For commercial off-the-shelf antennas to used for medical body area networks they must be able to operate in an environment that they were not generally designed for, next to the human body instead of in free space, writes Simon Kingsley
Radio and wireless networks are finding an increasing role in improving our health and fitness.
In these health and fitness applications there may be a number of different biosensors placed on, or in, the human body and linked together or to a central unit so that the body’s performance can be monitored remotely. If on-body sensors are light enough, and the links are made wireless, then the patient or sportsman can have considerable mobility, thus improving patient/client acceptability.
These on-body sensor systems are often called wireless body area networks (WBANs) and might typically consist of several wearable body sensor units, each containing a biosensor, radio, antenna and some on-board control and computation.
Generally WBANs use low-power, short-range radio communication systems such as Bluetooth Low Energy (BLE) or ZigBee. BLE is a low power version of Bluetooth that is capable of reporting data from a sensor for up to a year from a small button battery without recharging.
Although the BLE data rate and radio range is lower than that of classic Bluetooth, the low-power and long battery life make it suitable for short range monitoring applications in medicine. ZigBee is a similar low-power, low data-rate radio protocol.
The short range limitations of these radio systems may be overcome by collecting the data from the various sensors at a local central unit and then retransmitting it over longer distances to locations where there are staff qualified to interpret the information.
In this case more powerful radio technologies such as GSM or 3G may be used for the long range link. This type of longer ranger monitoring is generally termed Telecare, Telehealth or m-Health (when mobile communications are used). These terms may also be used to describe the delivery of health information and advice by telecommunications networks.
Clearly the development of WBANs and radio monitoring of the body is an interdisciplinary activity requiring a number of different technologies to be brought together. When researching and developing these new and complex radio systems, an important question is whether the problem can be simplified and cost reduced by making use of standard commercial off-the-shelf (COTS) components.
Can commercial off-the-shelf antennas be used for telecare and body area networks? This is not a straightforward question because it requires antennas to work in an environment that they were not generally designed for, i.e. next to the human body instead of in free space.
COTS components have been a requirement in the military sector for many years and have been adopted by non-military sectors as well. To qualify as a COTS component it must be on sale in the general marketplace where it is available in significant numbers.
There are essentially three types of commercial antennas on sale – customised, standard and radio-antenna modules. Customised antennas are developed for products such as mobile phone handsets where space is limited, many different radio bands must be accommodated and there are other components such as speakers and cameras located in the same part of the phone.
The antenna must be specifically designed to cope with the idiosyncrasies of the phone but, when the device goes into production, the volumes are generally high enough to compensate for the development costs.
Standard antennas are COTS components and are not changed from application to application. They do not have to be purchased directly from the vendor as they can generally be easily obtained from component distributors.
Like most antennas, standard antennas usually require a matching circuit to ensure their input impedance is as close to 50ohms as possible. A matching circuit is just a small network of a few series and parallel capacitors and inductors lying between the antenna and the radio (which itself may not be a perfect 50ohms).
In most applications involving standard or COTS antenna the matching circuit needs to be specially designed, but this is not a significant effort or modification as it just represents a change to the bill of materials for the host application PCB and does not affect the physical construction of the antenna itself.
A low cost standard antenna with a suitable matching circuit would seem ideal for WBANs and telecare applications but there are two potential drawbacks that we must consider.
The first of these is that standard antennas are designed to work in free space or to be located on a PCB. The most important factor affecting the resonant frequency of an antenna is the relative permittivity of its immediate surroundings.
For free space this is defined as unity and for FR4, a common PCB material, it is around 4.4. The same antenna can generally be made to work in either location by making suitable changes to the matching circuit.
However, when worn next to the human body as part of an on-body sensor, we must consider the relative permittivity and the conductivity of the skin and body and how this affects the antenna. The relative permittivity of skin is about 37.5 and for blood it is about 58. The effect of these high dielectric constants, and the conductivity of the body, can cause both significant losses and a major detuning effect on the antenna.
The second drawback in using on-body antennas for health monitoring is that the patient/client is usually indoors and here the radio propagation can be severely disrupted by multipath.
Multipath is a well known radio propagation effect whereby the signal travels from transmitter to receiver by more than one path, often by scattering from internal surfaces such as walls, floors and ceilings. The interference between the main signal path and these other scattered signals can lead to cancellation of the radio signal, a deep signal fading effect and a consequent loss of data.
Because of the effect of the human body and the multipath problems, it is not immediately clear that COTS antennas can be used for WBANs and telecare.
The effect of the human body and the multipath propagation was the subject of a research programme at Body-Centric Wireless Sensor Lab (BodyWiSe) at Queen Mary University of London, led by Professor Yang Hao.
The work done by researcher Dave Waddoup used a COTS antenna. supplied by Antenova, which was a 2.4GHz surface mount antenna measuring 15 x 5 x 1 mm.
The antenna is essentially a monopole type of radiator and so must be used on a PCB having a partial ground-plane. Because there is no ground-plane directly under the antenna itself, it remains largely unshielded from the effect of the nearby body.
The researchers at QMUL found that, as expected, this antenna assembly became significantly detuned if placed in direct contact with the body. However, it remains adequately tuned if located 3mm away from the body and, consequently, this 3mm separation was used for all the subsequent ‘on body’ measurements.
The signal losses caused by the presence of the body were found to be about 19 dB compared to the same antenna in free space. This had the effect of reducing the radio link budget and reduced the range of the system.
Interestingly, the signal from the antenna was found to be almost completely depolarised whereas in free space the antenna is strongly linearly polarised. One way to counter both the signal fading due to multipath, and the loss of signal caused by the presence of the human body is to use a link architecture known as MIMO (Multiple Input, Multiple Output).
Here two or more antennas are used at either end of the radio link and multipath effects are exploited to improve the data link. MIMO is only successful, however, when there is good isolation between the closely spaced antennas at either end of the link.
The mutual coupling between antennas on the human body thus became an important measurement for QMUL to make. Widely spaced antennas are usually well isolated from each other but they would occupy a lot of space on the body and might not be convenient to wear.
Fortunately it was found that relocation of the antennas from free space to be ‘on body’ reduced the mutual coupling such that for a given level of coupling between two antennas, their separation could be reduced by a factor of about 2.5 times.
This substantial size reduction increases user acceptability and would be particularly significant for subjects wearing multiple on body sensors in a WBAN Telecare application.
I discussed custom and standard antennas but so far we have not mentioned the third type of commercial antenna, the radio-antenna module. Here the antenna and radio are designed as a single unit in order to increase efficiency and reduce size and cost.
It is also possible for radio-antenna modules to be made into COTS devices by arranging the pin-out so as to permit an external matching circuit to be used between the on-board antenna and radio.
There is every prospect that in future modules can be designed that incorporate the biosensor and control logic as well as the antenna and radio into a single small and lightweight unit. This should simplify the design of telecare systems and increase user acceptance even further.
Standard, or COTS, antennas are designed to work in free space and it is inevitable that they lose efficiency when used close to the human body. However, it has been found in work carried out at Queen Mary University London that the loss of RF performance can be tolerated and they can be used for short-range WBAN and Telecare applications.
Significantly, it has also been found that pairs of antennas may be placed closer to each other than they can be in free space, for the same degree of mutual coupling. This is an important result because it allows MIMO techniques to be used to improve the data link and opens the door for radio-antenna-biosensor modules to be developed that contain several antennas in a small space.
Simon Kingsley is chief scientist at Antenova