DIFFERENT TYPES OF HUMAN HEAD SHAPES FOR CELLULAR PHONE EXPOSURE ON ELECTROMAGNETIC ABSORPTION Mohammad Rashed Iqbal Fauque* #1, Mohammad Taiqul Islam* 2, Nobahiah Misan* #3 * Institute of Space Science (ANGKASA), # Dept. of Electical, Electonic and Systems Engineeing, Faculty of Engineeing and Built Envionment, Univesiti Kebangsaan Malaysia, 43600 UKM, Bangi, Selango, Malaysia. 1 ashedgen@yahoo.com, 2 titaeq@yahoo.com, 3 bahiah@vlsi.eng.ukm.my ABSTRACT In this pape, the local specific absoption ate (SAR) induced in spheical, cubical, and ealistic human head models exposed to a mobile phone is investigated. The highest degee diffeent shapes but almost the same volume of thee human head models is consideed. Obtained local maximum SAR induced in these thee human head models fo homogeneous cases ae establish to have an aveage diffeences of about 9% at 1800 MHz espectively. A compaison analysis of SAR induced in the ealistic human head model fo homogeneous and inhomogeneous cases is also discussed. Fom the esults, it can be obseved that the local maximum SAR induced in the homogeneous human head model is lage than that induced in the inhomogeneous human head model. Index Tems cellula phone, the finite-diffeence timedomain (FDTD) method, head size, specific absoption ate (SAR). 1. INTRODUCTION With the apid and eve moe widespead use of mobile phones, public concen egading the possible health hazads has been gowing, which bings an inceased equiement on electomagnetic (EM) dosimety fo mobile phones. The basic paamete in the EM dosimety is defined in tems of the specific absoption ate (SAR), o the absobed powe in unit mass of tissue [1]. The SAR is geneally evaluated using eithe phantom measuement o compute simulation. The finite-diffeence time-domain (FDTD) method is cuently the most widely accepted means fo the SAR computations [2]. In Euope the basic limit of SAR set fo the geneal public is 2 W/kg aveaged ove a volume equivalent to 10 gm and a pefect of 6 min. [3]. The ANSI/IEEE standad [4] defines a sticte limit fo an uncontolled envionment of 1.6 W/kg aveaged ove a volume of 1 gm and a peiod of 30 min. To analyze the possible ange of vaiations of the induced field stengths in the vaious tissues equies an extensive effot, since the local field stengths stongly depend on a lage numbe of paametes, such as: opeational fequency and antenna input powe; position of the device with espect to the head; design of the device; the oute shape of the head; the distibution of the diffeent tissues within the head, and the electic popeties of these tissues. The electic paametes of a human body vay with levels of physical and metabolic activity, health, and age. The vaiations in all these popeties lead to a stetch in the analyzed absoption distibution. The cuently most often used system fo testing handheld cellula telecommunications equipment is the measuement setup using a homogeneous anatomically-shaped phantom filled with a liquid simulating bain tissue [4-8]. The ationale behind this appoach comes fom the enegy absoption mechanism in the close nea-field of antennas [9-11], which concludes that the most detemining paametes fo volumeaveaged values ae the time-aveaged antenna input powe, opeational fequency, design of the device, and its position with espect to the head, and to a much lesse extent, on the physical popeties of the head. In this pape the authos focus on the effects of the factos of head popeties, such as size, shape, and electical popeties of tissues in ode to calculate the SAR induced in human head models exposed to a cellula phone. To find the effects of head size o shape on electomagnetic absoption chaacteistics, the ealistic head model, spheical head model, and a cubical head model ae scaled and then SAR distibutions fo these models, when exposed to a cellula telephone model have been investigating using the FDTD method.
2. HEAD MODELING Accuate phantoms of homogeneous human heads can be geneated on the basis of magnetic esonance imaging (MRI). The tanslation of the thee-dimensional (3-D) data sets of elaxation times into the tissue distibution is a difficult task and geneally equies a peson tained eithe in medicine o biology, which is able to distinguish both tansitional and maginal egions. MRI poduced in diffeent laboatoies, by diffeent scientists and fom diffeent test subjects pedictably contains diffeing discetizations. In these studies diffeent shapes, diffeent dielectic constant, diffeent conductivity, and diffeent mass density of homogeneous phantoms in diffeent tissues based on MRI data sets of thee diffeent adults ae used, also simulated to inhomogeneous phantoms and thei esults compaed. Fig. 1 shows the oute shapes and two coss-sectional views of these head phantoms and Table I gives the data of discetization. The ea closest to the antenna needs special attention, duing nomal use of a hand-held telephone; the ea is pessed against the head and, theefoe, changes its shape. In ode to avoid effects caused by the diffeent ea modeling, which may mask the effects of the head itself, the oute ight ea of the head phantoms was emoved. The fist phantom in ealistic head model was taken with the highest esolution. Its voxel size is 1 mm in all thee Catesian dimensions. Realistic head model has the lagest volume. The bain egion was segmented vey caefully. Fo the entie head 13 tissue types wee simulated. Howeve, the lowe pat of the head was assigned to only one tissue type. The second head phantom is consideed as cubical, which has nealy the same voxel size. It was developed fo the taining of medical students and distinguishes among 120 tissue types. Fo the EM analysis this lage numbe needed to be educed to 13 diffeent tissues fo which electic paametes ae available [5]. The thid head phantom is consideed as spheical that has taken with a voxel size of about 12 mm 3. The discetization is elatively cude. In the oiginal MRI model the skin was not identified. In the compute model the skin was added as an oute laye with a thickness of one voxel. The bain egion of head phantom Spheical is homogeneous and assigned only one tissue type compaing the data of diffeent publications eveals a spead in the values given fo the electic paametes of diffeent types of tissue. Table-2 and Table-3 show the electic paametes we have chosen fo this study. The pemittivity and the conductivity of the tissues of phantoms wee taken fom the dielectic database [7]. Table 1 Thee diffeent head models fo simulations Realistic head model [11] Spheical head model [5] Fig. 1 Test view of thee MRI phantoms. a).realistic head model b) Spheical head model c) Cubical head model 3. PHONE MODELING Cubical head model [7] Head 4.44 dm 3 3.35 dm 3 4.26 dm 3 Volume Tissues 13 120 12 Computational 175 230 159 208 159 206 Space 226 201 249 Voxel Size ( 1 mm ) 3 ( 1.075 mm) 3 (1.875 mm) 2 3 mm a) b) c) The numeical phone model opeating at 1800 MHz consists of a conducting box, a plastic casing, and a helix antenna, as shown in Fig.2. The antenna was fed using the coaxial feeding method. The electic field components of the same amplitude, tangential to the top suface of the phone, wee applied at the souce point, as shown in Fig.2 (c). Fo both of the head models, the phone was located on the efeence plane, which is defined by auditoy canal openings of both eas and the cente of the mouth. Regading the diection of the phone, the touch position that is a nomal opeating position was used fo the anatomical models, and fo the simple models, phone was diected paallel to the sagittal plane of the head, as shown in Fig.2.
mass density ρ of the thee homogeneous phantom-muscle ε = 53.5, σ = 1.34 S/m, ρ = 1.04 g/cm 3 ) head models ( at 1800 MHz ae obtained fom the liteatue [3-7], espectively. The cellula phone and human-head models ae assumed to be of nonmagnetic mateial ( µ = 1.0). 4. RESULTS Fig. 2 Phone model (unit: mm). (a) Font view. (b) Top view. (c) Feeding pat in FDTD mesh The cellula-phone model is constucted with a quatewavelength helix antenna mounted on the top of a ectangula box with dimensions height h = 108 mm, width w = 46 mm, and thickness d = 23 mm, which is filled with equivalent mateial, as shown in Fig. 2. The mateial on the outside suface of the handset box is adopted by a dielectic mateial ( ε = 3.78, σ = 1.0 10 5 S/m) at 1800 MHz, only. Accoding to a study in ef. [10-11], the dielectic mateial used fo the outside suface of a handset box has significant impact on the eduction of the SAR induced in a human head. The tansmitted powe of the cellula phone is assumed to be 0.125 W at 1800 MHz, espectively. Following the equation in ef. [7-9], V = P 8 R, whee the excitation souce voltage V at the feeding point of the monopole antenna can easily be calculated. Hee R and P ae the esistance and tansmitted powe of the cellula phone in fee space, espectively. Using the method of moments (MoM) [4-5], the impedance of the monopole antenna in ai obtained is 50.81 + j14.85 Ω at 1800 MHz, only. The calculations of SAR distibution induced in the human head ae made with an initial sinusoidal time-vaying electic.field Ez = (V/δ) sin(ωt) located at the gap between the helix antenna and the uppe cente of the box case, as shown in Fig. 2, whee δ = 2.0 mm is the cell size used in the FDTD simulation. A metallic mateial with ε =1.0 and σ = 3.72 10 7 S/m is employed to simulate the helix antenna. Thee homogeneous human-head models (Fig. 1) including spheical, cubical, and ealistic shapes, ae discetized into 656419, 656253, and 656132 cubic cells of 2.0 mm on each side in the FDTD simulations, espectively. The elative dielectic constant ε conductivity σ and The esults of maximum local SARs induced in thee homogeneous human-head models vesus the distance between the cellula phone and the human-head model ae shown in Fig. 3. Fom Fig. 3, it is clea that the spheical head model has the maximum value of the local maximum SAR, while the ealistic human head model has the minimum value of the local maximum SAR. The esults of the local maximum SAR induced in the cubical and spheical head models ae close and they ae between those obtained by the ealistic head models. The aveage diffeence of the local maximum SAR induced in these thee head models is appoximately within 9% at 1800 MHz, espectively. This obsevation emphasizes that the shape of the human head plays a mino ole in calculating the SAR induced in the human-head models. These yield that the conclusion made in ef [5-7] that the maximum local SAR is scacely affected by the shape of the human head exposed to a cellula phone. In the following study, the ealistic humanhead model is simulated six times as a homogeneous model with six electical popety sets, such as the bone tissues, the skin tissues, the blood tissues, the eye tissues, the muscle, and bain tissues, and one time as an inhomogeneous model with six tissues taken into account to calculate the SAR esults at 1800 MHz only. Local Maximum SAR [W/kg] 0.0 6 0.0 5 0.0 4 0.0 3 0.0 2 0.0 1 0 1800 MHz 5 10 15 2 0 Dis tanc e be twe e n the pho ne mo de l and the mo de l (mm) C ub ic a l S p e ic a l R e a lis t ic Fig. 3 SAR induced in fou homogeneous phantom-muscle head models with elative dielectic constant ε = 53.5, conductivity σ = 1.34 S/m, and mass density ρ = 1.04 g/cm3 at 1800 MHz
The electical popeties wee taken of muscle [8] is ε = 53.5, σ = 1.34 S/m, ρ = 1.04 g/cm 3 at 1800 MHz. A compaison of the data found in diffeent publications eveals a spead in the values given fo the electical popeties of diffeent types of tissue [3-7]. The electical popeties of the inhomogeneous human-head model with six tissues at 1800 MHz is shown in Table 2. Table 2 Dielectic tissue popeties at 1800MHz Tissue Type Density, ρ (1000 Kgm 3 ) Conductivity, σ ) (S-m 1 Dielectic Constant Ai 0.0012 0.0 1.0 Bone 1.85 0.59 19.3 Skin 1.10 1.21 41.4 Blood 1.06 2.04 59.37 Eye 1.01 2.03 68.6 Bain 1.03 1.15 43.5 Muscle 1.04 1.34 53.5 Compaisons of maximum local SAR induced in the ealistic human-head model fo homogeneous and inhomogeneous case at 1800 MHz is shown in Fig 4. It is found that local maximum SAR induced in homogeneous models has lage values than those induced in inhomogeneous models. It should be noted that the electical popety of a human head significantly affects the esult of the SAR induced in homogeneous o inhomogeneous head models. Today, coect calculation o measuement of SAR distibution in a human head while using a cellula phone has become an impotant issue. Local Maximum SAR (W/kg) 0.0 6 0.0 5 0.0 4 0.0 3 0.0 2 0.0 1 0 1800 MHz 5 10 15 2 0 Distance between the phone model and the head model (mm) Fig. 4 Compaisons of maximum local SAR induced in the eal human head model fo homogeneous and inhomogeneous cases at 1800 MHz ε m o de l (B o ne ) m o de l (S kin) m o de l(b lo o d) m o de l (Eye ) m o de l (M us c le ) m o de l (B a in) Inho m o g e ne o us m o de l 5. CONCLUSIONS In this pape, the esults pove that the spatial peak SAR is baely affected by the size, diffeent dielectic popeties, and the shape of the human head fo electomagnetic souces at a defined distance fom the human head. Compaed to othe factos, such as distance of the souce fom the head and design of the devices, the effects caused by the complex anatomy ae mino especially in the case of volumeaveaged values. The compaison of the esults obtained fom the inhomogeneous and homogeneous phantoms suggests that homogeneous phantoms ae highly suited to be used in compliance tests fo handheld cellula telecommunications equipment opeating in the 1800 MHz. The esults of ou study can also be used to investigate diffeences of biological effects between human species and ages. ACKNOWLEDGEMENT The authos would like to thank Institute of Space Science (ANGKASA), Univesiti Kebangsaan Malaysia (UKM) and the MOSTI Secetaiat, Ministy of Science, Technology and Innovation of Malaysia, e- Science fund: 01-01-02- SF0612, fo sponsoing this wok. REFERENCES [1] Repot of Telecommunications Technology Council fo the ministy of Posts and Telecommunications, Delibeation no. 89, Radio-Radiation Potection Guidelines fo Human Exposue to Electomagnetic Fields, Tokyo, 1997. [2] K. S. Kunz and R. J. Luebbes, The finite diffeence time domain method fo electomagnetic, Boca Raton, FL, CRC, 1993. [3] IEEE C95.1-2005, IEEE standads fo safety levels with espect to human exposue to adio fequency electomagnetic fields, 3 khz to 300 GHz," Institute of Electical and Electonics Enginees, New Yok, NY, 2005. [4] M. T. Islam, M. R. I. Fauque, and N. Misan, Design analysis of feite sheet attachment fo SAR eduction in human head, Pogess In Electomagnetics Reseach, PIER 98, 191-205, 2009. [5] K. H. Hohne, M. Bomans, M. Reime, R. Schubet, U. Tiede, and W. Liesc, A volume-based anatomical atlas, IEEE Compute Gaphics Applicati., pp. 72-77, 1992 [6] A. Hiata, K. Shiai, and O. Fujiwaa, On aveaging mass of SAR coelating with tempeatue elevation due to a dipole antenna Pogess In Electomagnetics Reseach, PIER 84, 221 237, 2008. [7] 0. P. Gandhi..I. Y. Chen, and D. Wu, Electomagnetic absoption in the human head fo mobile telephones at
835 MHz and 1900 MHz, Int. Symp Elecomag. Compat., Roma, 1994, pp. 1-5. [8] M. R. I. Fauque, M. T. Islam, and N. Misan, Evaluation of specific absoption ate (SAR) eduction fo PIFA antenna using metamateials, Fequenz Jounal, vol. 7/8, pp. 144-149, 2010. [9] Kiminami, K., A. Hiata, Y. Hoii, and T. Shiozawa, A study on human body modeling fo the mobile teminal antenna design at 400 MHz band, J. of Electomagnetic Waves and Appl, Vol. 19, 671 687, 2005. [10] Ae-kyoung, and Jeong-ki Pack, Effect of head size fo cellula telephone exposue on EM absoption, IEICE Tans. Commun., vol. E85-B, no. 3, 2002. [11] G. Bielke and S. Meindl, Deidimensionale segmentiete MR-Bilddatenatze. Tech. Rep. N. 6564/33038, Deutsche Klinik fu Diagnostik e. V., Foschungsvetdg DBP Telekom, 1993.