IoT). New devices can monitor a variety of physiological parameters such as heart rate, stress, position, and motion. BANs can collect information about an individual's health, fitness, and energy expenditure. There are applications of these concepts in gaming, where sensors can measure a player's motion and allow a video game to enrich the player's immersion experience. There is also a potential use in assistive robotics for medical patients or disabled people. Physical therapists may be able to conduct rehabilitation exercises using a robotics setup that controls and assists limb movements under remote surveillance. Lifestyle and sports applications could allow user technology to go beyond the classic chest strap monitors and wrist computers. In this context, the European Union project Wear-a-BAN1 (see Figure 1) aims to develop novel implementation concepts that will advance these new uses of HMI and BAN technologies.


Figure 1. A conceptual view of the Wear-a-BAN project. The project aims to realize unobtrusive human-to-machine applications, using wireless sensor nodes embedded in garments. BAN: Body area network.

Designers need to integrate BAN sensors directly into the fabric of garments using smart textiles, and sensors must transmit information wirelessly. A key technical objective of the Wear-a-BAN project is the integration of radio integrated circuits (ICs) and modules with body-worn antennas. The principal challenge in achieving this objective was to address the contradictory requirements of antennas. They need to be large enough to transmit signals even in complex propagation conditions. Conversely, they should also be as small and unobtrusive as possible for the comfort of the wearer.

To fulfill these needs, the Wear-a-BAN team developed a smart textile solution (see Figure 2). Our solution consists of a textile-based antenna.2 It is bendable and comfortable enough to be easily inserted into garments, while still yielding excellent performance. We used a textile system-in-package (SiP) concept3 for both the physical realization of the antenna and for the electrical connection from the antenna feeds to the radio module. The resulting unit underwent specific reliability and stress tests, including temperature cycling, performance with increased humidity, and bending.


Figure 2. A BAN module showing the conductive-textile-based antenna onto which a radio module is assembled with conductive adhesive. The antenna is then wrapped around the module. The module size is 3×2cm. Embedded in it are the icycom radio frequency system on a chip, sensors (accelerometer, magnetometer, gyroscope, microphone), a crystal, and a coin-cell battery holder.

Another key challenge is miniaturizing the electrical functions needed to realize a BAN node. The classic approach relies on application-specific standard product (ASSP) components with standardized wireless protocols, such as Bluetooth-Smart or Zigbee solutions. This approach takes advantage of the increasing number of standard ICs available from leading semiconductor companies. ASSP components offer versatile features for covering a broad range of applications, but they often come at the expense of nonoptimal specifications, like power consumption. We developed a dedicated system-on-chip (SoC), called icycom,4 that offers an optimal implementation of custom features that the mass of ASSP-based solutions cannot. While ASSPs can pack limited analog-digital functionalities on-chip—or else require high supply voltages—we were able to develop a complete radio SoC with digital, analog, and radio frequency (RF) functions operating at low 1V supply. Icycom integrates into a single silicon chip a mixed-signal sensor signal conditioning section, a 10-bit 10kS/s analog digital converter, a 32-bit digital signal processor with 96kB memory and a variety of digital peripherals, a complete 1V ultra-low-power RF operating in the 868/915MHz bands, and a versatile 3.3V/1.0V DC-DC power-management unit (see Figure 3). The total power consumption is less than 3.5mW in active receive (Rx) mode, and 4.5mW in transmit mode for −5dBm output power for the version integrated in Wear-a-BAN. The sensitivity is −100dBm for 200kb/s at voltages as low as 1V, making icycom optimal for operation with a tiny lithium coin-cell battery.


Figure 3. Die photograph of the icycom radio system on a 5×5mmchip, containing all functions required for use in a BAN.

On the networking software side, we developed and optimized an energy-efficient BAN-oriented communication protocol called Batmac5 for targeted BAN applications. Standard ASSP protocols can be used for basic BAN solutions where size and autonomy do not need to be considered. This is because their need to cover a broad range of applications results in high energy consumption, which makes them suboptimal for ultra-low-power operations. Further, they come with no or poor provision for addressing around-the-body and body-range propagation challenges. Conversely, the Batmac software includes self-organizing, adaptive, and flexible media access control protocol features that automatically detect the signal-reducing shadowing effect and quickly adapt the relaying scheduling to BAN changes related to close-to-the-body implementation of sensor nodes. A systematic approach was used to address the targeted HMI and BAN requirements thanks to hardware and software co-design aiming at energy optimization at all layers of wireless connectivity. We showed the Batmac protocol execution on the icycom platform within a real multinode BAN prototype built with several BAN nodes.

We demonstrated the relevance of a dedicated and customized approach within applications such as energy expenditure monitoring (BANs for healthcare), human-to-gaming interfaces, and HMIs. Icycom is now available for integration into novel products. The platform can be delivered with a complete hardware and software development tool set, thereby enabling easy adoption and smooth development of next-generation HMI and BAN applications.

Although we have created excellent miniaturized features with the Wear-a-BAN project, we are already conducting research to reduce the size of wireless modules even further by using MEMS (microelectromechanical systems) devices combined with ICs.

The author acknowledges technological contributions from all 13 partners of the Wear-a-BAN consortium. The Wear-a-BAN consortium acknowledges the funding of the project by the European Commission.


Vincent Peiris
Swiss Center for Electronics and Microtechnology (CSEM)
Neuchâtel, Switzerland

Vincent Peiris is head of the RF and Analog IC Design group at CSEM. He holds a PhD from the Swiss Federal Institute of Technology in Lausanne. His areas of interest are ultra-low-power RFICs, wireless sensor networks, and BANs. He is the scientific coordinator of the Wear-a-BAN project.


References:
1. http://www.wearaban.com Website of EU project 242473, Wear-a-BAN, for unobtrusive wearable human-to-machine interfaces. Accessed 20 October 2013.
2. M. Hirvonen, C. Böhme, D. Severac, M. Maman, On-body propagation performance with textile antennas at 867MHz, IEEE Trans. Antennas Propagat. 61(4), p. 2195-2199, 2013. doi:10.1109/TAP.2012.2234713
3. C. Boehme, R. Vieroth, M. Hirvonen, A novel packaging concept for electronics in textile UHF antennas, Proc. 45th Int'l Symp. Microelectron., p. 425-432, 2012.
4. D. Manic, D. Severac, E. Le Roux, V. Peiris, Cost-effective and miniaturized System-on-Chip based solutions for portable medical & BAN applications, Proc. 5th Int'l Symp. Med. Inf. Commun. Technol., p. 15-19, 2011. doi:10.1109/ISMICT.2011.5759787
5. M. Maman, L. Ouvry, BATMAC: an adaptive TDMA MAC for body area networks performed with a space-time dependent channel model, Proc. 5th Int'l Symp. Med. Inf. Commun. Technol., p. 1-5, 2011. doi:10.1109/ISMICT.2011.5759784
PREMIUM CONTENT
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research