The Environmentalist, 25, 215 221, 2005 C 2005 Springer Science + Business Media, Inc. Manufactured in The Netherlands. Interaction Between GSM Handset Helical Antenna and User s Head: Theoretical Analysis and Experimental Results A. ALEXIOU, P. KOSTARAKIS * and V.N. CHRISTOFILAKIS University of Ioannina, Physics Department, Ioannina, Greece T. ZERVOS National Centre for Scientific Research Demokritos, Institute of Informatics and Telecommunications, Athens, Greece; University of Patras, Department of Electrical & Computer Engineering, Patras, Greece A.A. ALEXANDRIDIS and K. DANGAKIS National Centre for Scientific Research Demokritos, Institute of Informatics and Telecommunications, Athens, Greece C. SORAS University of Patras, Department of Electrical & Computer Engineering, Patras, Greece V.V. PETROVIĆ, B.M. KOLUNDŽIJA, A.R. DORDEVIĆ University of Belgrade, Faculty of Electrical Engineering, Belgrade, Serbia Summary. The performance of a cellular phone commercial helical antenna at 900 MHz band, both in free space and in the presence of a human head phantom was studied. Numerical simulation of the phone model for the latter case has been performed giving 3D radiation diagrams. The effect of the phantom head on radiation diagrams is presented. The relative amount of the EM power absorbed in the head was obtained for several distances of the phone. Measurements were carried out in a RF anechoic chamber using standard horn antennas and a calibrated measuring system. Absolute radiation patterns of the antenna gain were obtained in the three principal planes. Significant reduction of the absorbed power could be achieved just by moving the phone 1 cm away from the head. Keywords: helical antenna, GSM handset, phantom head, interaction, anechoic chamber, WIPL-D, radiation pattern, gain patterns, absorbed power Introduction As a result of the quick growth of mobile communications system many efforts have been made towards the improvement of the characteristics of cell phone antennas. Handheld antennas had to follow the dramatic decrease in size and weight of cell phone, while main- * E-mail: pkost@cc.uoi.gr E-mail: aalex@iit.demokritos.gr E-mail: soras@ee.upatras.gr E-mail: vpetrovic@telekom.etf.bg.ac.yu taining the same antenna performance in terms of radiation pattern, gain and bandwidth. These changes lead to a rapid evolution of antenna structures for portable phones (Elsherbeni et al., 2001). Also, as the usage of mobile communication systems keeps expanding, a large number of studies related to the interaction between antennas and human body are taking place. Taking into account that mobile terminals operate in close proximity with the human head, the studies examine the interaction of electromagnetic fields with the head from the two following points of view. The first one is related to the
216 Alexiou et al. degradation of the antenna performance due to the operator s influence that affects radiation pattern, input impedance and radiation efficiency. The second aspect is referred to the absorption of the microwave energy from the user s head, in order to establish precise safety standards. Several types of antennas in mobile phones operating in vicinity with the human body have been studied in other works. The majority of them used monopole antennas (Okoniewski and Stuchly, 1996; Watanabe et al., 1996; Zervos et al., 2004a, 2004b). Other studies used dipoles (Watanabe et al., 1996), helical (Mangoud et al., 2000) or patch antennas (Jensen and Rahmat-Samii, 1995a) at several frequencies. Dual band monopole-helix antennas near the human head have been studied (Bernardi et al., 2001). Also, such mobile handset antennas have been analyzed and designed with the presence of the user s head model (phantom) (Jensen and Rahmat-Samii, 1995b; Lazzi et al., 1998; Zervos et al., 2004a). The aim of our work is to study the interaction between a helical antenna and a homogeneous spherical head model (phantom). For the theoretical analysis, we have accurately modeled the radiation of a normal mode helical antenna attached on a mobile handset in vicinity with a human head. For the electromagnetic simulation a commercial package, called WIPL- D, was used. This software is based on the Method of Moments (Galerkin method) which is a very suitable method for the open electromagnetic (EM) structures analyzed in this frequency domain. For the experimental analysis, a head phantom filled with a liquid of certain electrical permittivity and conductivity was used to emulate the human head exposure in the field of a commercial helical antenna at 914 MHz. The far field radiation patterns of the handset in free space were measured in the three principal planes, xy, xz and yz, relative to a coordination system attached to the handset and were compared with the radiation patterns measured when the phantom head is near the handset. Models of the head and the handset In order to analyze theoretically the EM radiation of a mobile handset in the presence of a human head, an accurate geometrical model of a handset and a phantom head was made using a dedicated software package (WIPL-D). The model of the handset consists of rectangular plates. Its dimensions are: 125 40 20 mm. A helical antenna similar to those commercially used for 900 MHz GSM phones has been modeled. The model was made using a wire of 0.4 mm radius in the shape of helix with 8 turns. The radius of helical antenna was 1.6 mm and its height was 10.5 mm. At 6.5 mm distance above the top of the helix there was an additional 1 turn with 3.4 mm radius made of similar wire. The helical antenna was covered with a dielectric insulator of ε r = 3, in the shape of truncated cone, with bottom radius 5.6 mm, top radius 5 mm and height 23.3 mm (Fig. 1). A point-like (delta-function) generator is assumed to be attached at the plate-to-wire junction and all conductors are assumed to be perfect. Figure 1. Phantom and handset model with helical antenna.
Interaction Between GSM Handset Helical Antenna and User s Head 217 A homogeneous brain tissue spherical model was used, with 90 mm radius which was cut by two vertical planes (Fig. 1). The dielectric properties of the head model at 900 MHz are: ε r = 41.5,σ = 0.97 S/m (CENELEC ES 59005, 1998). The head has been geometrically modeled with 320 volume entities called bilinear hexahedra. The handset was modeled with its distance from the phantom head ranging between 0 and 40 mm. Its vertical position was set in an ear-to-speaker manner, while it was placed at a symmetrical position across y-axis. Theoretical analysis The basic numerical task of EM simulation software (WIPL-D) is to calculate the distribution of electric current density over the handset s plates and wires and over the hexahedra surfaces of the head. Surface currents distribution is approximated by polynomial functions that automatically satisfy continuity equation across the junctions. Current distribution is obtained by solving the electric field integral equation (EFIE) by the Method of Moments (Galerkin method). The relative absorbed power, P abs %, can be calculated using the following formula: P abs % = P g P rad P g 100 (1) where P g is the input power and is equal to P g = YU 2, where Y is the driving point admittance and U is the RMS value of the generator voltage. The term P rad is the radiated power in the 3D space and is calculated over the pre-specified dense 3D mesh. Modeling results From the numerical simulations that have been performed, the three dimensional gain radiation patterns of the handset with and without the phantom head can be drawn. They can be used as a graphical illustration of the EM power absorbed in and reflected by the head. These diagrams for a distance of 10 mm between the head and the handset are shown in Fig. 2. A significant change in the shape of radiation diagrams due to the presence of the head can be noticed. More precisely, in the direction towards the head (positive x-axis), radiation is reduced due to the power absorbed by the head, but we can notice that in the top side of the head near the antenna (positive z-axis) there is a relatively big lobe, possibly due to partial reflection from the head. In the opposite direction (negative x-axis), the radiated power looks similar to the one without the head. In order to have a quantitative approach to the absorbed power, the relative absorbed power in the head was calculated using equation (1). We have also made a model of a monopole antenna 83.3 mm high (λ/4 at 900 MHz), and we compared them according to the absorbed power from the head. In Fig. 3 the relative absorbed power is presented for both anten- Figure 2. Theoretical gain patterns (in dbi) for the handset alone (a) and for the handset with phantom head (b).
218 Alexiou et al. Real model Figure 3. Percentage of absorbed power for a monopole and a helical antenna versus distance from the head to the handset. nas as a function of the distance between the handset and the phantom head. For both types of antennas, it can be seen that the absorbed power is being reduced relatively fast by moving the handset away from the head. This is an expected result as it has also been derived for a monopole antenna in our previous work (Elsherbeni et al., 2001) but for different frequency (1800 MHz). Focusing on the comparison of the performance of the two antennas, it can be noticed that the amount of absorbed power is generally larger for the case of the handset with the helical antenna than for the case with the monopole one. More specifically, the difference becomes larger for small head-phone distances (with maximum value of the difference in the relative absorbed power of about 23% for 0 mm distance). As the distance increases, the difference between the relative absorbed power by the head for both antennas tends to be negligible (in 40 mm distance), where the value of the relative absorbed power falls at 19%. A real model of a commercial helical antenna has been used and was mounted on the top of a commercial mobile telephone model (Fig. 4). A sphere with a radius of R = 90 mm made of glass which is cut by two vertical planes at 5 R/6 = 150 mm, has been used as a shell in order to model the human head. The head phantom filled with a liquid of certain permittivity and conductivity according to the related standards (CENELEC ES 59005, 1998) was used to emulate the human head at 900 MHz. The shape and dimensions of the head phantom are shown in Fig. 4. The position of the handset was set to an ear-to-speaker manner, and it was placed 10 mm away from the phantom head. Measurements procedure The measurements setup is shown in Fig. 5. All measurements were taken in an RF shielded anechoic chamber 10 m long, 5mwide and 5mhigh, fully lined with 0.9 m height pyramids of absorbing foam. This is a commonly used setup for measurements of antenna radiation patterns. The handset under test is mounted on a tripod made by a RF transparent material standing on a turntable. The antenna of the handset is connected to the Port 1 of a Network Analyzer set to 914 MHz central frequency. At the other side of the chamber, at distance 4575 mm, a reference horn antenna is mounted and connected to the Port 2 of the Network Analyzer. The turntable is connected through an optical fiber with the positioning controller. All the devices are controlled by a PC with the appropriate software. Device Figure 4. Helical antenna model and head phantom used in measurements.
Interaction Between GSM Handset Helical Antenna and User s Head 219 Figure 5. Measurements setup. interconnections are obtained through a standard IEEE Bus. The turntable makes a scan from 0 to 360 degrees in step of 5 degrees clockwise. For every step we have the following formula for the calculation of the gain from the measured S-parameters: S 21 2 ( ) λ 2 (1 S 22 2 )(1 S 11 2 ) = G T G R, (2) 4πd where G T is the power gain of the handset, G R is the power gain of the reference antenna and d is the distance between the reference antenna and the handset. Using the above measuring procedure, we measure the gain of the handset alone (G HS ) and the gain of the handset with the presence of the phantom (G PH ). At each position, the horizontal and vertical components of the field (gain) are measured by the dual polarized horn antenna. Two series of measurements are taken, one with the reference antenna horizontally polarized and the other with the antenna vertically polarized. The sum of these two gain components at each position gives the corresponding total gain. Measurements results In this section the influence of the head phantom on the radiation characteristics of the helical antenna handset is studied. Far field radiation patterns for the examined antenna under the presence of the phantom have been calculated for the xy plane (azimuth pattern), xz plane (elevation pattern for ϕ = 0degrees) and yz plane (elevation pattern for ϕ = 90 degrees) at 914 MHz and compared with these obtained in free-space (handset alone). In Fig. 6 the total gain diagrams with and without the phantom are shown in combination with the simulation results. The changes on the shape of radiation patterns are related to the power absorption by the human head. Although the performed measurements were made in only the three principal planes, the set of measurement points can be considered as a very coarse nonuniform 3D mesh. From this mesh it is possible to estimate the total EM power absorbed by the head in the following way. Assuming that P g is the input power of the handset antenna and P abs is the absorbed power in the head, the radiated power is calculated from the total gain by the
220 Alexiou et al. Figure 6. yz-plane. Simulated and experimental total gain patterns (in db) with and without a phantom head: (a) xy-plane, (b) xz-plane, and (c) classical formula: P rad = P g 4π 4π G d = P g G average (3) The average gain in the last equation can be approximated by the average value of the N measured gain values: G G average (4) N The radiated power for the handset and the phantom can be expressed as: P PH = P g P abs (5) The relative absorbed power, η, for the three planes can be calculated using equations. (3) (5), as: η = P abs P g = 1 GPH N (6)
Interaction Between GSM Handset Helical Antenna and User s Head 221 Using the measurements results in the above formula we derived that η = 0.635. Hence it is estimated that 63.5% of the power delivered to the helical antenna handset is being absorbed by the head. The influence of the presence of the user s head on the radiation diagram of the handset is obvious by comparing the diagrams with and without the presence of the head (Fig. 6). From the measured diagrams, an averaged reduction in gain of 4.1 db can be noticed, in the azimuth plane diagram (Fig. 6(a)) while the maximum reduction is of order of 9 db. In the same way, in certain directions of the other two planes, the gain is reduced at maximum of about 6 db (Fig. 6(b) and (c)). On the contrary there are some directions where we observed an increase in gain up to 7.4 db (Fig. 6(b)) or 8 db (Fig. 6(c)), which was caused by the partial reflection of the EM energy by the user s head. Conclusions A normal mode helical antenna attached on a handset was accurately modeled and measured near a phantom head. The changes in radiation patterns, due to the presence of the head, were demonstrated. The helical antenna was compared theoretically with a monopole, and it was found that under the same conditions, the amount of the power absorbed by the head is larger in the case of a handset with a helical antenna. It was found that for both antennas, the relative absorbed EM power in the head reduces very fast by moving the handset away from the head. When the handset was in touch with the head, the absorbed power in the case of the helical antenna was calculated at 85% of the power delivered to the antenna, while for the monopole it was 62%. The experimental results showed that the antenna performance is affected by the presence of the human head. The absorption of the radiated power by the head phantom causes reduction in the average antenna gain. Using measurements, the absorbed power was estimated to be about 60% of the input power when the handset is 1 cm away from the head, while the corresponding theoretical result is about 55%. References Bernardi, P., Cavagnaro, M., Pisa, S. and Piuzzi, Em.: 2001, Power Absorption and Temperature Elevations Induced in the Human Head by a Dual-Band Monopole-Helix Antenna Phone, IEEE Trans. Microwave Theory Tech. 49, 2539 2546. CENELEC ES 59005.: 1998, Considerations for the Evaluation of Human Exposure to Electromagnetic Fields (EMFs) from Mobile Telecommunication Equipment (MTE) in the Frequency range 30 MHz 6 GHz. Elsherbeni, Z., Huang, C.-W.P. and Smith, C.E.: 2001, Handbook of Antennas in Wireless Communications, Chapter 12, Handheld Antennas. Jensen, M.A. and Rahmat-Samii, Y.: 1995a, EM Interaction of Handset Antennas and a Human in Personal Communications, Proceedings of the IEEE 83, 7 17. Jensen, M.A. and Rahmat Samii, Y.: 1995b, EM Interaction of Handset Antennas and a Human Head in Personal Communications, Proc. IEEE 83, 7 17. Lazzi, G., Pattnaik, S.S., Furse, C.M. and Gandhi, O.P.: 1998, Comparison of FDTD Computed and Measured Radiation Patterns of Commercial Mobile Telephones in Presence of the Human Head, IEEE Trans. on Antennas and Propagation, 46(6). Mangoud, M.A., Abd-Alhameed, R.A. and Excell, P.S.: 2000, Simulation of Human Interaction with Mobile Telephones Using Hybrid Techniques Over Coupled Domains, IEEE Trans. Microwave Theory Tech. 48, 2014 2021. Okoniewski, M. and Stuchly, M.A.: 1996, A Study of the Handset Antenna and Human Body Interaction, IEEE Trans. Microwave Theory Tech. 44, 1855 1864. Watanabe, S.-I., Taki, H., Nojima, T. and Fujiwara, O.: 1996, Characteristics of the SAR Distributions in a Head Exposed to Electromagnetic Fields Radiated by a Hand-Held Portable Radio, IEEE Trans. Microwave Theory Tech. 44, 1874 1883. Zervos, T., Alexandridis, A.A., Petrović, V.V., Dangakis, K., Kolundžija, B.M., Dordević, A.R. and Soras, C.: 2004a, Dependence of the EM Power Adsorbed in the Head of a Mobile Phone User on the Phone Head Distance, in Proc. of International Symposium on Electromag. Theory, URSI. Zervos, T., Alexandridis, A.A., Petrović, V.V., Dangakis, K., Kolundžija, B.M., Dordević, A.R. and Soras, C.: 2004b, Mobile Handset Radiation Efficiency as a Function of the Antenna Position Relative to the Human Head, in Proc. 8th WSEAS Int. Multiconf. CSCC.