Human Body Effects on the Matching of a 2.45 GHz Coplanar-fed Antenna

Transcription

1 1 Human Body Effects on the Matching of a 2.45 GHz Coplanar-fed Antenna Jorge Roc as*, Nuno Pires* and Anto nio Moreira* Instituto de Telecomunicaco es, Instituto Superior Te cnico, Universidade Te cnica de Lisboa Av. Rovisco Pais 1, Lisboa, Portugal jorge.rocas@ist.utl.pt Academia Militar, Exe rcito Portugue s Rua Gomes Freire, Lisboa, Portugal Laboratoire d Electromagne tisme et d Acoustique, E cole Polytechnique Fe de rale de Lausanne STI-IEL, Station 11, CH15 Lausanne, Switzerland Abstract This paper presents a coplanar-fed antenna designed to operate close to the human body in the 2.45 GHz Industrial Scientific Medical (ISM) band. The antenna matching is studied in free space and taking into account body proximity. Two models of the human arm were used in the simulation and the measurements were conducted in a human user. In addition, antenna efficiency in free space was simulated and measured using a cavity-based approach. The simulated radiation patterns and Specific Absorption Rate (SAR) are also presented. It was confirmed that the proposed antenna detunes near the body, but still has good matching considering a -6 db criteria. Both arm models show a similar behavior. A more simplified arm model is sufficiently accurate for use in reflection coefficient simulations. The results of the measurements, in proximity to the body show a behavior already observed in the simulations. The simulated SAR values are below the recommendations for the two human models used. Index Terms coplanar-fed antenna, 2.45 GHz Industrial Scientific Medical (ISM) band, body proximity, models of the human arm, Specific Absorption Rate (SAR). presented Specific Absorption Rate (SAR) values below the recommendations [9], 4 W kg 1, for an average weight of 1g. II. S TUDY AND DESIGN OF THE ANTENNA Based on a previous study [1] a new structure of a coplanar antenna operating in the ISM band 2.45 GHz is presented. The proposed geometry of the antenna has an overall dimension of 41.4 mm 46. mm, being the central element an extension of coplanar line (CPW) of 5 Ω. Surrounding the central element are two ground planes containing the elements responsible for the radiation. Fig. 1 shows the geometry with respective dimensions in millimeters and the built prototype. The proposed antenna is printed in TM a.75 mm thick RO33. This material has a relative dielectric constant of 2.45 GHz and a tangent of the loss angle 2.45 GHz. I. I NTRODUCTION Body Area Networks (BAN) have increasingly sparked interest in diverse areas, from health care [1] to entertainment business [2]. These networks operating in proximity (offbody), on (on-body) or inside the human body (in-body) cause changes in the parameters of the antennas, affecting their overall performance. Typically the proximity of the body detunes the antenna, modifies the radiation pattern and decreases efficiency. Several studies have shown those effects, either for antennas operating in the 2.45 GHz ISM band [3], [4], [5], in Ultra-Wideband band (UWB) [6], [7] or Medical Implant Communication Service band (MICS) [8]. In this study we present a coplanar antenna working at 2.45 GHz ISM band. Its behavior in free space and in presence of the human body was studied. In order to obtain a good association between measured and simulated results, two models of a human arm were used, characterized with the respective physical features in the band of interest. Simulations showed that the simplest model is accurate enough to be used in simulations of the reflection coefficient. Both models Fig. 1: Proposed antenna: geometry with dimensions in mm; prototype picture. To increase the mechanical stability of the antenna, the substrate extends 1 mm beyond the metallization, except at the power supply component. In order to reduce possible return currents to the cable, during measurements, the ground

2 2 planes have a length of approximately λ/4 in the guide ( GHz), predicting therefore a decrease in the effects of measuring electrically small antennas [11], [12]. Using the CST TM Microwave Studio software, simulations show that the antenna has a similar behavior to an aggregate of two dipoles, Fig. 2. Here we can identify two areas where the concentration of current at the frequency 2.45 GHz is higher, being these the elements responsible for the radiation ( dipoles ). xy plane the antenna presents a maximum value relative to the xx axis, although not significant (around.25 dbi). In the xz horizontal plane the antenna also has a symmetrical behavior about the yy axis, in which the minimum values of the three planes predominate (xz, yz and xy). In this plane the maximum values are oriented in the the zz axis. This type of diagram is a typical example of an aggregate of dipoles, which has a higher directivity in front and back (zz axis) of the aggregate. (c) Fig. 4: Radiation pattern of the simulated antenna in free space at 2.45 GHz (CST): In the xz plane; In the yz plane; (c) In the xy plane. Fig. 2: Analysis of current distribution, in simulation, of the antenna at 2.45 GHz (CST). The reflection coefficient, S 11, both simulated and measured (using the vector analyzer Agilent E8361A (VNA)), displays a very pronounced resonance at 2.45 GHz, Fig. 3. Based on this parameter the operating band of the antenna, using the db criteria, was set. The obtained operation band of the simulation was 2.29 to 2.73 GHz, translating into a of 17.53% bandwidth according to Equation 1 [13]. The obtained experimental band operation was 2.32 to 2.73 GHz resulting in a 16.24% bandwidth according to Equation LB % = 2 f max f min f max + f min (1) Simulation Measurement Reference db ,5 4 Fig. 3: Comparison between simulation and measurement of S 11 of the antenna in free space. The radiation patterns in simulation, at 2.45 GHz, Fig. 4, shows in yz and xy vertical planes, an behavior substantially symmetrical about the yy axis. In the yz plane the antenna presents a maximum value according to the zz axis. In the In the laboratory, the radiation patterns weren t studied, because of the difficulties in doing so for small antennas [14].In these antennas the radiation efficiency is a more interesting parameter [15], [16], therefore radiation patterns were not included in this study. In simulation the radiation efficiency and total efficiency of the antenna at 2.45 GHz showed values of 99.77% and 98.13% respectively. In laboratory using a method suggested in [17], [18], [19], we obtained a radiation efficiency and total efficiency of 95.%. An study on antenna behavior, when some of its parameters are changed, was also conducted. This included: 1. changing the CPW line width in ± 5% (parameter F of Fig. 1a), 2. changing the relative dielectric constant of RO33 TM (within the range of values related to the manufacturing process, 3.35±.4) and 3. changing the width of the elements responsible for radiation (parameter K of Fig. 1a), both within ± 5%. The results showed that the antenna is very sensitive to changes of CPW line width, which may even compromise the coverage of the band in question. Changes in the dimensions of the radiator elements and in the relative dielectric constant of the substrate do not affect the antenna performance at 2.45 GHz ISM band. III. STUDY IN THE PRESENCE OF THE HUMAN BODY After studying the antenna in free space, in this section, a study of its behavior near the human body was conducted. Here two models of the human arm were used, as shown in [7], Fig. 5. The model presentd in Fig. 5a is more similar to a real arm and presents an elliptical shape, with a 57 mm and 45 mm radius in major and minor axis, respectively. The model presents a 2 mm thick skin layer, a 4 mm fat layer, a 17.5 mm radius bone layer with a 4 mm offset from the center of the ellipse and the muscle layer that fills the space between the bone and fat. The arm section is 14 mm long. To reduce simulation time in calculating the reflection

3 3 coefficient, radiation patterns, radiation efficiency and total efficiency, a simpler model is proposed. This model, see Fig. 5b, is composed of four layers. A 2 mm thick skin layer, followed by fat, muscle and bone deeper layers with 4 mm, 17.5 mm and 17.5 mm thickness respectively. This flat model of the human arm is 14 mm long by 114 mm wide. Fig. 5: Models used in the simulation (CST): Elliptical; Flat. Since the human body is a complex biological structure, with various organs and tissues endowed with particular dielectric properties, is important a good characterization of the tissues that constitute the model. Therefore an online tool is used [2] to characterize the dielectric properties of different tissues. The density of each tissue was previously defined, as required for calculating the SAR. The used average values were, as follows: skin ρ = 11 kg m 3 ; fat ρ = kg m 3 ; muscle ρ = 16 kg m 3 ; bone ρ = 15 kg m 3. To perform the simulations on the body, the antenna is placed in proximity of the models. Here, the antenna is centered with the structure of each of the models, leaving a space of 6 mm between the back of the antenna and the model. The simulation results of S 11, Fig. 6, clearly show a slight change of the antenna tuning in direction of lower frequencies. A decrease in the magnitude of the S 11 parameter, due to misadaptation of the antenna caused by the absorption of energy by the models (human body) was found. 5 the arm, ranges from 2.4 to 2.8 GHz, corresponding to a bandwidth of 31.4%. In the case of the flat arm model, the band of operation ranges from 2.5 to 2.81 GHz, which represents a bandwidth of 31.69%. In comparison with the band of operation simulation at -6 db, of the antenna in free space ( GHz; LB % =37.7), there exists a lower bandwidth in the proximity of the models. As depicted in Fig. 6, if using the -6 db criteria in determining the bandwidth of operation of the simulation, in the presence of the models of the human arm, the ISM band (2.45 GHz) is fully covered. Fig. 7 shows the radiation patterns in xz, yz and xy planes, in the presence of the two models of the human arm, for a 2.45 GHz frequency. The green and blue curves describe the behavior of the antenna in the presence of the two models, elliptic and flat respectively. In xz and yz planes, it appears that nearly the entire lobe of radiation, which propagates in the direction of the models, is absorbed. This absorption is due to the high relative dielectric constants, ɛ r, characteristic of the tissues that compose the models. It is also apparent an amplification around 4 db, in the radiation lobe oriented in opposite direction to the models, when compared to the simulation in free space. This amplification results from the models work with reflectors, in the presence of the antenna. In the xy vertical plane, there is a symmetrical behavior about the yy axis. Note that in this plane, the magnitude of radiation is very small (see scale). (c) Fig. 7: Radiation pattern of the simulated antenna, in the presence of the models of the human arm, for the 2.45 GHz frequency (CST). Green curve, elliptical model, blue curve, flat model: In the xz plane; In the yz plane; (c) In the xz plane -2-3 Next to elliptical model Next to flat model In free space Reference -6 db Fig. 6: Simulation of the reflection coefficient of the antenna in free space and in the presence of the elliptical and flat models of the arm. According to the -6 db criteria, the antenna bandwidth of the simulation, in the presence of the elliptical model of The values of radiation efficiency and the total efficiency of the antennas were, respectively, 18.5% and 16.6% for the elliptical arm model and 13.47% and 15.65% for the flat arm model. Regarding the values obtained from the radiation efficiency and the total efficiency of the antenna in free space, 99.77% and 98.13% respectively, in the arm models there is a efficiency reduction of around 8%. After the simulations of the antenna near the human body, the built prototype was subjected to laboratory tests. Here were studied the effects of the human body, in three conditions adjacent to S 11 and in the operating band of the projected antenna. All the tests were performed using the vector analyzer Agilent E8361A (VNA). In the measurements of the antenna close to the body, three scenarios were created. In the first scenario, condition I, the antenna is placed along the arm of a volunteer, recreating the

4 4 best possible the simulation scenario. A small 6 mm thick strip of polyethylene ( r 1), was used in order to avoid contact with the arm. For adjusting the antenna near the arm, a band of latex is used. Fig. 8 shows the first measurement position (condition I). Fig. 1: Third situation of the measuring (Condition III), of antenna near the human body. can be compared to the simulations near the human body. Fig. 8: First scenario created (Condition I), of antenna near the human body. In the second scenario, condition II, the antenna is tested exactly as in condition I, except for the volunteer wearing a military jacket. This jacket is part of the Portuguese Army uniform being worn in places where atmospheric conditions, such as cold, rain and snow prevail. In Fig. 9, we can observe the jacket worn by the subject, with the respective details during the measurement. -2 Simulation (flat model) Simulation (elliptical model) Measurement (in arm) Measurement (in arm with jacket) Measurement (in arm with wet jacket) ,5 4 Fig. 11: Measurement of reflection coefficient of the antenna in the three conditions (condition I, II and III) and simulation in the presence of elliptical and flat model of the arm. (c) Fig. 9: Second scenario (Condition II), of antenna near the human body: Coat used (front and back); Detail of the antenna within the sleeve; (c) Detail of the measuring cable in coat sleeve Finally, in a third scenario, condition III, the behavior of the antenna under rainy weather conditions is tested. In addition to condition II the coat sleeve under which the antenna is placed is sprayed with water. Although it is a simple test, far from recreate the plenty of water during a rainy day, allows observing its effect on the antenna performance. Fig. 1 illustrates the condition III, the measurements performed in the laboratory, and the coat after being sprayed with water. The measurements of the S11 parameter, in the three scenarios (condition I, II and III) are shown in Fig. 11, and A reasonable concordance between simulation and measurement curves up to 2.75 GHz can be observed, with differences in higher frequencies. There are also differences in the magnitude of S11, showing a decrease in its module values of the measured curves when compared to the simulated curves. Although only the condition I curve recreates the simulation scenario, the curves of conditions II and III only slightly differ compared to the former. In these latter additional measurement scenarios (condition II and III) the response of the antenna, regarding the S11 parameter, shows no significant changes. As shown in the graph of Fig. 11, when determining the bandwidth of the measurements using the db criteria, the ISM band (2.45 GHz) is not fully covered. On the other hand if applying the -6 db criteria, the band of operation can not be defined, due to the behavior of the curves (no intersection with the -6 db reference line). Despite the lack of a determinable band of operation, the ISM band (2.45 GHz) is completely covered, to a range between -8 S11-6 db. A major concern related to new technologies operating in the proximity of the human body, is the electromagnetic radiation emitted by the antennas. Since the Wearable Antennas (WA) are close or even in contact with the human body, it is of paramount importance that the level of electromagnetic radiation does not exceed the official recommendations [9]. The study of the SAR is obtained by simulation with the TM CST Microwave Studio software, for both models Fig. 5a and Fig. 5b, at 2.45 GHz. These results were calculated for an average mass of 1g of tissue. The models show a

5 5 simulation volume of 1.13 dm 3 and.65 dm 3, elliptical and flat, respectively. Based on the density of the tissues present, the above software also calculates its mass, which is 1.24 kg to the elliptical model and.81 kg to the flat model. The power value at the input, to calculate the SAR, was determined for a maximum equivalent isotropically radiated power (EIRP) [21] of 2 dbm, and an antenna gain of dbi in free space, at 2.45 GHz. Conversly, applying the Equation 2, was obtained the maximum input power of the antenna (P Twatt =.5). ( EIRP ) [dbm] G [dbi] P T watt = (2) The resulting SAR values were, in both models (1.515 W kg 1 and 1.42 W kg 1, for elliptic and flat models, respectively), below the recommended by the International Commission on Non-Ionizing Radiation Protection (ICNIRP), at 4 W kg 1. IV. CONCLUSION This paper presented a new structure of a CPW antenna for applications within the BAN, at the 2.45 GHz ISM band. The measurements and simulations in free space parameters: S 11, radiation efficiency and total efficiency have shown a good correlation, in part due to a good computational model in CST TM Microwave Studio software. The simulated radiation patterns had an aggregate of two dipoles behavior. To study the performance of the antenna near the human body, two models of a human arm were suggested. The dielectric characteristics of the tissues on the frequency range of 1 to 4 GHz, as well as the respective densities were accounted for. The flat model proved to be accurate enough to calculate the S 11 parameter. The radiation patterns, as tested in the models, show a decrease in the lobe directed towards the human tissues and an amplification in the lobe oriented in the opposite direction. This was the expected and desirable behavior, reinforcing the idea that the antenna is suitable for off-body connections. In the laboratory three testing scenarios were created, the first of which (conditon I) best correlated with the software simulation. Additionaly, the results of the S 11 parameter were very similar for the other two scenarios (conditions II and III). Both in free space and in the presence of the human body, in simulation and in the laboratory, a coverage of the entire 2.45 GHz ISM band was achieved, using the -6 db criteria. The SAR values of W kg 1 for the elliptical model and 1.42 W kg 1 for the flat model, for an average mass of 1g with a power input of.5 W, were inferior to the international recommendations of 4 W kg 1. [3] A. Aroul and D. Bhatia, Study of performance and propagation characteristics of wire and planar structures around human body, in Engineering in Medicine and Biology Society,EMBC, 211 Annual Interna- tional Conference of the IEEE, sept , pp [4] S. Lee, U. Kim, K. Kwon, W. Seo, and J. Choi, Design of on-body antenna for wireless body area network, in Advanced Communication Technology (ICACT), th International Conference on, feb. 212, pp [5] T. Kellomaki and W. Whittow, Bendable plaster antenna for 2.45 ghz applications, in Antennas Propagation Conference, 29. LAPC 29. Loughborough, nov. 29, pp [6] Z. N. Chen, A. Cai, T. See, X. Qing, and M. Chia, Small planar uwb antennas in proximity of the human head, Microwave Theory and Techniques, IEEE Transactions on, vol. 54, no. 4, pp , june 26. [7] M. Koohestani, N. Pires, A. K. Skrivervik, and A. A. Moreira, Influence of the human body on a new coplanar-fed ultra-wideband antenna, in Antennas and Propagation (EUCAP), 212 6th European Conference on, march 212, pp [8] J. Choi, U. Kim, S. Lee, and K. Kwon, Design of an implantable antenna for wban applications, in Antenna Technology (iwat), 212 IEEE International Workshop on, march 212, pp [9] ICNIRP, Guidelines for limiting exposure to time-varying electric, magnetic, and electromagnetic fields (up to 3 ghz), [Online]: Available: July 212. [1] Q. Bai and R. Langley, Bending of a small coplanar textile antenna, in Antennas and Propagation Conference (LAPC), 21 Loughborough, nov. 21, pp [11] O. Staub, J.-F. Zürcher, and A. Skrivervik, Some considerations on the correct measurement of the gain and bandwidth of electrically small antennas, Microwave and Optical Technology Letters, vol. 17, no. 3, pp , [12] J.-F. Zürcher, O. Staub, A. K. Skrivervik, and M. Hermanjat, Accurate measurement of the maximum gain of electrically small antennas, Microwave and Optical Technology Letters, vol. 23, no. 6, pp , [13] C. A. Balanis, Antenna Theory: Analysis and Design, 2nd ed. John Wiley & Sons Ltd, [14] C. Icheln, Methods for measuring rf radiation properties of small antennas, Ph.D. dissertation, Helsinki University of Technology, 21. [15] K. Rosengren, P.-S. Kildal, J. Carlsson, and O. Lunden, A new method to measure radiation efficiency of terminal antennas, in Antennas and Propagation for Wireless Communications, 2 IEEE-APS Conference on, 2, pp [16] V. Tavakol and P.-S. Kildal, Relations between radiation characteristics of small antennas located in free space and close to lossy phantom, in Antennas and Propagation Society International Symposium 26, IEEE, july 26, pp [17] C. Mendes, Radiation efficiency of small printed antennas, Master s thesis, Instituto Superior Técnico, 26. [18] C. Mendes and C. Peixeiro, Theoretical and experimental validation of a generalized wheeler cap method, in Antennas and Propagation, 27. EuCAP 27. The Second European Conference on, nov. 27, pp [19] C. Mendes and C. Peixeiro, Radiation efficiency of several handset antennas obtained with a modified wheeler cap method, in Antennas and Propagation, 27. EuCAP 27. The Second European Conference on, nov. 27, pp [2] [Online]: Available: June 212. [21] S. Kawade, T. Hodgkinson, and D. Wisely, Impact of eirp limit changes on the performance of 2.4 ghz wifi systems, in Vehicular Technology Conference, 27. VTC27-Spring. IEEE 65th, april 27, pp REFERENCES [1] W. Scanlon, G. Conway, and S. Cotton, Antennas, propagation considerations for robust wireless communications in medical body area networks, in Antennas and Propagation for Body-Centric Wireless Communications, 27 IET Seminar on, april 27, p. 37. [2] H.-B. Li, K.-i. Takizawa, B. Zheri, and R. Kohno, Body area network and its standardization at ieee mban, in Mobile and Wireless Communications Summit, th IST, july 27, pp. 1 5.