IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, VOL. 11, 2012
693
Switched Parasitic Dipole Antenna Array for High-Data-Rate Body-Worn Wireless Applications Md. Rashidul Islam, Student Member, IEEE, Nowrin H. Chamok, and Mohammod Ali, Senior Member, IEEE
Abstract—A switched parasitic dipole antenna array is introduced for 2.45-GHz body-worn wireless applications. The array consists of a driven folded dipole and four parasitic dipoles whose excitation currents are controlled using p-i-n diode switches. Radiation beam can be steered in the azimuth plane using selective switching of the parasitic dipoles. An experimental prototype array was fabricated and tested to demonstrate performance both in free-space and next to the human body. Measured (dB) data clearly demonstrate that the array meets the 2.45-GHz ISM band bandwidth requirements when operated on or near the body. Simulation results of the array against an anatomical human body model show that the array beam can be scanned within an angular region of 65 with 6.9–8 dBi of peak gain. Significant improvement in radiation efficiency and considerably lower specific absorption rate (SAR) are achieved with this array. Measured radiated field strength test shows that the array performance is 2–5 dB superior to a single folded dipole for body-worn applications. Index Terms—Beam scanning, body-worn antenna, dipole array, parasitic array, specific absorption rate (SAR), temperature rise.
I. INTRODUCTION NTENNA design for body-worn communications is challenging because the body being a large lossy dielectric material deteriorates the antenna performance. For high-power devices, the issue of electromagnetic energy absorption by the body and subsequent gain degradation are both to consider while designing such antennas. Wearable antennas for law enforcement personnel, soldiers, firefighters, first responders, and astronauts must be robust enough that they will allow the necessary data throughput for voice and video communication in the presence of fading and interference. Clearly, a beam-steering antenna array will be an excellent choice because in the presence of strategic obstructions and fading the likelihood of a beam-steering array in allowing successful communication is significantly higher. In order to achieve higher data rate and distributed coverage around the body, a 2.45-GHz eight-element phased array was proposed in [1]. Phased arrays require large antenna-to-antenna separations, delay lines, and switches, which make them space-hungry, costly, and difficult to realize in practice. Alternately, a body-worn multiple-input–multiple-output (MIMO) antenna system operating at 2.4 GHz was proposed for higher data throughput [2]. In a MIMO system, strong mutual coupling between antenna elements is a critical challenge that leads to strong channel correlation and degradation in the mean
A
Manuscript received April 28, 2012; accepted June 07, 2012. Date of publication June 15, 2012; date of current version July 06, 2012. The authors are with the University of South Carolina, Columbia, SC 29208 USA (e-mail:
[email protected]). Color versions of one or more of the figures in this letter are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/LAWP.2012.2204949
Fig. 1. Proposed switched parasitic antenna array. (a) Detailed dimensions. (b) Top view. (c) 3-D view for placement in a body-worn application and a p-i-n , nH, diode switching control circuit. Circuit parameters: pF. PIN diode: SMP 1345 (Skyworks, Inc.).
effective gain (MEG) [2]. A switched parasitic antenna array can be a good alternative that does not require large space and provides reasonably good beam steering without the use of any delay lines. In a switched parasitic antenna array, element-to-element mutual coupling is exploited and used to achieve beam steering and high antenna gain. Previously, microstrip patch antennas were reported for nonswitched fixed-beam wearable applications [3]–[6]. Although microstrip patch antennas have been proposed and used for switched parasitic array design, integrating and implementing them for body-worn applications is difficult because: 1) patches are larger in length and width; and 2) they require relatively large ground planes to achieve directional beams. These constraints will require that both the antenna and the ground plane surfaces are continuous. Conversely, if a switched parasitic beam-steering array can be developed where the array elements are made of narrow strips
1536-1225/$31.00 © 2012 IEEE
694
Fig. 2. Computed antenna gain (
IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, VOL. 11, 2012
) at 2.45 GHz in the azimuth plane (
-plane) for parametric variation of (a)
representing dipoles, it would be much easier to implement that for wearable applications. This is the objective of this letter. Here, we propose a 2.45-GHz switched parasitic antenna array comprised of a driven folded dipole and four parasitic dipoles that are nearby and are controlled using p-i-n diode switches. Our proposed array offers nearly the same directive gain as that of the fixed-beam microstrip patch antennas reported in [3] and [4] in addition to its beam-scanning feature. Although a switched parasitic array composed of 24 V-shaped bent wires was proposed for beam steering in the azimuth plane in free space and at 30 intervals [7], the 3-D array is not suitable for body-worn applications. Thus, in this letter we introduce a wearable antenna array consisting of V-shaped dipoles in planar arrangements operating at 2.45 GHz. Preliminary work on this concept was presented in [8]. This letter uses that concept, but conducts a more in-depth analysis and provides additional numerical and experimental data both from antenna performance characteristics and electromagnetic exposure points of view. II. PROPOSED ANTENNA GEOMETRY The switched-beam body-worn array is designed for operation in the 2.45-GHz ISM band. The proposed array has a single folded-strip dipole as the driven element [Fig. 1(a)]. There are four V-shaped parasitic dipoles that are turned ON or OFF using p-i-n diode (SMP 1345, Skyworks, Inc.) switches. Two of the four V-shaped parasitic dipoles are in the same plane as the driven element, thus defining Layer 1. The other two V-shaped parasitic dipoles are in Layer 2. Dipoles in Layer 2 are 10 mm away from Layer 1 [Fig. 1(b)]. Fig. 1(b) and (c) shows the proposed placement of the array for body-worn applications where Layer 1 elements are near the human body separated by a distance . By turning the switches ( , , , and ) ON and OFF, it is possible to steer the radiation beam of the array. The switches control the magnitudes and phases of the induced currents on the four parasitic dipoles. The parasitic dipoles take the shape of the letter V to allow for better impedance matching. The lengths of the driven and parasitic dipoles are close to at 2.45 GHz, where is the free-space wavelength. The isolated input impedance of the V-dipoles is capacitive at 2.45 GHz. This makes them act as directors [9]. However, in
when
is ON and (b)
when
Fig. 3. Computed phase difference (in degrees) between (a) and . (b)
and
is ON.
and
and [Fig. 1(a)] determine whether an array, the distances the parasitic dipoles will act as directors or reflectors. To finalize the distances and , the array was simulated in free space using HFSS (ANSYS, Inc.) . The OFF state of each switch was modeled using a small 0.15-pF capacitor, while the ON state was modeled using a relatively large (10 pF) capacitor. The latter was done in order to represent the presence of two dc blocking capacitors (20 pF each) that were connected in series with the switch [Fig. 1(c)]. Considering only the elements in Layer 1 being present, distance was varied from 1 to 15 mm (0.008 to ). Considering ON and OFF, the simulated gain patterns ) are shown in Fig. 2(a). As apin the azimuth plane ( parent, for distance from 1 to 9 mm, parasitic dipole 1 acts as a director; the beam points toward itself, i.e., . For mm, parasitic dipole 1 acts as a reflector; the beam points in the opposite direction, i.e., . For mm, the phase of the currents in parasitic dipole 1, , starts to lead the phase of the currents in the driven element, [Fig. 3(a)], which makes parasitic dipole 1 act as a reflector. The phases of these currents were simulated following the method reported in [10]. Considering similar cases where only the elements in Layer 2 were present, gain patterns were computed [see Fig. 2(b), which shows the case for ON and OFF]. From Fig. 2(b), it is clear that parasitic dipole 3 acts as a director or a reflector based on its distance from the driven dipole. The phase of the current in parasitic dipole 3, , leads when mm, which makes dipole 3 a reflector [Fig. 3(b)]. The same dipole functions as a director for mm ( lags ).
ISLAM et al.: SWITCHED PARASITIC DIPOLE ANTENNA ARRAY FOR HIGH-DATA-RATE BODY-WORN WIRELESS APPLICATIONS
695
Fig. 5. Simulated normalized radiation patterns of the array next to an anatomical human body model as function of various switching states of the array. Fig. 4. Photographs of the fabricated antenna array and measured free-space (dB) versus frequency data with switch states as the parameter. Aninput , . tenna dimensions are given in Fig. 1,
III. MEASURED FREE SPACE INPUT
PERFORMANCE
Based on simulations of input and radiation patterns, mm and mm were selected. The proposed antenna array was fabricated on foam. The p-i-n diode (SMP 1345, -Skyworks, Inc.) switches were fabricated on taconic substrates, and dc bias wires were attached (see photographs shown in Fig. 4). Measured (dB) data versus frequency with the switch states as parameters are shown in Fig. 4. As apparent, the array operates in the vicinity of 2.45 GHz for a whole host of switching conditions. Considering the dB as the criterion, the array free-space bandwidth is about 150 MHz, which is much larger than the ISM-band bandwidth requirement (85 MHz). Note that by “ ON,” it is meant that all other . switches are OFF except IV. SIMULATED ARRAY PERFORMANCE NEXT HUMAN BODY MODEL
TO A
Since the primary goal of the antenna array is for it to be used next to a human body, its performance was simulated against an anatomically correct human body model. For array performance, specific absorption rate (SAR) and temperature rise calculations XFDTD (Remcom, Inc.) was used. The body model used was the Duke model. A detailed description of the model can be found in our earlier work [8]. Array Layer 1 was placed near the body at a distance of 25 mm from the body, while array Layer 2 was placed 10 mm above of array Layer 1. Thus, array Layer 2 was facing the outside. In XFDTD, the lengths of the parasitic dipoles were reduced from 28 to 26 mm to compensate the rectilinear representation of inclined (i.e., 45 ) dipole arms. As before, in XFDTD simulations, the switches were modeled as they were modeled in HFSS. Simulated radiation patterns of the proposed array under different switching conditions are shown in Fig. 5. Note that, the beam-scanning angle of 90 seen in Fig. 2(a) due to parasitic dipole 1 is not present in Fig. 5. The array scans the beam from 330 to 35 . The scanning angular range is about 65 , which is narrower than the corresponding scanning angular range in free space. This occurs because of the blockage presented by the human body. The array peak gains for all five cases shown in Fig. 5 are between 6.9–8 dBi (ignoring impedance mismatch). Radiation efficiencies for the five cases were between 67%–87%. The highest efficiency occurred when both switches
TABLE I AT 2.45 GHZ DUE TO THE PROPOSED ARRAY AND SAR AND PEAK A DRIVEN FOLDED DIPOLE. SEPARATION BETWEEN THE BODY AND mm THE ARRAY LAYER 1,
and were ON. For comparison, the radiation pattern and efficiency of a folded dipole [Fig. 1(a)] without any V-shaped parasitic dipoles were also computed. The efficiency was 68%, and the peak gain was 6 dBi for the folded dipole at the same distance from the body as Layer 1 of the array was. SAR and peak temperature rise, , induced by the proposed array were computed for the same separation between Layer 1 and the body, i.e., 25 mm. Computed peak 1-g and 10-g average SAR and peak normalized to 1 W of power are listed in Table I. Since XFDTD ignores the effects of impedance mismatch, the SAR and data represent worst-case scenarios. Comparing the SAR induced by the array and that induced by a single folded driven dipole (last row in Table I), SAR is higher when or is ON due to stronger electromagnetic radiations in the direction of the body. By contrast, SAR reduces when either or both or are ON. Specifically for the case when and are both ON, both 1-g and 10-g SAR decreases by 74% compared to that induced by a driven folded dipole alone. Also from Table I, the induced by the array is only about 27% of that induced by the driven folded dipole (0.09 compared to 0.33 ). Peak 10-g SAR and radiation efficiency of the array ( and both ON) and a single folded dipole were studied using as XFDTD simulations for different separation distances listed in Table II. It is clear that the array performs better than a single folded dipole for any of the distances listed in Table II. When the distance was reduced to 2 mm from the body, the array radiation efficiency was 4.4 dB, and the peak 10-g SAR was 10 W/kg. This very short-distance scenario is unlikely to be the case for vest- or jacket-installed antennas for firefighters or law enforcement personnel.
696
IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, VOL. 11, 2012
Next, the array was placed against the body of the volunteer, and received signal strengths were measured for the array on the body, at a distance of 25 mm from the body and at a distance of 50 mm from the body when both and were ON and OFF. The array received signal strength was 2 dB higher when on the body and 25 mm from the body, and 5 dB higher when 50 mm from the body. These figures are comparable to those shown in Table II. VI. CONCLUSION
Fig. 6. Measured (dB) versus frequency with various switch configuramm and (b) mm. is the distance tions as parameters for (a) between Layer 1 and the human body. TABLE II PEAK 10-g SAR AND RADIATION EFFICIENCY, , OF THE PROPOSED ARRAY ( AND BOTH ON) AND A SINGLE FOLDED DIPOLE FOR DIFFERENT SEPARATION DISTANCES FROM THE BODY
V. MEASURED ARRAY PERFORMANCE NEXT TO HUMAN BODY The fabricated antenna array was placed next to the body of a volunteer on the chest area. The volunteer was only wearing a very thin T-shirt. The volunteer was a 6-ft male. Since the transmit power from the vector network analyzer (VNA) was only 10 dBm, there was no issue of electromagnetic exposure. Two cases were considered: 1) array Layer 1 was touching ( mm) the volunteer’s T-shirt; and 2) array Layer 1 was spaced 25 mm ( mm) from the volunteer’s T-shirt. The 25-mm space was created by placing a block of foam. Referring to the photographs shown in Fig. 4, Layer 1 was facing the body, while the balun was away from the body. Measured input data for these two cases are shown in Fig. 6. It is clear that both for on body and 25 mm away from the body, the array performs well within the ISM frequency band. The array was measured for radiated field strength inside the Microwave Engineering Laboratory, University of South Carolina, Columbia. A transmit broadband 400 MHz–4 GHz log-periodic dipole array (vertically polarized) with 4-dBi gain was placed at a distance of 2.56 m from the receive switched parasitic array. The transmit array was connected to the signal source, while the receive array was connected to a spectrum analyzer. Measurements were conducted primarily at 2.45 GHz by doing a peak search within a narrow frequency band. Measurements were conducted when both and were ON and and OFF). for a single folded dipole (basically both In either case, the beam direction was broadside to the array plane. Measurements were performed both in free space and in body-worn conditions. In free-space, the array received signal strength was 5 dB higher (with and both ON) compared to the single folded dipole. Overall, signal strength difference within the frequency range of 2.35–2.5 GHz was between 2–5 dB between the array and the single folded dipole.
A switched parasitic antenna array is proposed for beam steering in the azimuth plane. It is shown that by switching a number of parasitic elements, the array beam can be steered in different directions. Measured data in free space as well as next to a human body show that the array can operate in the 2.45-GHz ISM frequency band. Finite-difference time-domain (FDTD) simulations of the array against an anatomical human body model show that the array beam can be steered within an angular region of 65 and with peak gain of 6.9–8 dBi. The array has the advantage of significantly lowering the SAR and temperature rise for a number of switching configurations. Performances at different distances and very close to the body show that the array retains its gain advantage over a single folded dipole antenna. Measured signal strengths on or near the body show that the array signal strength improves from 2–5 dB depending on its distance from the body. Using conductive fabrics/threads, this array can be implemented in the vest/garments of a wearer, ensuring ease of movement. ACKNOWLEDGMENT The authors would like to thank N. Alam for assisting with some of the measurements. REFERENCES [1] T. F. Kennedy, P. W. Fink, A. W. Chu, N. J. Champagne, II, G. Y. Lin, and M. A. Khayat, “Body-worn E-textile antennas: The good, the low-mass, and the conformal,” IEEE Trans. Antennas Propag., vol. 57, no. 4, pp. 910–918, Apr. 2009. [2] Y. Ouyang, D. J. Love, and W. J. Chappell, “Body-worn distributed MIMO system,” IEEE Trans. Veh. Technol., vol. 58, no. 4, pp. 1752–1765, May 2009. [3] C. Hertleer, H. Rogier, L. Vallozzi, and L. V. Langenhove, “A textile antenna for off-body communication integrated into protective clothing for firefighters,” IEEE Trans. Antennas Propag., vol. 57, no. 4, pp. 919–925, Apr. 2009. [4] A. Tronquo, H. Rogier, C. Hertleer, M. Moeneclaey, and L. V. Langenhove, “Robust planar textile antenna for wireless body LANs operating in 2.45 GHz ISM band,” Electron. Lett., vol. 42, no. 3, pp. 142–143, Feb. 2006. [5] F. Declercq and H. Rogier, “Active integrated wearable textile antenna with optimized noise characteristics,” IEEE Trans. Antennas Propag., vol. 58, no. 9, pp. 3050–3054, Sep. 2010. [6] E. K. Kaivanto, M. Berg, E. Salonen, and P. de Maagt, “Wearable circularly polarized antenna for personal satellite communication and navigation,” IEEE Trans. Antennas Propag., vol. 59, no. 12, pp. 4490–4496, Dec. 2011. [7] A. Sutinjo, M. Okoniewski, and R. H. Johnston, “An octave band switch parasitic beam-steering array,” IEEE Antennas Wireless Propag. Lett., vol. 6, pp. 211–214, 2007. [8] M. R. Islam and M. Ali, “A novel wearable antenna array for 2.45 GHz WLAN application,” in Proc. IEEE Antennas Propag. Soc. Int. Symp., Spokane, WA, Jul. 2011, pp. 2754–2757. [9] R. S. Elliot, Antenna Theory and Design, ser. Electromagnetic Wave Theory. Piscataway, NJ: IEEE Press, 2003. [10] M. R. Islam and M. Ali, “Ground current modifications of mobile antennas and its effects,” IEEE Antennas Wireless Propag. Lett., vol. 10, pp. 438–441, 2011.
