MODERN hospitals have a complex infrastructure, which. Simulation of Radiowave Propagation in Hospitals Based on FDTD and Ray-Optical Methods

Transcription

1 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 53, NO. 8, AUGUST Simulation of Radiowave Propagation in Hospitals Based on FDTD and Ray-Optical Methods Thomas M. Schäfer, Student Member, IEEE, and Werner Wiesbeck, Fellow, IEEE Abstract This paper addresses the simulation of radiowave propagation in hospitals in the frequency range from 42.6 MHz to 5.2 GHz. Due to the special construction of the walls, wave propagation in hospitals is different from other buildings. These walls contain metallic layers within their structure, e.g., operating rooms with CrNi-steel faced walls, X-ray rooms with lead shielded walls or magnetic resonance tomography rooms with an explicit EMC shielding made of copper. Wave propagation in such rooms has not been simulated yet. This work uses a simulation tool based on the finite-difference time-domain method that allows a detailed simulation of the walls, floors and ceilings for frequencies from 42.6 to 300 MHz. For higher frequencies up to 5.2 GHz a ray-optical simulation tool must be used and the walls must be modeled homogeneously by multiple layers of different materials. In order to accurately model the walls, the electromagnetic properties of the different material layers inside the walls have to be known. The results of extensive wave propagation measurements in four different hospitals are used for the determination of these parameters. This paper presents the methods used to obtain these material parameters and their performance in a ray-optical simulation tool. Index Terms Clinic, hospital, mobile communications, ray optics, simulation, wall attenuation, wave propagation. I. INTRODUCTION MODERN hospitals have a complex infrastructure, which requires the collaboration of multiple information services to provide reliable medical care. Many trend-setting innovations have led to the increased performance of computers and medical equipment. More and more radio networks are used to assure high mobility and to keep wiring low. The new techniques lead to significant chances in the hospital environment [1] [3]. One example of these new products is the electronic ward round, which is a wireless PDA that provides doctors with a digital patient file. This reduces paperwork to a minimum [4]. Further examples are radio location and monitoring of all medical devices in a hospital, wireless data transmission between medical devices and patient monitoring equipment, and internet connections via handheld PCs for bedridden patients, which has been tested in a pilot study in Germany [5]. The installation of these services requires radio networks in hospitals. However, it is necessary to ensure the electromagnetic compatibility (EMC) of medical devices. Therefore, crit- Manuscript received September 25, 2004; revised March 20, This work was supported by the Deutsche Forschungsgemeinschaft (DFG) under the framework of the Sonderforschungsbereich 425 (Electromagnetic Compatibility in Medical and Industrial Environments). The authors are with the Institut für Höchstfrequenztechnik und Elektronik (IHE), University of Karlsruhe, Karlsruhe, Germany ( Thomas.Schaefer@ihe.uka.de). Digital Object Identifier /TAP ical areas must be checked for compliance with international standards for medical equipment [6] [8]. These tasks require knowledge of wave propagation in hospitals. There have only been a few studies on wave propagation in hospitals [9], [10], primarily because it is difficult to perform measurements in hospital traffic, especially in sterile operating rooms. This work is the first fundamental contribution to this topic. Extensive measurements were performed in four different hospitals. The results of those measurements are presented in [11] [13]. This paper uses those measurements to determine parameters required for a ray-optical tool to simulate wave propagation in hospitals in the frequency range from 600 MHz to 5.2 GHz. The paper presents the results of model runs with those parameters and a comparison of the model results with the measurements. For lower frequencies from 42.6 to 300 MHz a tool based on field theory finite-difference time-domain (FDTD) is used for detailed investigations MHz is the operating frequency of a 1-Tesla magnetic resonance tomography (MRT) system. II. DESCRIPTION OF THE HOSPITAL ENVIRONMENT The walls in modern hospitals are built with drywall that is composed of varying materials, depending on the room usage. The basic construction is composed of steel studs upon which the different types of walls are hung. A hollow space that has various uses, including electric cabeling, water pipes, pipes for medical gases or ventilation ducts remains between walls, leading to an inhomogeneous wall structure that can have strongly varying electromagnetic properties. The type of wall construction depends on the room usage in the hospital. These rooms include: Operating rooms, X-ray, MRT and computer tomography rooms (CT), as well as ordinary hospital rooms (therapy rooms, patient rooms). All of these rooms are quite different in their arrangement and in the construction of their walls, which means each room may have different wave propagation properties and must be considered individually. The detailed construction of the different types of walls follows. A. Operating Rooms The walls in operating rooms consist of 1.8 cm of plasterboard and 0.8 mm of CrNi-steel paneling on both sides. This metallic facing is used for hygienic reasons, better fire protection and its flexibility in modifications. It also has a considerable influence on the propagation of EM waves. Another wall construction used in modern operating rooms consists of a Trespa facing (solid panels that consist of tightly-pressed wood and glue that are extremely durable [14]) and an additional shielding X/$ IEEE

2 2382 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 53, NO. 8, AUGUST 2005 TABLE I COMPOSITION OF WALLS IN HOSPITALS AND THEIR MATERIAL PARAMETERS FOR RAY-OPTICAL SIMULATIONS Fig. 1. Distribution of the field strength in the partition wall between two X-ray rooms for 42.6 MHz and vertical polarization (grayscale: 0 to 50 db). tool with this feature was developed [15] and used in the present study in order to investigate the wave propagation between adjacent rooms. The tool is based on the FDTD method. of 0.5 mm of lead. In contrast to the wall with CrNi-steel, where the metallic facing is mounted on both sides of the wall, the wall with Trespa has only one metallic layer in its structure. B. X-Ray and CT Rooms Rooms for X-ray and CT examinations have walls of 2 cm of plasterboard on both sides and an additional layer of lead of 0.5 to 2.5 mm thickness on the inside of the wall, depending on the strength of the X-ray apparatus. C. MRT Rooms Rooms for MRT systems have an explicit EMC shielding, which is commonly constructed with a surrounding foil of copper (0.16 mm thickness). This is an important component in this wall structure with respect to EM wave propagation. D. Ordinary Hospital Rooms Walls in ordinary hospital rooms (therapy rooms, patient rooms) consist of the previously mentioned skeletal steel structure with 2 cm of plasterboard on both sides. III. FDTD SIMULATIONS IN THE VHF-BAND For a highly accurate simulation of the wave propagation in an environment with elements that are small compared to the wavelength, numerical tools based on the solution of Maxwell s equations (field theory) are necessary. Using such a tool it is possible to simulate wave propagation through a slot that is smaller than the wavelength. This feature is important for the simulation of hospital rooms with metallic layers in the walls. A simulation A. Simulation Results Three different types of rooms are investigated: An X-ray room (4.7 m 5.0 m 2.85 m), an operating room (7.2 m 6.0 m 3.0 m) and an MRT room (6.0 m 5.0 m 2.7 m). The composition of the walls is listed in Table I and the floor and the ceiling consist of 33.5 cm of concrete. Two rooms are considered in all simulations: The room under test and an adjacent room of the same type that contains the transmitter. No furniture is included in the simulations, because the rooms were empty during the measurements. The simulation area is surrounded by open boundary conditions. The simulations are performed for the frequencies 42.6, 100, and 300 MHz. As a transmitter, a vertically polarized Hertzian dipole is used for all simulations. Comparable simulations with a vertically polarized -dipole lead to no significant changes in the attenuation between adjacent rooms. The transmitter is located in the middle of one of the two adjacent rooms at a height of 1 m. The cell sizes are chosen to be variable between 1 and 8 cm. The time steps have a length of s with a total simulation time of s. As an example, the results for the X-ray room are presented. Walls in X-ray rooms contain a layer of lead, while floor and ceiling consist of concrete without additional shielding. The lead shielding is made for X-rays and therefore does not shield high frequencies (HF). Slots in real walls are primarily vertically oriented because of the typical vertical construction of the walls. Therefore, two vertically-oriented slots with a length of 1 m are integrated into the lead shielding in the simulations as characteristic examples. The width of the slots is much smaller than the wavelength. The lead can be modeled as a perfect electric conductor, because the penetration depth ( m at 42.6 MHz) is much smaller than the width of the lead (0.5 to 2.5 mm). The simulation results in Figs. 1 3 and the cross section in Fig. 4 (see Fig. 5 for grayscale description) show that the waves propagate via the floor and the ceiling into the adjacent room. The propagation through the slots (two vertical light cylinders in the pictures) increases in importance for higher frequencies but still remains negligible, as long as the width of the slots is much

3 SCHÄFER AND WIESBECK: SIMULATION OF RADIOWAVE PROPAGATION IN HOSPITALS 2383 Fig. 2. Distribution of the field strength in the partition wall between two X-ray rooms for 100 MHz and vertical polarization (grayscale: 0 to 50 db). Fig. 6. rooms. Concept for the measurement of the room attenuation between adjacent Fig. 3. Distribution of the field strength in the partition wall between two X-ray rooms for 300 MHz and vertical polarization (grayscale: 0 to 50 db). Fig. 7. Comparison of measured and simulated room attenuations. Solid line: measurement; dashed: simulation; squares: MRT room; circles: operating room with CrNi-steel paneling; x-marks: X-ray room. Fig. 4. Cross section of the field-strength distribution in the two X-ray rooms under test for 300 MHz and vertical polarization (grayscale: 0 to 50 db). Fig. 5. Gray-scale used in Figs. 1 to 4. smaller than the wavelength [16], [17]. As a consequence, the length of the vertical slots has a negligible influence on vertical polarization. B. Measurement of the Room Attenuation Many measurements of the wave propagation in hospitals were performed and the results are presented in [11] [13], including the room attenuation of different hospital rooms. In order to determine the so-called room attenuation, a receiver was located inside a room and the path loss to the adjacent rooms was measured. Fig. 6 shows the measurement concept for the room attenuation. The path loss is measured for all transmitter-receiver combinations and then the average path loss for this room is determined. Many measurements in different rooms of the same type are performed in order to obtain a meaningful average for the room attenuation for one type of room. The detailed measurement setup is described in [11] [13]. C. Simulations Compared to Measurements The room attenuation was also determined using the simulation results from Section III-A. In order to determine the room attenuation from the simulations, the field strength is analyzed at 15 different points spread all over the room under test. The points are positioned at the vertices of a cuboid (8), at the center of its faces (6) and at its center point (1). The size of the cuboid is chosen to ensure a 1 m distance between it and the walls. The room attenuation is the result of the average attenuation at all 15 receiver points. Fig. 7 depicts the comparison between measurement and simulation. Both, simulations and measurements, are performed for vertical polarization. Some slots with realistic lengths are exemplarily integrated in the metallic layers of the X-ray, operating and MRT rooms [13]. Free space attenuation is not included in the data of Fig. 7. The results reveal that there are big differences in the attenuation of different rooms. Especially the MRT room shows a

4 2384 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 53, NO. 8, AUGUST 2005 very high attenuation of more than 70 db. The attenuation of the operating room and the X-ray room is about 30 db and about 20 db, respectively. For comparison, the attenuation of an ordinary room with walls that consist of plasterboard is only 10 db or less [11]. The simulations are in good agreement with the measurements. When using numerical tools, the RAM resources of a computer limit the upper simulation frequency of a given scenario, because the simulation area has to be divided into cuboids that are much smaller than the wavelength. With 4 GB of RAM and a simulation area of 10 m 5m 3 m for example it is possible to simulate frequencies up to about 500 MHz. Therefore, other simulation tools must be used for higher frequencies. For example, it would be possible to use hybrid ray-tracing/fdtd techniques that allow for a simulation of large areas in combination with a detailed consideration of the walls [18]. However, detailed information on the assembly of the walls (e.g., the location of steel studs or slots in the shielding) is not available. Ray-optical tools turn out to be most suitable and effective. IV. MATERIAL PARAMETERS FOR RAY-OPTICAL SIMULATIONS A ray-optical simulation tool that was developed for indoor environments is used to simulate wave propagation in hospitals for frequencies from 600 MHz to 5.2 GHz [19]. The tool is able to handle reflection and transmission, but not diffraction. Furniture and other instruments are not included in the simulations, because they are also not included in the measurements (the measurements were performed before inauguration). Wall construction in hospitals is complex and inhomogeneous, with the exact location of inner wall elements unknown, which makes an exact modeling of each wall virtually impossible (for all types of simulation tools). To further complicate wall modeling, some wall constituents (e.g., the steel studs) have dimensions smaller than the wavelengths used and thus cannot be processed with geometrical optics. Nor can numerical simulations (field theory) be used because of the enormous demand for computational resources when doing calculations over large areas and high frequencies. A reasonable solution to this problem is to heuristically model the walls in order to obtain correct results on average. The walls are then modeled as being homogeneous, composed of multiple layers that can be calculated with ray-optical simulation tools. The procedure for determining the material parameters (permittivity, permeability and conductivity) of the different wall layers is described in the following paragraphs. For common wall types, the material parameters of most of the wall s materials (e.g., plasterboard, chipboard, concrete, etc.) can be found in the literature [20], [21]. However, the parameters cannot be found for special hospital walls that contain metallic layers in their structure. If the real conductivity of these metals (copper, lead, CrNi-steel) were to be used in the simulations, it would lead to infinite shielding in the walls. For this reason, the material parameters of the metallic layers must be determined by measurements, which were performed in different hospitals. Two different kinds of measurements were Fig. 8. Measurement-based determination of material parameters for ray-optical simulations in a patient room at 2.45 GHz. used to characterize the wave propagation: Individual wall attenuation and the attenuation between adjacent rooms [11], [12]. The measurements were performed at five discrete frequencies between 600 MHz and 5.2 GHz. The attenuation of individual walls was measured by positioning horn antennas on both sides of the wall and measuring the attenuation of the transmitted power directly through the wall. This kind of measurement is a common procedure. For practical purposes, however, the propagation behavior of an entire room (not of an individual wall) is the primary interest. Therefore, the path loss between two adjacent rooms was additionally measured by placing an omnidirectional monoconical receiver inside the room and an omnidirectional biconical transmitter in all adjacent rooms. The measurement of path loss gives an average attenuation between adjacent rooms [11], [12]. The results of both types of measurements were used to find proper material parameters for the critical layers of the walls. In ordinary rooms with plasterboard walls, the characteristics of the middle layer of sheet rock are unkown, because it also contains steal studs and other metallic components. In the other rooms, the critical layers in the walls are the metallic layers (e.g., lead, CrNi-steel, etc.). In order to determine the effective conductivity of these (metallic) layers, the measurement setup was copied by the simulation and the conductivity was optimized to obtain agreement between simulation and measurement. The parameters and remained unchanged for all materials. This optimization was done for the wall attenuation and the room attenuation respectively. Consequently, two sets of parameters were obtained, one based on the measured attenuation of individual walls and the other based on the measured attenuation between adjacent rooms. Fig. 8 shows the simulation results for the material parameters determined for an ordinary room. For measurement point 1, the receiver and the transmitter were both located inside the room; for measurement points 2 5, the transmitter was positioned in adjacent rooms. The simulation was performed for both parameter sets. Both simulations fit the measurement results well.

5 SCHÄFER AND WIESBECK: SIMULATION OF RADIOWAVE PROPAGATION IN HOSPITALS 2385 Fig. 9. Measurement based determination of material parameters for ray-optical simulations in an operating room with CrNi-steel paneling at 2.45 GHz. TABLE II FREQUENCY-DEPENDENT CONDUCTIVITY (f ) OF THE EMC SHIELDING OF MRT ROOMS FOR RAY-OPTICAL SIMULATIONS Fig. 10. Path loss in the second floor of a hospital at 900 MHz with unmodified material parameters. A detailed performance of more simulation results compared to measurements will be outlined in Section VI. It should be mentioned that the parameters do not depend on the size of the room; they only depend on the composition of the walls. Fig. 9 shows the simulation results for the determination of the material parameters in an operating room with CrNi-steel faced walls. The simulation results are shown for both parameter sets again. For the measurement points 1 3, the receiver and the transmitter were both positioned inside the operating room; for points 4 13, the transmitter was positioned in adjacent rooms. The figure shows that the simulation with the material parameters gained from the wall attenuation leads to a much higher path loss than the measurement. This behavior is due to the metallic layer of CrNi-steel in the wall. The measurement of the wall attenuation (which only considers a small part of the wall) does not take into account all propagation paths that arise between two adjacent rooms and therefore leads to a higher attenuation. The simulation with the parameters gained from the room attenuation, however, fits the measurements well. For most of the walls it was possible to obtain material parameters that are valid for all frequencies in the range from 600 MHz to 5.2 GHz. This was not possible for the MRT room, however, because the EMC-shielding of this room induces a more frequency-dependent behavior. Therefore, the material parameters had to be determined separately for all five frequencies. The components of the investigated hospital walls and the material parameters of the different material layers are listed in Tables I and II. The conductivity is not included in (1) V. IMPLICATIONS OF THE NEW MATERIAL PARAMETERS This section presents a comparison between the performance of the old (unmodified from literature [20], [21]) and the new material parameters (from Tables I and II). Therefore two simulations were performed over a large area in a hospital with plasterboard walls, lead shielded walls and an EMC shielded MRT room. A Hertzian dipole was used as a transmitter at a frequency of 900 MHz. The antenna was positioned in a therapy room, which has the same walls as a patient room. A total number of 20 reflections and/or transmissions were taken into account in the simulations and the maximum path loss for a single propagation path was set to 140 db. Fig. 10 shows the simulation results for the unmodified material parameters. The walls with metallic layers in their structure cause an almost infinite shielding in the corresponding rooms, e.g., the X-ray rooms on the left side of Fig. 10, the MRT room on the lower left and the CT room right beside it. The abovementioned termination criteria are fulfilled within these rooms because of the metal (too many reflections or a too high attenuation), which results in totally black areas. It is striking that walls made of plasterboard cause virtually no attenuation between rooms, which results in an almost homogeneous distribution of the field strength over a large area. Fig. 11 shows the same simulation with the new material parameters presented in Section IV. The same parameters are used over the entire simulation area for the same type of wall. The black areas have disappeared, which means that the path loss could be determined in all rooms. Nevertheless, the attenuation of the walls with metallic layers in their structure is still high compared to the other walls, especially in the MRT room (lower left). The attenuation of the ordinary rooms (walls made of plasterboard) is also higher than in Fig. 10, which leads to an overall higher path loss, especially for the rooms on the upper

6 2386 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 53, NO. 8, AUGUST 2005 Fig. 13. Positions of the receiver (Rx) and the transmitter (Tx) for the simulations and the measurements in Fig. 12 (dimension of the CT room: 6.1 m m m). Fig. 11. Path loss in the second floor of a hospital at 900 MHz with new material parameters (see Tables I and 2. Fig. 14. Measurement and simulation of the room attenuation in an X-ray room with lead shielded walls at 900 and 2450 MHz. Fig. 12. Measurement and simulation of the room attenuation of a CT room with CrNi-steel faced walls at 900 and 2450 MHz. left that are additionally shadowed by the X-ray rooms in front. The higher path loss of the ordinary rooms is due to the metallic components inside the walls (steel studs, etc.) that have not been taken into account by the unmodified parameters. Altogether, the simulation with the new parameters leads to much better results and is closer to reality. A comparison between simulations and measurements is presented in the next section. VI. SIMULATIONS COMPARED TO MEASUREMENTS The simulations in this section are performed with the new material parameters from Section IV. Two different kinds of rooms are considered as an example. First, Fig. 12 shows the measured and the simulated path loss for 900 and 2450 MHz in and around a CT room with CrNi-steel-faced walls for seven different measurement positions. The CT room is adjacent to an operating room and is used for CT examinations during surgery. The room was completely empty at the time of the measurements. The positions of the receiver and the transmitter are shown in Fig. 13. is the receiver position for the measurements inside the room (transmitter positions Tx1-Tx3) and is the receiver position for the transmitter positions outside the CT room (Tx4-Tx7). The transmitter and the receiver are simulated as -dipoles with vertical polarization, which corresponds to the measurements. The measurements with the receiver and the transmitter both located inside the room (Tx1-Tx3) result in a low attenuation. Tx4 is located in an adjacent room with an observation window in the wall, which leads to a lower attenuation than the measurements for Tx5-Tx7. The window has also been taken into account in the simulation. The attenuation at Tx5-Tx7 of about 70 db for 900 MHz and about 80 db for 2450 MHz is very high for the relatively short distance of about 4 to 5 m between transmitter and receiver. This attenuation is caused by the CrNi-steel surface of the walls. Both frequencies, 900 and 2450 MHz, show good agreement between simulation and measurement. Fig. 14 shows the comparison of the measured and simulated path loss in an X-ray room with lead-shielded walls. Again, the two frequencies 900 and 2450 MHz are considered. In addition to the measurements inside and outside the room, Fig. 14 also shows two measurements in the adjacent floors (points 8 and 9) for the determination of the attenuation of floor and ceiling.

7 SCHÄFER AND WIESBECK: SIMULATION OF RADIOWAVE PROPAGATION IN HOSPITALS 2387 Fig. 15. Positions of the receiver (Rx) and the transmitter (Tx) for the simulations and the measurements in Fig. 14 (dimension of the X-ray room: 7.0 m m m). Measurement point 8 is located in the floor above and point 9 is located in the floor below the X-ray room. The other measurement points are located in the same floor as the X-ray room and are shown in Fig. 15. The path loss for the measurements inside the room (Tx1, Tx2) is again clearly lower than that of the others. In the wall toward measurement point Tx3 there is a lead-shielded observation window which has a 15 cm wide gap at the upper end of the window in order to keep conversation contact with the patients. Because of this gap, the path loss is lower at this position than at the points Tx4 Tx7. The window was also considered in the simulation. It also stands out that the path loss is very high at the points 8 and 9. In contrast to the ceiling (point 8), there is an additional sheet-metal shell in the composition of the floor, which causes a higher path loss at point 9. As in the case for Fig. 12, the path loss is higher for 2450 MHz than for 900 MHz. This effect is due to the effective antenna area and not due to the material of the walls (free space loss is included in Figs. 12 and 14). Both frequencies result in good agreement between simulation and measurement. Similar results are obtained for the other types of rooms and for the other frequencies [13]. This implies that the proposed modeling of the walls with ray-optical methods is indeed able to predict wave propagation in hospitals. VII. CONCLUSION This paper presents the first fundamental approach for simulations of radiowave propagation in hospitals. It shows that the usual measurement of the wall attenuation cannot be used for the characterization of the wave propagation in hospital areas where the walls have a metallic layer in their structure. Therefore, the room attenuation method is introduced and is found to deliver more accurate results because it considers the room as a whole and takes into account all important propagation paths between adjacent rooms. The results of the room attenuation experiments are used to determine material parameters for the hospital walls for ray-optical simulation tools. The material parameters enable the simulation of wave propagation in hospitals in the frequency range from 600 MHz to 5.2 GHz, even for rooms with walls that contain a metallic layer. For lower frequencies from 42.6 to 300 MHz, a tool based on field theory (FDTD) is used for a detailed investigation of the wave propagation between adjacent rooms. The simulations show that the floor and the ceiling give rise to important propagation paths if the walls contain a metallic layer. The propagation through slots in the metal increases with frequency and depends on polarization. Wireless networks (Bluetooth, WLAN, etc.) are installed in hospitals to enhance and facilitate work. These systems must be planned before their installation, which means that the number and the transmission power of the base stations need to be determined. Therefore the results of this work can be used in order to predict the coverage of the base stations in hospitals. It must be kept in mind that, depending on the hospital area (wall structure), the coverage of a base station can be quite different and needs individual configuration. When wireless networks are installed in hospitals, the effect of the transmitted electromagnetic waves on sensitive medical equipment must be checked. The transmission power of the base stations can be reduced by using smaller radio cells, which, at the same time, allow a higher capacity of the system. Simulations of the wave propagation can also help to evaluate and prevent EMC problems between medical devices that may interfere, e.g., sensitive diagnosis devices (MRT, ECG, EEG) and powerful therapy devices (HF-cancer-therapy, HF-surgery, hyperthermia) that work at high frequencies in the power range from picowatt to kilowatt, respectively. ACKNOWLEDGMENT The authors would like to thank the Deutsche Forschungsgemeinschaft (DFG) for the financial support of this work in the framework of the Sonderforschungsbereich 425 (Electromagnetic Compatibility in Medical and Industrial Environments). REFERENCES [1] H. Schmidt, Clear view into the human body: Europe s first wholebody magnet resonance tomography in Berlin (in German), Kmnkenhaus Technik + Management, vol. 29, no. 10, p. 26, Oct [2] R. Garcia, Hightech and workflow are important for modern hospitals (in German), Krankenhaus Technik + Management, vol. 29, no. 5, pp , May [3] R. Frenkiel, A brief history of mobile communications, IEEE Veh. Technol. Society News, vol. 49, no. 2, pp. 4 7, May [4] H. Schäfers, Quick ward round with a PDA (in German), Krankenhaus Technik + Management, vol. 29, no. 5, pp , May [5] H.-U. Schmidt, Internet access for bedridden patients (in German), Kmnkenhaus technik + management, vol. 29, no. 10, p. 10, Oct [6] IEC , Medical Electrical Equipment, Part 1: General Requirements for Safety. 2. Collateral Standard: Electromagnetic Compatibility Requirements and Tests, Geneva, Switzerland, [7] D. Adler, L. Margulies, Y. Mahler, and A. Israeli, Measurements of electromagnetic fields radiated from communication equipment and of environmental electromagnetic noise: Impact on the use of communication equipment within the hospital, Biomed. Instrumentation Technol., vol. 32, no. 6, pp , Nov./Dec [8] H. Grant, Managing electromagnetic compatibility between wireless and medical devices, Compliance Engineering European Edition, pp , May/Jun [9] E. Hanada, Y. Watanabe, Y. Antoku, Y. Kenjo, H. Nutahara, and Y. Nose, Hospital construction materials: Poor shielding capacity with respect to signals transmitted by mobile telephones, Biomed. Instrumentation Technol., vol. 32, no. 5, pp , Sep./Oct

8 2388 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 53, NO. 8, AUGUST 2005 [10] D. Davis, B. Segal, C. W. Trueman, R. Calzadilla, and T. Pavlasek, Measurement of indoor propagation at 850 MHz and 1.9 GHz in hospital corridors, in Proc. IEEE APS Conf. Antennas and Propagation for Wireless Communications, 2000, pp [11] T. M. Schäfer, J. Maurer, and W. Wiesbeck, Measurement and simulation of radio wave propagation in hospitals, in Proc. IEEE 56th Vehicular Technology Conf., Vancouver, BC, Canada, Sep. 2002, pp [12], Experimental characterization of radio wave propagation in hospitals, IEEE Trans. Electromagn. Compat., vol. 47, no. 2, pp , May [13] T. M. Schäfer, Experimental and Simulative Analysation of the Ra-Diowave Propagation in Hospitals (in German). Karlsruhe, Germany: University of Karlsruhe, Institut fur Hochstfrequenztechnik und Elektronik, 2003 [Online] Available: [14] Panels for Facade Cladding and Interiors (2004). [Online]. Available: [15] J. Haala, Analysing Microwave Heating Processes by Self-Consistent Finite Integration Techniques (in German). Karlsruhe, Germany: University of Karlsruhe, Institut fur Hochstfrequenztechnik und Elektronik, [16] J. van Bladel, Field penetration through small apertures: The first-order correction, Radio Sci., vol. 14, no. 2, pp , Mar./Apr [17] J. von Hagen, Wide Band Electromagnetic Coupling to a Cavity: An Integral Representation Based Model (in German). Karlsruhe, Germany: University of Karlsruhe, Institut fur Hochstfrequenztechnik und Elektronik, [18] Y. Wang, S. Safavi-Naeini, and S. K. Chaudhuri, A hybrid technique based on combining ray tracing and FDTD methods for site-specific modeling of indoor radio wave propagation, IEEE Trans. Antennas Propag., vol. 48, no. 5, pp , May [19] D. J. Cichon, T. Zwick, and J. Lätheenmäki, Ray optical indoor modeling in multi-floored buildings: Simulations and measurements, in Proc. IEEE Int. Antennas Propagation Symp. AP-S 95, Newport Beach, CA, Jul , 1995, pp [20] A. von Hippel, Dielectric Materials and Applications. Boston, MA: Artech House, [21] M. N. Afsar and K. J. Button, Millimeter-wave dielectric properties of materials, Infrared and Millimeter Waves, vol. 12, pp. 1 2, Werner Wiesbeck (SM 87 F 94) received the Dipl.- Ing. (M.S.E.E.) and the Dr.-Ing. (Ph.D.E.E.) degrees from the Technical University of Munich, Munich, Germany, in 1969 and 1972, respectively. From 1972 to 1983, he was with AEG-Telefunken in various positions including the Head of Research and Development, Microwave Division, Flensburg, Germany, and Marketing Director in the Receiver and Direction Finder Division, Ulm. During this period he had product responsibility for millimeter-wave radars, receivers, direction finders and electronic warfare systems. Since 1983, he has been Director of the Institut für Höchstfrequenztechnik und Elektronik (IHE), University of Karlsruhe, Karlsruhe, Germany, where he is presently Dean of the Faculty of Electrical Engineering. In 1989 and 1994, respectively, he spent a six month sabbatical at the Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA. He serves as a Permanent Lecturer for radar system engineering and for wave propagation for the Carl Cranz Series for Scientific Education. He is a Member of an Advisory Committee of the EU-Joint Research Centre (Ispra/Italy), and he is an Advisor to the German Research Council (DFG), the Federal German Ministry for Research and to industry in Germany. His research topics include radar, remote sensing, wave propagation and antennas. Dr. Wiesbeck has received a number of awards including the IEEE Millennium Medal, the IEEE GRS Distinguished Achievement Award and an Honorary Doctorate (Dr.h.c.) from the University of Budapest. Since 2002, he has been a Member of the Heidelberger Akademie der Wissenschaften. He was a Member of the IEEE GRS-S AdCom from 1992 to 2000, Chairman of the GRS-S Awards Committee from 1994 to 1998, Executive Vice President IEEE GRS-S from 1998 to 1999, President IEEE GRS-S from 2000 to 2002, and past and present Treasurer of the IEEE German Section ( , ). He was an Associate Editor IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION from 1996 to He has been General Chairman of the 1988 Heinrich Hertz Centennial Symposium, the 93 Conference on Microwaves and Optics (MIOP 93), and has been the Technical Chairman of International mm-wave and Infrared Conference He has been a Member of the scientific committees of many conferences. Thomas M. Schäfer (S 99) was born in Karlsruhe, Karlsruhe, Germany, in He received the Dipl.-Phys. (M.S.Phys.) degree in physics from the University of Karlsruhe (TH) and the Dr.-Ing. (Ph.D.E.E.) degree in electrical engineering from the Institut für Höchstfrequenztechnik und Elektronik (IHE), University of Karlsruhe (TH), in 1999 and 2003, respectively. His work includes the investigation of EMC in hospitals, the measurement and simulation of radio wave propagation, especially in indoor environments like buildings and in tunnels, as well as mobile-to-mobile communications between vehicles. He is participating as an expert in the European COST 273 Toward Mobile Broadband Multimedia Networks. For the Carl Cranz Series for scientific education, he serves as a lecturer for EM wave propagation and radio network planning.