printed dipole antenna for energy harvesting is presented. A very small reflector is placed behind the antenna to increase its gain. This antenna can be used in ...
Design and Optimization of a GSM Printed Dipole Antenna for Energy Harvesting Applications Ali Alaeldine, Mohamed Latrach, Hedi Raggad and Zaher Sayegh ESEO - Radio & Microwave Team 4, rue Merlet-de-la-Boulaye, 49009 Angers - France Email: http://perso.eseo.fr/∼aalaeldine/ Telephone: (33) 241-86-67-03 Abstract—In this paper the design and optimization of a GSM printed dipole antenna for energy harvesting is presented. A very small reflector is placed behind the antenna to increase its gain. This antenna can be used in indoor or outdoor energy harvesting applications. An interesting characteristic is its planar structure which allowing an easy fabrication with low cost. Simulations and measurements have been carried out using HFSS simulator tool and an anechoic chamber to examine the antenna characteristics as gain and radiation patterns. Comparisons between measurements and simulations are done in order to validate the suggested design.
I. I NTRODUCTION Within the recent years, micro devices, including ultra low power consumption wireless sensors, have become to be inherently emerged for controlling of structural health. The most concerned domains are aerospace, bridges, high speed train rails where the manual non-destructive testing is relatively expensive and requires many time consuming. Wireless sensors can be used also to predict earth’s interior quiz, volcano eruption and to verify the structural health of skyscraper, nuclear reactors and missiles. Therefore, the fast development of electronic systems of low power consumption associated with sensors placed in structures difficult to reach pushes the engineers to find an efficient power supply solution. The objective of engineers also is to reduce the size and weight of these systems. In general, systems are powered by batteries that occupy 90% of the device volume. Improving the energy density (and other features such as cost, number of charging cycles, and power density) of batteries has been, and continues to be, a major research field, however, alternative methods of powering the devices that make up the wireless networks are desperately needed. The possible approaches to this challenge is to develop technologies that enable a node to generate or harvest its own power. The cost of procuring, storing and getting someone to change a battery can easily cost as much as an energy harvester. Depending on their type, harvesters utilize different energies such as solar, motion and vibration, temperature difference and ambient energy [1] [2]. The efficiency of these methods depends on the existence of an enough mechanical vibration or temperature variation. This dependence makes the control of harvesting systems so complex. The other source for energy harvesters that is being considered is use of RF signals. In RF power Harvesting, the RF harvester can harvest energy from propagating RF
978-1-4244-9561-0/11/$26.00 ©2011 IEEE
waves which can be ambient RF waves generated by nearby electronic components such as cellular phones, or generated from a dedicated RF power sources. The radio frequency radiation energy represents the most common mode of distributing power to embedded electronics [3]. The power that can be achieved from them could be high enough to run a low power sensor or a low power circuit. For RF power harvesting, the power that can be transmitted by an RF power source through an antenna is limited by EEC (Europe) and FCC (USA) regulations [4]. Power induced at GSM frequency can be relatively high to be used to power an ultra low consumption application. In free space, the path loss of the RF signal is given by Frii’s equation [5] as: 2 4πR (1) LP = λ Where R is the distance between the power source and power harvester, and λ is the wavelength of the RF signal transmitted from the power source. To reach the best results, the design and conception of small, omnidirectional, low cost and wide band antennas with an important gain are becoming necessary. Printed dipole antennas on low-cost substrates like FR4 are appropriate candidates since, in addition to their low costs; they can provide a large bandwidth and can also present better pure polarization in comparison with other structures [6]. Besides, the printed dipole antenna is a good candidate for phase arrays antenna with large number of elements, since they can easily be integrated [7] [8]. This paper presents the design and optimization of a GSM dipole printed antenna 915 MHz with large bandwidth and high gain modeled on low-cost FR4. Sect. II describes the antenna structure. Sect. III discusses the antenna gain simulation (HFSS) and measurement (anechoic chamber) using the three-antenna-method. In this section we discuss also the integration of this antenna in a power harvesting application. Finally, conclusions and perspectives are summarized in Sect. IV. II. A NTENNA STRUCTURE AND OPTIMIZATION In this section, the structure of a GSM printed dipole antenna (Fig. 1) using for RF energy harvesting is discussed. The harvester is used to power an ultra power consumption
1780
AP-S/URSI 2011
cient can be observed as shown in Fig. 2.
Fig. 2. cient
Fig. 1. Suggested GSM printed dipole antenna, before optimization (top) and after optimization (bottom)
system. The suggested antenna is printed in the top layer of the FR4 substrate. The thickness of this substrate is 1.6 mm and its relative permittivity r = 4.5. the input impedance of the antenna should be 50 Ω. A SMA connector is connected in its input. The antenna is designed and simulated using Ansoft-HFSS 12.1. The initial dimension if the first antenna version (top side of the Fig. 1) is 68 mm for each side. Its width is 18 mm. Some geometric optimizations has been entered into the antenna structure (bottom side of the Fig. 1). The objective is to reduce the antenna size, to increase the impedance adaptation (50Ω), to center the operational frequency and to increase the gain. Tab. II shows the simulated antenna gain and reflection coefficient at 915 MHz before and after optimization. Antenna Simulation (before optimization) Simulation (after optimization)
Frequency 915 MHz
S11 -16.82
Gain (dBi) 2.37
915 MHz
-19.72
2.49
Comparison between simulated and measured reflection coeffi-
The measurement has shown also the wide operation band of the antenna. Its going from 840 MHz to 1 GHz and represents 22% for ||S11 || ≤ 10dB. According to Pozar [9], the far-field pattern of a half-wave dipole is: cos( π2 cos(θ)) e−jk0 r , EΦ = 0 (2) Eθ = V 0 sin(θ) r In order to be able to calculate the radiation intensity of the far-field pattern, the Eq. 2 is simplified: e−jk0 r (3) r The radiation pattern can be plotted by using the Eq. 3 for the radiation intensity. Fig. 3 shows the simulated radiation pattern of the optimized antenna at 915 MHz at magnetic plane H (Φ = 90◦ ) and electric plane E (Φ = 0◦ ). These plots show that the half wave dipole antenna is a very omnidirectional and therefore well suited for radio frequency energy harvesting. Eθ ≈ V0 sin(θ)
TABLE I Comparison between antenna simulated gain and reflection coefficient before and after optimization
In the internal face of the optimized antenna, we have obtained a geometric form proportional to its length (antenna shapes). The increasing or decreasing of antenna length plays on its frequency. However, the resizing of antenna shapes allows the obtaining of better impedance adaptation and increases the gain for each selected frequency. To measure its adaptation, a vector network analyzer is used. An encouraging correlation between measured and simulated reflection coeffi-
Fig. 3. cient
Comparison between simulated and measured reflection coeffi-
In order to increase the gain, mini-reflector (20 mm of large)
1781
is placed behind the antenna at a distance of λ/4 (75 mm) (Fig. 4).
Fig. 4.
Optimized antenna with a mini-reflector placed behind
Fig. 6.
A significant gain improvement has been observed. Tab. IV shows the antenna gain with and without reflector. Antenna Without reflector With reflector
Frequency 915 MHz 915 MHz
Gain (dBi) 2.49 6.28
TABLE II Comparison between antenna simulated gain and reflection coefficient before and after optimization
Fig. 5 and Fig. 6 represent the simulated antenna gain with and without reflector.
Simulated gain of the antenna with reflector at 915 MHz
have a standard environment surrounding the antenna. Ideally, measurements should be made with the measured antenna so far removed from any objects causing interferences that it can be considered in open space. This is an impractical situation. Professional laboratories use electromagnetic anechoic chambers that simulate almost perfectly the open space situation. These are very expensive and requires many time and space for set-up. In this section, the gain and radiation patterns of a small 915 MHz printed dipole antenna are measured in a anechoic chamber. However, to achieve gain measurement, three antennas are required: • •
The antenna under test Two antennas of unknown gains
Three measurements are required to determine the gain of the antenna under test. In each measurement, the forward transmission coefficient S21 is obtained (Eq. 4). Pr Si,j (dB) = (dB) (4) Pe i,j
Fig. 5.
Simulated gain of the antenna without reflector at 915 MHz
III. A NTENNA GAIN MEASUREMENT USING THREE - ANTENNA METHOD Several studies have demonstrated that the best method to measure the antenna gain is this of three-antenna. The advantage of this method is that no reference antenna is needed [10]. The gain of three used antennas are unknowns. To be coherent in comparing different antennas, it is necessary to
Measurement procedure is doing as follows. In the first step, one antenna with unknown gain is connected to the input S11 of the vector network analyzer (VNA) HP and represents the transmitter. In the other side, the antenna under test is connected to the input S11 of the VNA and represents the receiver (Fig. 7). The distance R between the antennas is 2.5 meters (far-field). In each measurement, a relation between both used antennas gains is obtained (Eq. 5). 4πR (5) Gi (dB) + Gj (dB) = Si,j (dB) + 20log λ where Si,j is the forward transmission coefficient as indicated in the Eq. 1, Gi and Gj are the gains of the transmitter and receiver antennas and λ is the wave length. However, the three used antennas used in our case are: the dipole antenna under test, a horn antenna (frequency range of 700 MHz to 18 GHz) and a patch antenna. Three equations
1782
Antenna Realized Dipole Patch Horn
Measured gain (dBi) 1.94 4.05 7.6
Given gain (dBi) 4.1 6.8
TABLE IV M EASURED ANTENNA GAIN USING THE THREE - ANTENNA METHOD
Antenna Without reflector With reflector
Simulated gain (dBi) 2.49 6.28
Measured gain (dBi) 1.94 4.7
TABLE V S IMULATED AND MEASURED ANTENNA GAIN WITH AND WITHOUT REFLECTOR
Fig. 7.
An anechoic chamber set-up
IV. C ONCLUSION
with three unknowns gains values are obtained: GD (dB)+GH (dB) = S(DH)2,1 (dB)+20log GD (dB) + GP (dB) = S(DP )2,1 (dB) + 20log GP (dB) + GH (dB) = S(P H)2,1 (dB) + 20log
4πR λ 4πR λ 4πR λ
(6) (7) (8)
The value of λ is given by the equation 9: λ=
c 3 ∗ 108 = = 0, 328m f 915 ∗ 106
(9)
Tab. III shows transmission coefficient values measured in the anechoic chamber. These values includes measured losses caused by cables using for transmission and receiving (9 dB). Antennas Dipole-Horn (DH) Dipole-Patch (DP) Patch-Horn (PH)
In this paper, the design and optimization of a GSM printed dipole antenna for energy harvesting is described. A good correlation between simulated and measured reflection coefficient shows a good impedance adaptation of 50 Ω. Also the simulated gain and radiation pattern of the optimized antenna with and without reflector have been demonstrated using Ansoft HFSS. Antennas gain have been measured in an anechoic chamber using the three antenna method. Results show a interesting antenna gain for energy harvesting application. Integration of this antenna in a RF energy harvester to power a very low power consumption wireless sensor is in progress.
S2,1 (dB) -30,13 -33,68 -28,02
TABLE III F ORWARD TRANSMISSION COEFFICIENTS
Therefore, we have to insert obtained values of S2,1 into Eq. 6, Eq. 7 and Eq. 9. Three equations with three unknowns are formed. Gain results are shown in Tab. IV. A good correlation between obtained results and those supplied by the manufactured of the two used antennas confirm the validity of our measurements. An encouraged correlation between the measured and simulated gain of the suggested antenna is also observed (Tab. V). We are following the same procedure to measure the gain of the antenna with the mini-reflector. Results are showing in Tab. V.
R EFERENCES [1] S. Priya and D. Inman. Energy Harvesting Technologies. Springer, second edition, 2009. [2] J. P. Thomas, M. A. Qidwai, and J. C. Kellogg. Energy scavenging for small-scale unmanned systems. Journal of Power Sources, 159:1494– 1509, February 2006. [3] S. J. Roundy. Energy Scavering for Wireless Sensor Nodes with a Focus on Vibration to Electricity Conversion. PhD thesis, The University of California, Berkeley, 2003. [4] Federal Communications Comission. Fcc codes of regulation part 15. http://www.access.gpo.gov/nara/cfr/waisidx_03/47cfr15_03/htm. [5] H.T. Frii’s. A note on a simple transmission formula. In proceedings of the institute of Radio Engineers and Waves and Electron, pages 254– 256, May 1946. [6] K. Fujimoto. Mobile Antenna System Handbook. Artech House Publishers, Boston, MA, second edition, 2001. [7] J. R. Bayard, M. E. Cooley, and D. H. Schaubert. Analysis of infinite arrays of printed dipoles on dielectric sheet perpendicular to a ground plane. IEEE Transactions on Antennas Propagation, 39:1722–1732, December 1991. [8] A. J. Parfitt, D.W. Griffin, and P. H. Cole. Analysis of infinite arrays of substrate-supported metal strip antennas. IEEE Transactions on Antennas Propagation, 41:191–199, February 1993. [9] David M. Pozar. Microwave and RF design of Wireless systems. John Wiley and sons, Inc., first edition, 2001. [10] Hsin-Chia Lu and Tah-Hsiung Chu. Antenna gain and scattering measurement using reflective three-antenna method. In IEEE International Symposium on Antennas and Propagation, pages 374–377, Jully 1999.
1783
