Measurement and Performance of Textile Antenna ... - IEEE Xplore

4 downloads 217 Views 2MB Size Report
Efficiency on a Human Body in a. Reverberation Chamber. Stephen J. Boyes, Student Member, IEEE, Ping Jack Soh, Student Member, IEEE, Yi Huang, Senior ...
IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 61, NO. 2, FEBRUARY 2013

871

Measurement and Performance of Textile Antenna Efficiency on a Human Body in a Reverberation Chamber Stephen J. Boyes, Student Member, IEEE, Ping Jack Soh, Student Member, IEEE, Yi Huang, Senior Member, IEEE, Guy A. E. Vandenbosch, Fellow, IEEE, and Neda Khiabani, Student Member, IEEE

Abstract—With the advent of on-body communication research in recent years, there is a growing need for antenna developments that satisfy a wide criteria (one being minimal efficiency degradation) in order to be integrated successfully onto human subjects; one promising development is the textile antenna. In this paper we investigate the efficiency performance of some newly designed small sized textile antennas on live human subjects using a reverberation chamber. First, we show that the material selection of these textile antennas can have a crucial effect on the on-body frequency detuning and efficiency levels, as via a comparison we determine that a lossier textile antenna in free space can actually outperform a higher free space efficient textile antenna when placed on-body. This has a profound impact on the material design choices for these small sized antennas. Second, we investigate the performance effects under bent conditions and finally we show that the overall performance of the textile antenna can be mitigated somewhat by variations in on-body distances from the human subject. It is revealed that in some cases a small 20 mm distance from the body is sufficient for the radiation efficiency to approach the free space levels. Theoretical, simulated and experimental evidence is presented to verify the conclusions. Index Terms—Antenna efficiency, on-body communications, reverberation chamber, textile antennas.

I. INTRODUCTION

A

GROWING interest is evident concerning body centric wireless communication technology, aimed at providing solutions for a wide range of applications from the medical/ healthcare industries, the consumer electronic industries, wearable technology for the fashion industry and the military sector to name but a few. A crucial component in the body centric wireless communication ‘chain’ concerns the antenna device itself; the antenna should be ergonomically suitable for integration onto a human body. Furthermore, the antenna should not be Manuscript received April 13, 2012; revised June 20, 2012; accepted October 11, 2012. Date of publication October 19, 2012; date of current version January 30, 2013. This work was supported in part by the Engineering and Physical Science Research Council (EPSRC) U.K, the Malaysian Ministry of Higher Education (MOHE), and the EU-FP7 CARE Project. S. J. Boyes, Y. Huang, and N. Khiabani are with the Department of Electrical Engineering and Electronics, University of Liverpool, Liverpool L69 3GJ, U.K. (e-mail: [email protected]; [email protected]; [email protected]). P. J. Soh and G. A. E. Vandenbosch are with the ESAT-TELEMIC Research Division, Department of Electrical Engineering, Katholieke Universiteit Leuven, 3001 Leuven, Belgium (e-mail: [email protected]; [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TAP.2012.2225817

obtrusive, it should be able to maintain flexibility, exhibit minimal degradation in terms of bandwidth, reflection coefficient and efficiency performance, and be manufactured in a low cost manner. To satisfy this broad criterion, the textile antenna has been developed and continues to receive ever growing attention [1]–[3]. When communicating wirelessly in the proximity of a human body, the propagation channel is dependent upon the body condition, the human activity being performed, the antenna position, the immediate surrounding environment and any interaction between the human body and the antenna [4]–[6]. Therefore, the radio propagation channel in this (on-body) scenario directly includes the body effect and is not usually stationary. It becomes important therefore to be able to characterize any antenna, specifically designed for use “on-body,” in a realistic scenario (i.e., on a human being) to attempt to accurately quantify the effect of the human body on any associated antenna parameters of interest. In this paper, the quantity of interest mainly concerns the radiation and total radiation efficiency of some newly designed textile antennas. The measurement facility chosen to perform the characterization is the reverberation chamber (RC). The RC is now well known from studies such as [7]–[12], and can be described as an electrically large shielded metallic enclosure which is designed to work in an “over-mode” condition. The fields inside this environment can be perturbed by various means to engender the average field distribution statistically homogeneous, as long as the field location is situated approximately from the chamber walls [13]. The field statistics inside the chamber have received much theoretical and practical treatment over the years and are well known, for example see [14]. To review, the received complex signal in this environment with no line of sight path manifests itself as being complex Gaussian (normally distributed), the magnitudes are Rayleigh distributed, the power exponentially distributed and the phase uniformly distributed over . It should also be noted however that a Rician distributed environment can also be emulated in the chamber as proposed in [15]. The choice of measurement facility here has been chosen in part because of the statistically emulated environment it can offer, but also because we feel it can offer an easier, more robust/less uncertain measured solution to the problem of textile antenna efficiencies when acquired on a human body. To explain further, it is inevitable that some human movements will be present during any measurement procedure (from breathing, etc.). However,

0018-926X/$31.00 © 2012 IEEE

872

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 61, NO. 2, FEBRUARY 2013

II. DEFINITIONS A. Antenna Efficiency For the quantity of radiation efficiency we refer to the IEEE definition of “Ratio of total radiated power to net power accepted by the antenna at its terminals.” Mathematically, when acquired using a RC, (1) applies [17] (1)

Fig. 1. Single band textile antenna radiating elements [ShieldIt radiating patch (left) and Copper radiating patch (right)].

Fig. 2. Single band textile antenna ground planes [ShieldIt ground plane (left) and Copper ground plane (right)].

in the RC environment these can be tolerated as they would add to the randomness of the fields inside the chamber which can be viewed positively. To the best of our knowledge only one experiment concerning antenna radiation efficiency using live human test subjects in an RC exists, see [16]. The conclusions drawn from this piece of work indicated that the RC was a repeatable test bed for wearable antenna measurements using live human subjects; this potentially validates the prior stated reasoning on the choice of facility here. Further, the results presented indicated that the RC results obtained with human test subjects were sufficiently representative of results acquired using tissue phantoms—validating the live human approach against phantom models often used in the field [16]. The antennas used in the study consisted of four low profile microstrip patch antennas constructed on a dielectric substrate and a quarter wavelength monopole, with the results indicating that the type of the antenna plays a more dependent role in determining the overall radiation efficiency performance when placed “on-body” as compared to the characteristics of any human subject. Elements for future work stated in [16] highlighted a need for characterizing flexible/fabric antennas on body and also an assessment of movement related bending effects. This, therefore, is what this paper seeks to address. This paper is organized as follows. Section II will first define the quantities of interest in this study and the mathematical formula behind them to avoid any ambiguity, and second, introduce the antennas under test (AUT). Section III presents detailed information concerning the RC measurement procedures used in this manuscript and Section IV will disclose all results pertinent to the investigation. An assessment of the measurement uncertainty is provided in Section V to benchmark the accuracy of the results presented.

where: average of the scattering parameters, absolute value, transmission coefficient, reflection coefficient, antenna under test, reference antenna and known reference efficiency. In our case the reflection coefficient quantities here are not signified as being an ensemble average, this is because we choose to acquire this particular parameter in an anechoic chamber as it was quicker for us to do so. It should be noted however that if the reflection coefficients are acquired in an RC then the quantity should be signified as being an ensemble average. For the quantity of total radiation efficiency (or total efficiency for short) we imply the definition of “Ratio of total radiated power to the power incident on the antenna port”. Mathematically, this is a product of the radiation efficiency and mismatch efficiency of an antenna as detailed in (2) (2) Again, the reflection coefficient parameter in our case is acquired in an anechoic chamber hence the omission of the ensemble average. B. Antenna Descriptions It is beyond the scope of this manuscript to present any optimization analysis of the antennas under test; therefore in this section they will only be briefly described. For further detailed information pertaining to the designs please refer to [3]. One distinct category of textile antenna is to be investigated in this paper—designed for single band use at the 2.45 GHz industrial scientific and medical (ISM) frequency. The single band antennas are based around the same PIFA topology and both are constructed using a 6 mm thick felt fabric ( and at 2.45 GHz [3]). The felt fabric is sandwiched between a ground plane and a slotted radiating patch, with two different conductive textile materials being subject to investigation [18], [19]: 1) copper textile (plain woven and coated, 0.08 mm thick, at 2.45 GHz); 2) ShieldIt™ conductive fabric (0.17 mm thick, at 2.45 GHz). The profile of each antenna can be viewed in Figs. 1 and 2. III. MEASUREMENT PROCEDURES The measurements utilized the RC at the University of Liverpool. The chamber is constructed from galvanized steel and the dimensions are , and

BOYES et al.: MEASUREMENT AND PERFORMANCE OF TEXTILE ANTENNA EFFICIENCY ON A HUMAN BODY IN A RC

. The chamber is fitted with two principle sets of mechanical stirring paddles that rotate about a central shaft; one set configured for vertically polarized waves mounted towards a corner of the chamber from the floor to the ceiling, the other configured for horizontally polarized waves mounted from the front to back wall at ceiling height. The paddles are therefore configured to be able to stir orthogonal volumes of the chamber. As with any efficiency measurement in this facility, the chamber required an initial reference measurement for calibration purposes. This was performed using a dual ridge horn antenna (Satimo SH2000), the antenna being selected as its unidirectional pattern characteristics were similar to that of the AUT’s. For the single band antennas a frequency range of 2–3.5 GHz was selected using 801 frequency data points, with the number of data points being selected to ensure that a sufficiently large number of modes would be excited in the chamber throughout the measured range. The calibration needed to be performed while the human subject was located inside the chamber, as the presence of the human subject would severely load the chamber. As such, it was expected to be the dominant contributor to any loss mechanisms that exist (more so, than for example, any wall losses, losses from aperture leakage and any antenna loss). The stirring sequences used throughout (same for reference calibration and AUT measurement) comprised the following parameters: 1) mechanical stirring—5 degree (71 measurements); 2) polarization stirring—two orthogonal linear polarizations; 3) position stirring—5 separate receiver locations The above stirring mechanisms gave rise to a total of 710 measured samples per frequency point for both the reference and AUT measurement. The reasons behind the stirring sequences were twofold: firstly, a large number of samples (a percentage of the total being statistically independent) are required to keep the uncertainty in the measurements to an acceptable level, and secondly, the sequences meant that the human subject did not have to spend prolonged amounts of time inside the chamber to complete each measured sequence (i.e., the subject could take regular breaks if need be). During the reference measurements, the human subject had the AUT attached to the body via a rigid cable which was terminated in 50 ohms. During the on-body AUT measurements, the reference antenna remained inside the chamber—connected via the same rigid cable and terminated in 50 ohms. By this, the chamber loading was configured exactly the same from the reference to the AUT measured sequences to ensure accuracy; no aspects differed in this regard whatsoever. Indeed, even the contents inside the human subjects pockets, etc. was always kept constant (no mobile phone, coins, etc.) to ensure that consistency prevailed. Fig. 3 illustrates the set up used during the reference measurements. The distance between the consecutive position locations shown above was configured such that each location was larger than the correlation distance (about ½ wavelength) to create independent samples. The exact locations are summarized in the following table. Further, throughout all position stirring locations (for both the reference and AUT measurements), the separation between

873

Fig. 3. Measurement setup (reference in this case).

TABLE I POSITION STIRRING LOCATIONS

human subject and reference antenna was always kept as a prescribed distance apart to avoid any coupling issues that could potentially corrupt any measured results. IV. MEASUREMENT INVESTIGATION In this investigation we will compare the radiation and total radiation efficiency values of each AUT on different body locations with respect to their free space efficiency values. By this we not only want to assess which antenna yields the highest efficiency on body and in what location, but also chart the difference in magnitude of the efficiency quantities to assess what degradation occurs when the textile antennas are located on-body in specific locations and proximity distances. Please note that all free space efficiency values have been measured in the same chamber by the same principle author, using the same stirring mechanisms to be able to compare directly. The measured free space radiation and total radiation efficiency values that are supplied are compared by simulated efficiency values provided from CST Microwave Studio (based on the finite integration technique) by an assessment from the gain/directivity [20]. All efficiency values herewith for the single band antennas have been frequency stirred by 20 MHz (i.e., over 11 frequency data points); the 20 MHz frequency stirring value was selected such that it was far less than the bandwidth of the antennas to minimize the reduction in frequency resolution. This section is subdivided into three smaller sections to assess both antennas performance at given body locations and proximity distances from the human test subject. From here on in, the AUT’s will be referred to by their manufactured name: Copper textile radiating element and ground plane, ShieldIt textile radiating element and ground plane.

874

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 61, NO. 2, FEBRUARY 2013

Fig. 4. On chest 0 mm measured setup.

Fig. 5. FLSL 0510 radiation efficiency space.

at 0 mm for on-body chest & free

A. On Body 0 mm Measured Results 1) Chest 0 mm: In this investigation the antenna was located on the centre of the human subjects’ chest, 1.38 m from the floor. The 0 mm dimension in this case does not mean that the antenna was physically touching the human subjects’ skin; rather, the human subject was wearing a woolen jumper 1.5 mm thick (constant throughout all experiments) and it was this which the antenna was placed against. The antenna element was not physically strapped to the human by any means—standard Velcro straps were used to fix the attached cable to the human being which was sufficient to hold the antenna to the chest of the subject. Throughout the measurement sequences this aspect was rigorously checked and at no time did the antenna shift location. Fig. 4 illustrates the measured set up. In Fig. 4, the human subject faces the back wall with the antenna on his chest. This aims to block any line of sight path from the transmitting antenna which is located behind the human subject (off shot). The proximity distance between the body and the antenna in this particular investigation has been chosen to assess the ‘worst case’ possible estimate that could be witnessed; that is, we expect the severest loss. The antenna orientation in Fig. 4 (normal to the chest) has been chosen because of the location of the SMA connector in the antenna design, meaning that it was easier to secure the antenna in position at the 0 mm proximity distance in the manner shown. The radiation efficiency results for FLSL placed on the chest (0 mm) are disclosed in Fig. 5. Three separate measurements have been completed for this antenna and body location to provide an indication as to the measurement repeatability. A comparison with the free space radiation efficiency levels is also provided so that the degradation in radiation efficiency can be easily assessed. The total radiation efficiencies are shown in Fig. 6. The discrepancy between the measured and simulated cases witnessed can be attributed in part to differences between the measured and simulated reflection coefficients. This in turn resulted from slight differences in the fabricated dimensions between the simulated model and physical design, and also the presence of the SMA connector. To address the nature of the antennas orientation at the 0 mm proximity distance; specifically the fact that the radiating element is mounted perpendicular to the chest as opposed to being

Fig. 6. FLSL 0510 total radiation efficiency free space.

at 0 mm for on-body chest &

Fig. 7. 0 mm measured set-up for alternative antenna polarization.

parallel, an investigation was undertaken to compare if any differences existed. This investigation utilized the exact same measurement parameters and procedures as previously discussed to ensure that an accurate comparison could be formed. The on-body antenna location was configured such that the cable was mounted and secured under the arm, enabling the antenna radiating patch to be away from the human body. The exact same (0 mm) proximity distance was also maintained with Fig. 7 illustrating this particular set up.

BOYES et al.: MEASUREMENT AND PERFORMANCE OF TEXTILE ANTENNA EFFICIENCY ON A HUMAN BODY IN A RC

Fig. 8. FLSL 0510 comparison of deduced efficiency versus antenna orientation at the 0 mm proximity distance.

The main purpose for this investigation was to assess if any differences are observed with respect to the antenna’s polarization. For example, in on-body applications for consumer electronics, the users’ movements could potentially alter the antennas polarization when worn on the clothing, which could have an impact on the antennas performance. From Fig. 8 we see that the resultant magnitudes of efficiency at 0 mm are very close indeed, irrespective of the antennas perpendicular or parallel on-body orientation. The differences in the largest case are in the order of 4%. The reason for the similarity in the results is believed in part to be down to the emulated multipath environment which has contributed to help overcome any differences in the antenna radiation patterns with respect to the antennas on-body orientation [21]. Furthermore, because the antennas ground plane is small in dimension, we believe that this also a contributing factor as to why the different antenna orientations produce similar results—i.e., the results shown are valid irrespective of the (0 mm) antenna orientation in this case. The parallel results measured in Fig. 8 were acquired on a different human being than the perpendicular case which collectively shows that any efficiency results here are not too sensitive to the on-body antennas orientation within reason, and furthermore that the efficiency results are reproducible to different human subjects. As stated prior, the reflection coefficients on body were acquired in the anechoic chamber (AC) at the University of Liverpool. All antennas were strategically located in exactly the same place as where the RC measurements took place on body to ensure accuracy in the measured reflection coefficient—this aspect was checked and consistent results were obtained from both RC and AC facilities. The measured on body (chest 0 mm) reflection coefficients are issued in Fig. 9 and are compared with the free space performance. The discrepancy between measured and simulated quantities can be explained by slight differences in the fabricated dimensions between the simulation and measurements, and the effect of the SMA connector represented by a simplified model in the simulation. From Figs. 5 and 6 we see that the levels of radiation and total radiation efficiency are degraded considerably from the free space efficiency levels. At the start of the measured frequency band (2 GHz) this is as much as 78.15% for the radi-

875

Fig. 9. FLSL 0510 reflection coefficients (dB) at 0 mm for on-body chest and free space.

Fig. 10. SHSL 0510 radiation efficiency free space.

at 0 mm for on-body chest and

ation efficiency and 72.65% for the total radiation efficiency. The trend of the on body efficiency is also seen to rise over the first 600 MHz measured band; this believed to be attributable to the coupling between the human body and the antenna (remembering the antenna is not physically touching the human skin). Thus as the frequency begins to increase, the inter element (human body and antenna) spacing is also increasing. From the measured reflection coefficients the antenna is also seen to exhibit a degree of detuning. At 2.45 GHz the mismatch efficiency measured on the chest is 78% as compared to the free space value of 91%. With regards to the measurement repeatability, it is seen that the three separate measured runs undertaken yield very consistent levels of efficiency; this provided increased confidence that the measurement procedures undertaken were thus consistent and not liable to wild fluctuations from one measurement to the next. The difference between the measured runs is in the order of 2% at its maximum. For the ShieldIt material (SHSL 0510) the measured set up and procedures differed in no way from the prior (FLSL 0510) 0 mm chest investigation so that a direct comparison could be formed; therefore, Fig. 4 again illustrates the set up. The radiation efficiency results for SHSL 0510 at 0 mm on the chest are presented in Fig. 10. The total radiation efficiency results for SHSL 0510 are presented in Fig. 11.

876

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 61, NO. 2, FEBRUARY 2013

Fig. 11. SHSL 0510 total radiation efficiency (%) at 0 mm for on-body chest and free space.

Fig. 13. On-body (elbow 0 mm) measurement set up.

Fig. 14. FLSL 0510 bent elbow (0 mm) radiation and total efficiency versus free space and bent elbow reflection coefficients (dB). Fig. 12. SHSL 0510 reflection coefficients (dB) at 0 mm for on-body chest and free space.

Comparing Figs. 5 and 6 with Figs. 10 and 11 we find that the textile antenna constructed with the ShieldIt type material for the radiating element and ground plane is lossier in free space conditions than the copper textile material—this is as predicted since we expect the higher conductivity material to yield higher radiation efficiencies. However, when placed on the body, the lower conductivity (ShieldIt) material exhibits a larger amount of radiation and total radiation efficiency than its copper based counterpart—differences of 30% are evident for the radiation efficiency and 25% for the total radiation efficiency at 2.45 GHz. Theoretical and simulated evidence to explain and support these trends can be found in section C. The reflection coefficient of the SHSL 0510 antenna on-body (chest 0 mm) is shown in Fig. 12. By comparing Figs. 9 and 12, we find that the ShieldIt textile antenna also exhibits a lesser amount of detuning as compared to the copper based counterpart. The mismatch efficiency at 2.45 GHz for the ShieldIt material is calculated at 90% as opposed to the 78% recorded for the copper textile material. 2) Bent (Elbow) 0 mm: This investigation looked to assess the effect on radiation and total radiation efficiency under a bent configuration. The body area selected was the elbow. Fig. 13 illustrates the set up used. This time the antenna had to be secured with the standard Velcro ties owing to the loading placed on the antenna through

bending. The bending radius (measured from the centre of the elbow to the edge of the elbow was 55 mm). The antenna element was selected as the copper based textile (FLSL 0510) and was located 1.2 m from the chamber floor. The human subjects’ arm was fixed in place by a Velcro tie which was strapped around the arm and shoulder. This was sufficient to hold the subjects’ arm in place to prevent movement. The radiation efficiency, total radiation efficiency and reflection coefficient results can be viewed in Fig. 14. Comparing to the 0 mm chest measurement (see Fig. 5) the radiation efficiency values at 2 GHz would appear slightly higher, believed to be due to the absence of major human organs near the antenna, but in the mid and upper frequency range the values became comparable. The comparable nature of the mid and upper frequency range results are partly expected owing to the overmoded characteristics of the RC—(the fact that the angle of arrival of plane waves come from every conceivable direction with equal probability meaning the radiation patterns of the AUT play no part [22]). However, the main effect here concerned the total radiation efficiency. The levels are 12% to 15% lower than the 0 mm chest values (Fig. 6) owing to the severe detuning in the magnitude of reflection coefficient; hence the mismatch efficiency of the antenna at 2.45 GHz was reduced to 62% as compared to the 78% witnessed on the chest at 0 mm and 91% in free space. The conclusion drawn from this investigation was that severely loading the antenna under a bent configuration is not good practice, and on the evidence of the results

BOYES et al.: MEASUREMENT AND PERFORMANCE OF TEXTILE ANTENNA EFFICIENCY ON A HUMAN BODY IN A RC

Fig. 15. FLSL 0510 chest 20 mm radiation and total efficiency space and chest 20 mm reflection coefficients (dB).

versus free

seen here perhaps should be avoided when an on-body antenna location is to be chosen. In respect of these conclusions, this experiment was not repeated. B. On Body 20 mm Measured Results 1) Chest 20 mm: The next investigation undertaken looked to assess the effect of having different spacing between the antenna and the human body—i.e., if the interelement spacing were to be slightly increased, what difference in the magnitude of the radiation and total radiation efficiency could be obtained. The theoretical rationale guiding this investigation we considered to be a function of (expected) decreased coupling between the antenna and human subject; noting that the medium that is coupling to the antenna (i.e., the human being) will present a significant loss if the antenna is close enough. The antenna was again located in the centre of the subjects’ chest, 1.38 m from the floor. The 20 mm chest separation was achieved by employing a 90 degree elbow connector to keep the antenna at a fixed distance. The radiation efficiency, total radiation efficiency and reflection coefficient performance of the copper textile (FLSL 0510) at this location and proximity distance can be viewed in Fig. 15. Comparing Figs. 5, 6 and 15 we find that the 20 mm distance from the human subjects’ chest yields a clearly higher value of both radiation and total radiation efficiency than the levels witnessed on the chest at 0 mm. A 20% to 30% increase was apparent in radiation efficiency in this case, and a 15% to 19% increase in total radiation efficiency across the measured band. Further, the radiation efficiency levels in Fig. 15 are also seen to rise with increasing frequency which gave confidence to believe that the coupling theory could explain this trend. Further, we can also see that the added distance from the human subject has reduced the loading effects sufficiently enough such that no negative detuning is evident at this proximity distance. The ShieldIt material (SHSL 0510) performance at this location and proximity distance from the human subject can be viewed in Fig. 16. Comparing Fig. 16 with Figs. 10 and 11 we find that again the 20 mm distance off body can yield a higher level of radiation and total radiation efficiency as compared to the 0 mm distance between antenna and human body. This trend can be explained again by the reduced coupling between the antenna and human body due to the added (electrical) distance between both

Fig. 16. SHSL 0510 chest 20 mm radiation & total efficiency space and chest 20 mm reflection coefficients (dB).

877

Vs free

elements. Further, in this case we see that the antenna towards the end of the measured band has decoupled itself sufficiently from the human body such that the radiation efficiency levels approach the free space levels witnessed. This fact could prove useful when antennas with multi band operational frequencies are used (progressively higher frequencies), such that the higher frequency bands suffer less of a radiation efficiency degradation due to antenna/body coupling—particularly so when located 20 mm off the body as seen here. If we assess the effect on the measured reflection coefficient we find that again no negative detuning has taken place at all at the 20 mm distance using this antenna—the antenna is operational over a similar bandwidth as free space. It is proved therefore that certain antennas can be placed in a (relatively) close proximity to the human body in an operational role, and suffer only minor detrimental performance effects. C. Theoretical & Simulated Evidence Drawing conclusion from Section III parts A and B we can clearly see that the antenna constructed from the lower conductivity, thicker textile material (ShieldIt—SHSL 0510) outperforms the thinner, higher conductivity copper textile antenna (FLSL 0510) when placed in various proximity distances of the human body in terms of both the efficiency and the frequency detuning levels. This observation is in stark contrast to the free space (efficiency) case and counter-intuitive to what we would normally expect. The reason for the increased levels of efficiency and decreased levels of detuning can be explained as follows. It is stated in [23] that demands placed upon antennas with small ground planes, when placed near the proximity of a human body, can result in an interaction with the reactive near fields of the antenna and cause a loss. The reason here for the difference in the magnitude of the radiation efficiency is due to the fact that the lower conductivity material based antenna has given rise to lower electric fields in the body as more power has been lost in the antenna itself for the same input power. This is understood to have had the effect of causing lower losses in the human body as opposed to the higher conductivity material based antenna. To help reinforce this statement and provide evidence that the theory can explain the trend witnessed, two simulated models have been adopted using CST Microwave Studio. The models

878

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 61, NO. 2, FEBRUARY 2013

From a design perspective, the use of a specific human Voxel model is believed would improve the agreement between the simulated and measured on-body efficiency values. As a means of simple verification however, the simulation model is seen to be sufficient enough in this case. V. MEASUREMENT UNCERTAINTY

Fig. 17. Comparison of simulated radiated power (mW) and radiation effiat 0 mm on emulated muscle material. ciency

For the measurement uncertainty we consider the following aspects. Firstly, it is known from [25] that direct coupling can be a major source of uncertainty inherent during over the air (OTA) measurements in an RC; therefore direct coupling, usually expressed as a Rician K factor, should be as small as possible. Models were also presented in [25] that equate the total standard deviation of the average power transfer function to be comprised of the non line of sight number of independent samples , the line of sight number of independent samples and the Rician K factor as detailed in (3) (3) , where: and average Rician K factor, comprising the samples obtained from mechanical stirring, the various receiver position locations and different transmit polarizations used for polarization stirring. Equation (4) details the calculation of the Rician K factor [15]. (4)

Fig. 18. Comparison of simulated electric fields (V/m) at 10 mm depth inside muscle emulated material.

have been mounted (at 0 mm) onto a structure whose material parameters have been chosen to emulate muscle at 2.45 GHz ( and [24]). Fig. 17 depicts the simulated radiated power and radiation efficiency from the two models, showing clearly that the copper based textile antenna (FLSL) radiates less power into free space than the SHSL counterpart. Further, by definition, it is also seen to be less efficient when placed in conjunction with the simulated lossy structure. Fig. 18 depicts the simulated electric field magnitudes in the emulated muscle structure at a depth of 10 mm. A clear difference is shown between the two antenna models which reinforces the theory that explains this trend. By comparing the simulation results in Fig. 17 with the measured results from Figs. 5 and 10 (that is, the simulated and measured radiation efficiencies at 0 mm) we find that the same conclusions are upheld and validated from the measurement cases, but the measurement and simulated efficiency values do vary between 7% to 20% in some cases. The reason for this can be attributed to the simplistic nature of the adopted simulated model in that it has only taken into account the muscle parameter. We know obviously that the physical human being is far more complex. However, the primary purpose here was to underpin and supplement the theoretical and measurement evidence, with a secondary aim being to provide a simple and resource efficient simulation model which can be easily reproduced.

Equation (5) presents the calculation of the standard deviation in dB form [26] (5) To calculate the non line of sight number of independent samples we employed the use of the autocorrelation function as defined in [27]. After which, we referred to [28] and repeated this experiment in terms of our own number of measured samples (in this case 710 per frequency point) to obtain the correct critical value for use in the autocorrelation calculation at a 99% conficriterion was not used in dence interval; thus the this instance. For the line of sight number of independent samples we calculated via (6) (6) where:

number of position locations and number of independent transmitting antenna locations used in the . Before any standard deviations are issued, it is important to assess the statistics of the measurement throughout the different measurement locations used in the investigation. Any difference in the measured statistics at different on-body locations could result in different uncertainties in the measurements which is unacceptable. Fig. 19 depicts the calculated Rician K factors from different on-body locations. It shows that the measured statistics are

BOYES et al.: MEASUREMENT AND PERFORMANCE OF TEXTILE ANTENNA EFFICIENCY ON A HUMAN BODY IN A RC

Fig. 19. Measured Rician K factor at different on-body locations.

879

Fig. 21. Standard deviations in linear and dB formats for various on-body locations.

the different on-body locations tested. The uncertainty therefore is seen to be comparable irrespective of the on-body location chosen which means that any different on-body locations or any slight movements that may have occurred in the measurements do not affect the measurement repeatability and/or accuracy. VI. CONCLUSION

Fig. 20. Measured Rician K factor from different human beings.

comparable from one scenario to the next. Therefore, we would expect that the uncertainty levels will be comparable from one on-body location to the next if the exact same measurement parameters are employed. It is also important to check how the statistics deviate from one human being to another. Fig. 20 illustrates the effect of different human beings on the proportion of direct power in the chamber. Three male subjects were used with heights ranging from 1.74 to 1.8 m and weights ranging from 70. 5 to 81.3 kg, with each subject wearing different clothing. All measurements were made on the chest at a 0 mm proximity distance. Fig. 20 shows that the statistics are comparable irrespective of the human being, thus the same emulated scenario can be realized from one human to the next. Overall, when assessing Figs. 19 and 20, we see that any uncertainty contribution from line of sight coupling will be small. The figures prove that the fields inside the chamber are well stirred; thus the chamber is a Rayleigh environment (in terms of measured magnitudes) and this does not deviate with respect to the on-body measurement location or different human test subjects. This helps to realize a consistent platform from which to measure antenna parameters on human beings. The standard deviations are issued in Fig. 21, plotted in both linear and dB format as a function of different on-body measurement locations used in this investigation. The standard deviation inherent in the antenna measurements presented here is seen to be in the order of 0.22 dB for all of

In this paper we have shown that efficiency measurements performed using textile antennas on live human beings can be performed in an accurate and controlled manner, with the repeatability in a 0 mm chest location being as close as 2%. The magnitude of on-body losses experienced by a given textile antenna with a small ground plane is seen in this case to be a function of the material properties of that antenna—a lower resistivity, thinner (copper based) textile material was seen to perform worse when placed on body as compared to a larger resistivity, thicker material with the same overall design topology. This is believed to be due to the fact that the lower conductivity material based antenna has given rise to lower electric fields in the body as more power has been lost in the antenna itself for the same input power. This is a remarkable result from this study, quite unexpected by the authors, that the lower resistivity, thinner copper based textile was seen to perform worse as opposed to the higher resistivity, thicker textile material. This result can have a profound impact on the material choice for these small sized antennas in the sense that a higher conductivity material would appear not always to be the best option when operating in close proximity to a human being. The magnitude of efficiency losses on body has been experimentally shown to be mitigated somewhat by a variation in the distance from the body—a small 20 mm distance from the body (for antenna SHSL) in this case was sufficient to show that a reduction in radiation efficiency can be eliminated by up to 22%. For the single band (SHSL) antenna, at higher frequencies the 20 mm (off) body result approached the radiation efficiency value in free space. This result could prove useful when a location/on body distance is to be chosen for a given antenna and if the accession to multiple (higher) frequencies bands are envisaged. From the experiment that looked to assess the effect of loading the antenna via bending, one can conclude that this aspect is not good practice—the antenna was severely loaded for a considerable period of time (more so than for example

880

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 61, NO. 2, FEBRUARY 2013

if placed on a body part that relaxed the antenna from time to time), and thus the detuning performance was seen to be severe in this scenario. From the results seen here this condition cannot be recommended. With regards from the design aspect of on-body antennas, we feel it would be good practice to design and optimize with a given distance and body location in mind, preferably with the use of a body model if one is available. The use of multiband antennas is envisaged for future measurement work to assess the performance of larger sized textile devices. ACKNOWLEDGMENT The authors would like to thank M. Raja and M. Zurita for useful discussions and for their combined thoughts. REFERENCES [1] P. Salonen, L. Sydanheimo, M. Keskilammi, and M. Kivikoski, “A small planar inverted F antenna for wearable applications,” in Proc. 3rd Int. Symp. Wearable Comput. Digest, 1999, pp. 95–100. [2] P. Salonen, L. Sydanheimo, M. Keskilammi, and J. Rantanen, “A novel bluetooth antenna on flexible substrate for smart clothing,” in Proc. IEEE Int. Conf. Syst. Man Cybern., 2001, vol. 2, pp. 789–794. [3] P. J. Soh, G. A. E. Vandenbosch, S. L. Ooi, and N. M. A. Rais, “Design of a broadband all textile slotted PIFA,” IEEE Trans. Antennas Propag., vol. 60, no. 1, pp. 379–384, Jan. 2012. [4] Z. H. Hu, Y. I. Nechayev, P. S. Hall, and C. C. Constantinou, “Measurement and statistical analysis of on-body channel fading at 2.45 GHz,” IEEE Antenna Wireless Propag. Lett., vol. 55, no. 6, pp. 612–615, Jun. 2007. [5] D. Smith et al., “First and second order statistical characterizations of the dynamic body area propagation channel of various bandwidths,” Annal. Telecommun., pp. 1–7, Dec. 2010. [6] K. Minseok and J. I. Takada, “Statistical model for 4.5 GHz narrowband on-body propagation channel with specific actions,” IEEE Antennas Wireless Propag. Lett., vol. 8, pp. 1250–1254, Dec. 2009. [7] P. S. Kildal and K. Rosengren, “Correlation & capacity of MIMO Syst. & mutual coupling, radiation efficiency & diversity gain of their antennas: Simulations & measurements in a reverberation chamber,” IEEE Commun. Mag., pp. 104–112, Dec. 2004. [8] D. A. Hill, “Plane wave integral representation for fields in a reverberation chamber,” IEEE Trans. Electromagn. Compat., vol. 40, no. 3, pp. 209–217, Aug. 1998. [9] D. A. Hill et al., “Aperture excitation of electrically large lossy cavities,” IEEE Trans. Electromagn. Compat., vol. 36, no. 3, pp. 169–179, Aug. 1994. [10] P. S. Kildal and C. Carlsson, “Detection of a polarization imbalance in reverberation chambers and how to remove it by polarization stirring when measuring antenna efficiencies,” Microw. Opt. Technol. Lett., vol. 34, pp. 145–149, Jul. 2002. [11] L. R. Arnaut and G. Gradoni, “On distribution of fields and power in undermoded mode stirred reverberation chambers,” in Proc. General Assembly Sci. Symp., Oct. 2011, pp. 1–4. [12] L. R. Arnaut, “Effect of size, orientation and eccentricity of mode stirrers on their performance in reverberation chambers,” IEEE Trans. Electromagn. Compat., vol. 48, pp. 600–602, Aug. 2006. [13] D. A. Hill, “Boundary fields in reverberation chambers,” IEEE Trans. Electromagn. Compat., vol. 47, no. 2, pp. 281–290, 2005. [14] J. G. Kostas and B. Boverie, “Statistical model for a mode stirred chamber,” IEEE Trans. Electromagn. Compat., vol. 33, no. 4, pp. 366–370, Nov. 1991. [15] C. L. Holloway et al., “On the use of reverberation chambers to simulate a Rician radio environment for the testing of wireless devices,” IEEE Trans. Antennas Propag., vol. 54, no. 11, pp. 3167–3177, Nov. 2006. [16] G. A. Conway, W. G. Scanlon, C. Orlenius, and C. Walker, “In situ measurement of UHF wearable antenna radiation efficiency using a reverberation chamber,” IEEE Antennas Wireless Propag. Lett., vol. 7, pp. 271–274, 2008. [17] G. le Fur, C. Lemoine, P. Besnier, and A. Sharaiha, “Performances of UWB wheeler cap and reverberation chamber to carry out efficiency measurements of narrowband antennas,” IEEE Antennas Wireless Propag. Lett., vol. 8, pp. 332–335, 2009.

[18] J. Lilya and P. Salonen, “On the modeling of conductive textile materials for softwear antennas,” in Proc. IEEE Antennas Propag. Society Int. Symp. (APSURSI 2009), Jun. 1–5, 2009, pp. 1–4. [19] [Online]. Available: www.lessemf.com/fabric.html [20] [Online]. Available: http://www.cst.com [21] W. G. Scanlon and N. G. Evans, “Numerical analysis of bodyworn antennas,” Electron. Commun. Eng. J., pp. 53–64, Apr. 2001. [22] K. Rosengren and P. S. Kildal, “Study of distributions of modes and plane waves in reverberation chambers for characterization of antennas in multipath environments,” Microw. Opt. Technol. Lett., vol. 30, pp. 386–391, Sep. 2001. [23] P. S. Hall et al., “Antennas and propagation for on-body communication Syst,” IEEE Antennas Propag. Mag., vol. 49, no. 3, pp. 41–58, Jun. 2007. [24] P. S. Hall and Y. Hao, Antennas & Propagation for Body Centric Wireless Communications. London, U.K.: Artech House, 2006, ch. Ch.2, p. 14. [25] P. S. Kildal, S. H. Lai, and X. Chen, “Direct coupling as a residual error contribution during OTA measurements of wireless devices in a reverberation chamber,” presented at the IEEE AP-S, Charlestown, WV, Jun. 2009. [26] P. S. Kildal, X. Chen, C. Orlenius, M. Franzen, and C. L. Patane, “Characterization of reverberation chambers for OTA measurements of wireless devices: Physical formulations of channel matrix and new uncertainty formula,” IEEE Trans. Antennas Propag., vol. 60, no. 8, pp. 3875–3891, Aug. 2012. [27] BS EN 61000-4-21:2011 Electromagnetic Compatibility (EMC). Testing and Measurement Techniques. Reverberation Chamber Test Methods 2011. [28] H. G. Krauthauser, T. Winzerling, and J. Nitsch, “Statistical interpretation of autocorrelation coefficients for fields in mode stirred chambers,” in Proc. IEEE Int. Symp., Aug. 2005, pp. 550–555.

Stephen J. Boyes (S’08) entered academia full time in 2005 after a 10 year period of working full time in industry. He received the B.Eng. degree (Hons) in electronics and communications and the M.Sc. degree with distinction in microelectronic systems and telecommunications from the University of Liverpool, Liverpool, U.K., in 2008 and 2009, respectively. Since 2009, he has been working toward the Ph.D. degree at the University of Liverpool, U.K. He is a Dual Qualified Engineer, also holding a skilled level status in Mechanical/Manufacturing Engineering. His research interests include antenna measurements, electromagnetics, reverberation chambers, and on-body communications.

Ping Jack Soh (S’09) received the Bachelor and Master degrees in electrical engineering from Universiti Teknologi Malaysia (UTM), Johor, Malaysia, in 2002 and 2005, respectively. From 2002 to 2004, he was a Test Engineer working on new products’ test definition for manufacturing purposes, both hardware and software. Then in 2005, he joined Motorola Technology Malaysia as a Research and Development (R&D) Engineer for Electrical Design. There, he worked on the hardware development of two-way radios, focusing on design, characterization and testing of new radios’ antennas and RF front-ends. In 2006, he joined the School of Computer and Communication Engineering, Universiti Malaysia Perlis (UniMAP) as a Lecturer, before being promoted to Senior Lecturer in 2011. He is currently on study leave and working towards his Doctoral degree in the Telecommunication and Microwaves Research Division, Department of Electrical Engineering (ESAT-TELEMIC), Katholieke Universiteit Leuven, Belgium. His research interest includes the design, development and modeling of flexible, textile and planar antennas, on-body communications, metamaterials, passive microwave components, and microwave measurements.

BOYES et al.: MEASUREMENT AND PERFORMANCE OF TEXTILE ANTENNA EFFICIENCY ON A HUMAN BODY IN A RC

Yi Huang (S’91–M’96–SM’06) received B.Sc. degree in physics from Wuhan University, Wuhan, China, the M.Sc. degree in microwave engineering from the Nanjing Research Institute of Electronics Technology (NRIET), Nanjing, China, and the D.Phil. degree in communications from the University of Oxford, London, U.K., in 1994. He has been conducting research in the areas of wireless communications, applied electromagnetics, radar and antennas for the past 25 years. His experience includes three years spent with NRIET (China) as a Radar Engineer and various periods with the Universities of Birmingham, Oxford, and Essex at the U.K. as a Member of research staff. He worked as a Research Fellow at British Telecom Labs in 1994, and then joined the Department of Electrical Engineering & Electronics, the University of Liverpool, UK as a Faculty in 1995, where he is now a full Professor in Wireless Engineering, the Head of High Frequency Engineering Research Group and M.Sc. Programme Director. He has published over 200 refereed papers in leading international journals and conference proceedings, and is the principal author of the popular book Antennas: from Theory to Practice (John Wiley, 2008). He has received many research grants from research councils, government agencies, charity, EU and industry, acted as a consultant to various companies, and served on a number of national and international technical committees. He has been an Editor, Associate Editor, or Guest Editor of four international journals. Dr. Huang has been a keynote/invited speaker and organizer of many conferences and workshops (e.g., IEEE iWAT 2010, WiCom 2006, 2010 and LAPC2012). He is at present the Editor-in-Chief of Wireless Engineering and Technology, a U.K. National Rep. of European COST-IC1102, Executive Committee Member of the IET Electromagnetics PN, and a Fellow of IET.

Guy A. E. Vandenbosch (M’92–SM’08–F’13) received the M.S. and Ph.D. degrees in electrical engineering from the Katholieke Universiteit Leuven, Leuven, Belgium, in 1985 and 1991, respectively. From 1991 to 1993, he held a Postdoctoral Research position at the Katholieke Universiteit Leuven. Since 1993, he has been a Lecturer, and since 2005, a Full Professor at the same university. He has taught, or teaches, courses on electromagnetic Waves, antennas, electromagnetic compatibility, electrical engineering, electronics, and electrical

881

energy, and digital steering and measuring techniques in physics. His research interests are in the area of electromagnetic theory, computational electromagnetics, planar antennas and circuits, nanoelectromagnetics, EM radiation, EMC, and bioelectromagnetics. His work has been published in 150 papers in international journals and has been presented in 250 papers at international conferences. Dr. Vandenbosch has been a Member of the “Management Committees” of the consecutive European COST actions on antennas since 1993. Within the ACE Network of Excellence of the EU (2004–2007), he was a Member of the Executive Board and coordinated the activity on the creation of a European antenna software platform. At present, he leads the EuRAAP Working Group on Software and represents this group within the EuRAAP Delegate Assembly. He is the holder of a certificate of the postacademic course in Electro-Magnetic Compatibility at the Technical University Eindhoven, The Netherlands. From 2001 to 2007, he was the President of SITEL, the Belgian Society of Engineers in Telecommunication and Electronics. Since 2008, he is a Member of the board of FITCE Belgium, the Belgian branch of the Federation of Telecommunications Engineers of the European Union. In the period 1999–2004, he was ViceChairman, and in the period 2005–2009 Secretary of the IEEE Benelux Chapter on Antennas en Propagation. Currently he holds the position of Chairman of this Chapter. From 2002–2004, he was Secretary of the IEEE Benelux Chapter on EMC. He currently is Secretary of the Belgian National Committee for Radioelectricity (URSI), where he is also in charge of Commission E.

Neda Khiabani (S’12) received the B.Sc. and M.Sc. degrees in electrical engineering from K. N. Toosi University of Technology (KNTU), Tehran, Iran, in 2004 and 2006, respectively. She is currently working towards the Ph.D. degree in electrical engineering at the University of Liverpool, Liverpool, U.K. From 2004 to 2009, she was with Tele2Iran (Taliya) in Tehran as a Radio Planning and Optimization Engineer. Her current research interests include THz antenna design and analysis