I. INTRODUCTION
Rapid progression is done in the field of wireless communication and radio frequency electronics with the benefit of meeting the requirement of miniaturization. Compact planar antennas are very promising candidates in satisfying the design consideration. Microstrip patch antenna covering more than one wireless communication band is considered a dual/multiband [Reference Chung, Jeon, Kim, Ahn, Choi and Itoh1–Reference Wong3]. Challenge is to obtain enhanced bandwidth and better gain with the antenna designed to operate at dual bands. It is noted that majority of the earlier literature with different shaped slots and slits printed on the patches have drawn less attention due to less gain at low resonant frequencies. The use of defected ground structure (DGS) was explored earlier to miniaturize printed circuits, and this concept was adapted to the problem of antenna designs [Reference Mandal and Sanyal4]. Metallic ground plane may be considered a dominant portion of the radiating structure [Reference Liu, Li and Sun5]. Self-similar structure for the fractal shape and multiple scales of recurring geometry benefit the antenna to resonate at different frequency bands. Variety of simple DGS slots such as dumbbell shaped, arrow shaped dumbbell, circular shaped, and spiral shaped are available in the literature which increases the effective length on the ground plane [Reference Breed6–Reference Lim, Ahn, Han, Jeong and Liu8]. Efficient size reduction of low-pass filters and broad bandwidth with patches is discussed using sierpinski fractal DGS [Reference Hwang9].
Several techniques have been demonstrated using fractal elements and elevation of patches which are discussed in the available literature. Etching of a defect in the ground plane of patch antenna disturbs the current distribution and gives rise to increasing effective capacitance and inductance [Reference Weng, Guo, Shi and Chen10]. Hilbert fractal-shaped structures have been proposed to produce printed and microstrip dipole and monopole antennas with compact size and dual-band performance for different applications [Reference Anguera, Puente, Martinez and Rozan11]. Concentric ring-shaped DGS [Reference Guha, Biswas, Biswas, Siddiqui and Antar12] discusses about suppressing the harmonics in microstrip-based active antenna designs. Liu et al. presented a monopole antenna exciting tri-bands with DGS playing a major role [Reference Liu, Wu and Dai13]. Probe-fed broad band antenna with V-slot DGS [Reference Esa, Jamaluddin and Awang14] and Z-type DGS [Reference Kandwal, Sharma and Khah15] for bandwidth enhancement are few latest models developed in current literature. H-shaped DGS slot antennas with CPW fed obtain high bandwidth for WLAN 2.4 GHz band with very low gain [Reference Sujith, Mridula, Binu, Laila, Dinesh and Mohanan16]. Circular ring-shaped DGS proposed for wideband application, but resonates at X-band frequency [Reference Sharma, Kandwal and Khah17]. Polygonal DGS [Reference Farahbakhsh, Mosalanejad, Moradi and Mohanna18] is etched on the ground plane and is used for array applications for better gain value. The brief literature dealt about applying defect on the ground plane for multiple applications in enhancement of bandwidth and frequency shift property.
In the present design, it is observed that the slits truncated across diagonal corners on the radiating area, create additional resonant frequencies. Fractal DGS on the metallic ground would give rise to size reduction with lowering the frequencies. The size reduction property of the proposed antenna may be suitable for miniaturized planar and conformal antenna arrays that are suitable for personal communication devices. The antenna resonates at frequencies of handheld devices. The antenna has to be further modified to be integrated on hand held device. The problem of back radiation due to defect on the ground plane can be reduced with the usage of absorbers without degrading the actual performance of the antenna in the form of bidirectional patterns [Reference Yang and Rahmat-Samii19].
In this paper, Minkowski shaped fractal geometrical element is adopted as optimum choice that allow both miniaturization and dual-band behavior. The optimal feed positions of iterative shapes of proposed antenna are same, which indicates that the feed position is insensitive to the variation in the spur-line length. Owing to the spur-line perturbation, the radiation pattern of the lower operating frequency has a relatively larger cross-polarization component than that of the higher resonant frequencies [Reference Lu and Wong20]. The embedded spur lines are normally placed at non-radiating edges of the patch and the design resonates at three bands. When the spur-line length is greater than about one-half of the patch side length (L 6 > L 1/2), the antenna can have a new resonant mode at a frequency less than the fundamental frequency. Furthermore, this new resonant mode and TM10 dominant mode can both be excited with good impedance matching using single probe feed located with (−8, 8) at point P.
II. ANTENNA DESIGN
The conventional square-shaped patch antenna with single coaxial probe feed is initially considered. Dual L-shaped slits are cut across the diagonal corners on the radiating portion along with minkowski shaped defective element on the metallic ground plane are as shown in Fig. 1. It is printed on a dielectric substrate of 3.175 mm thick with a relative permittivity of 2.33 and loss tangent of 0.0012. The radiating patch is completely covered by 36 × 36 × 3.175 mm3 dimensions on top.
Geometry of the proposed antenna (all dimensions in mm).
The optimized dimensions of the top layer and bottom layer of antenna are displayed in Table 1. The purpose of metallization with fractal DGS on bottom layer is to shift the S 11 response to a lower frequency without increasing the size of the radiating patch. Further, the impedance matching at obtained frequencies also needs to be maintained.
Optimized geometrical dimensions of the proposed structure.
A) Design of DGS
Different techniques have already been used for the antenna size miniaturization such as using the substrate with high dielectric constant, edge shorted patch with shorting plates, slot loading, etc. The etching of a defect in the ground plane is also a unique technique for size reduction. This technique not only reduces the size, but also improves the antenna efficiency at low frequencies. Variety of slot geometries etched in the microstrip ground plane has been reported in the literature [Reference Breed6]. The different geometries of slots are arrowhead slot, H-shaped slot, open loop dumbbell, and fractal slots as shown in Fig. 2.
(a) Arrow head, (b) H-shaped slot, (c) open-loop dumbbell, (d) symmetric fractal shape, and (e) asymmetric fractal shape.
Asymmetric minkowski fractal shaped cut on the metallic ground plane is chosen to increase the impedance bandwidth for dual-band compact. The difference in indentations on either side of the fractal DGS increase route length of current and hence, effective inductance, which gives rise to lower cut-off frequency. Selection of DGS structure placed on ground plane shows the impact of frequency tuning and improvement of the bandwidth and gain at resonant frequencies. The dimensions of the DGS slot are selected as part of design specifications such that modes change the position from higher frequency to lower frequency.
B) Design simulations
The design iterations of the model are as shown in Fig. 3 and the simulations are carried using Hyper Lynx IE3D tool. Representative S 11 (in dB) characteristics with and without the fractal DGS compared with basic patch are presented in Fig. 4, to illustrate the significance of DGS. The ground plane is etched with minkowski fractal DGS configuration in order to achieve dual-band operation and resonates at lower frequencies. The asymmetry (L 2 ≠ L 3) and (L 4 ≠ L6) lengths on the sides of the fractal DGS are tuned to obtain loading reactance values that reduces the resonant frequency of the TMmn modes as desired. These design specifications are introduced to approach excitation of dual resonant modes accompanied with good impedance bandwidths over the operating bands for the proposed structure. The observations of gain and impedance bandwidth values from the simulations performed for all three design iterations are listed in Table 2.
Design iterations: (a) basic patch, (b) truncated L-shaped spur lines without DGS, and (c) truncated corners with Minkowski fractal DGS.
Simulated return loss curves for design iterations.
Simulated results of antenna design iterations.
The bandwidth is narrow operating at a frequency of 3.75 GHz for iteration of Fig. 3(a). Additional multi bands are generated for Fig. 3(b) (without DGS) with bandwidths maintained at all operated resonant frequencies. On the other hand, Fig. 3(c) (with DGS) provides higher bandwidth over generated dual bands at lower frequencies, when compared with previous two designs. The simulated values of S 11 show impedance bandwidth of 50 and 90 MHz for the operating bands of 1.202 and 1.685 GHz, respectively. From the observations, it shows that frequency is lowered from 2.56 to 1.2 GHz for first band and from 3.02 to 1.657 GHz for second band, respectively, for the third design iteration as indicated in Fig. 4. Simulated gain values at these resonant frequencies are listed in Table 2. From the generated text values, it indicates that gain value is compromised, especially at lower frequencies with the applied DGS.
C) Surface current distribution
The relationship between the excited surface waves and the cut off frequency (f c ) representing the existence of TE/TM modes is expressed by,
$${f_c}\, = \,\displaystyle{n \over {4h\sqrt {{\varepsilon _0}{\mu _0}} \sqrt {{\varepsilon _r}{\mu _r} - 1}}},$$
where n = 0, 2, 4, … for TM modes and n = 1, 3, 5, … for TE modes, h, the thickness of the dielectric; ε r and μ r are relative permittivity and permeability of the substrate; ε 0 and μ 0 are free space permittivity and permeability. Owing to high dielectric thickness (h = 3.175) for proposed antenna design, high energy is coupled to surface waves, thus increasing the surface waves. However, this results in increase in the antenna bandwidth and antenna efficiency. To decrease the resonant frequency of the antenna for a given surface area, the current path must be maximized with in the area. The key to reduce the size of the antenna is to maximize the current patch of the printed antenna. It is found that Minkowski fractal DGS with probe feed achieves a significant reduction in size of antenna using frequency shift and also shown that bandwidth is enhanced.
The IE3D simulated surface current distributions at the frequencies of 1.202 and 1.657 GHz are illustrated in Fig. 5. It can be clearly seen that the current distributions are different for dual bands. L-shaped spur lines and fractal DGS modify the current distribution for exciting the antenna at both frequencies. Most of the current flow is at the edges of the patch at 1.202 GHz as given in Fig. 5(a), whereas the surface current at 1.657 GHz is mainly concentrated around asymmetrical boundaries of fractal DGS indicating the presence of resonance as drawn in Fig. 5(b). This implies that for the high resonant band, significant improvement in the bandwidth is observed due to the placement of fractal DGS. Enhanced impedance bandwidth is achieved with dual L-shaped slit antenna loaded with fractal DGS by optimizing the dimensions to adjust its different resonances.
Average and vector current distributions at: (a) 1.202 GHz and (b) 1.657 GHz.
D) Simulated radiation patterns
The simulated E-plane radiation patterns are represented as shown in Fig. 6 for operating frequencies of 2.56, 3.02, 3.72, and 4.98 GHz, respectively, for the antenna without DGS. It is found that poor radiation characteristics are observed at high resonant modes. This is due to back propagation of the ground plane.
Simulated E-plane radiation patterns for (a) phi = 0° and (b) phi = 90° at resonant frequencies of 2.56, 3.02, 3.72, and 4.98 GHz (y–z plane) (without DGS).
III. EXPERIMENTAL RESULTS AND DISCUSSION
To validate the hypothesis made pertaining to the proposed antenna experimentally the antenna is fabricated and tested. Top view and bottom view of the proposed antenna shown in Fig. 7 is measured using Agilent E5071C vector network analyzer. The simulated and measured return loss curve for the antenna is presented in Fig. 8.
Photograph of the proposed antenna working for L 1 and L 2 bands: (a) top view and (b) bottom view.
Simulated and measured return loss curve for the proposed antenna.
The measured values of S 11 show a bandwidth of 6.3% (1.17–1.25 GHz) and 4.6% (1.5–1.65 GHz) at lower and upper bands, respectively. It is also found that the measured values are in good agreement with the simulated values. Figure 9 demonstrates the measured and simulated gain of the antenna versus frequency for antenna with DGS. The measured gain for the resonant frequencies is 3.9 and −0.5 dBi for L 1 and L 2 bands, respectively. As long as receiver is assumed to be near to the transmitter, negative gain can be acceptable. The gain in the upper band is about −0.5 dBi and the negative gain value is due to the low antenna efficiency at that band and the antenna designed is likely to be used for handheld devices up to −2 dBi. There is a reduction in gain predicted by the simulation which actually does not exist as given by the measurement. This reduction in gain shown by simulation may be due to limited number of grid points considered during simulation which may be affecting the results for the case of defected ground structure. Owing to the aforementioned reason, a difference of measured and simulated gain value is observed with maintained efficiency at 2.2 GHz. It is evident that the results obtained for the frequencies of operation smaller than 2 GHz via simulation approximate the respective measurements. Considering antenna performance requirements and the tradeoffs associated with operation over dual frequency bands, size constraint is typically required for handheld devices. The measured radiation efficiency of about 79 and 38% at lower and upper resonance frequencies, respectively, as shown in Fig. 10.
Measured and simulated gain response of the proposed antenna.
Measured and simulated antenna efficiency for the proposed antenna.
The measured and simulated radiation characteristics at the y–z plane of the antenna at 1.2 GHz (L 1 band for GPS receiver) and 1.65 GHz (L 2 band and Iridium), respectively, are presented in Fig. 11.
Measured and simulated radiation patterns in the y–z plane at 1.202 and 1.657 GHz.
Table 3 gives the performance of the proposed antenna when compared with current existing models. From the observed results, the improved −10 dB impedance bandwidth and ~72% of average size reduction of antenna are obtained.
Simulated results of antenna design iterations.
The size reduction effect has been calculated by comparing a traditional square patch antenna's first resonant frequency with that of the proposed DGS-based antenna. For example, a regular square patch antenna for 1.2 GHz required a length and width of 80 mm. The proposed antenna resonates at the same frequency with a size of 36 × 36 mm2. Thus, the obtained size reduction is 79.75% and calculation is performed [1−(36 × 36 mm2/80 × 80 mm2) × 100]. This concept defines the size reduction property for the designed antenna. With reference to the resonant bands, the average size reduction is calculated for both bands. ~80% size reduction for first band and ~64% for the second band. Overall size reduction is nearly ~72%. The proposed antenna has to be modified further to be integrated on handheld device and placing reflectors at a distance of λ g /4 of antenna produce unidirectional patterns. In handheld devices, the designed dual-band antenna can redirect the back radiation by proper selective dimensions of defect on back side metal of the patch without degrading the antenna performance. Typical frequencies in L 1 and L 2 bands generated are used for linearly polarized (LP) GPS receiver applications and they are unique to each application which can be placed onboard aircraft, ships, submarines, cars, and trucks. Iridium phone antenna is used for handheld telephone services works at 1.6 GHz.
VI. CONCLUSION
Minkowski fractal DGS-based truncated slit antenna is designed, fabricated, and tested for dual-band applications. Frequency shift property of DGS structure makes the proposed antenna to resonate at lower frequencies. Overall size reduction of 72% is achieved with this DGS-based antenna. Good agreement is found between the measured and simulated results of the operating antenna. The results show acceptable gain value at both LP frequency bands. Different wireless applications will be benefited with the designed single-layered fractal DGS antenna.
ACKNOWLEDGEMENTS
The authors would like to thank the editors and reviewers for their constructive comments and also wish to acknowledge the support of Prof. N.V.S.N.Sarma, Professor in Department of ECE, NIT Warangal for his helpful technical support and stimulating discussions.
B. Rama Sanjeeva Reddy received his Bachelor's degree in Electronics and Communication Engineering from Bangalore University, Bangalore, India and Master's degree in Microwaves from Government College of Engineering, Pune, India. Currently He is working in the field of antenna arrays at National Institute of Technology, Warangal, India.
D. Vakula obtained her Bachelor's degree in Electronics and Communication Engineering from Nagarjuna University, AP, India and Master's degree from Birla Institute of Technology, Mesra India, with Microwave specialization in 1992 and 1994, respectively. She obtained Ph.D. on Fault Diagnostics of Antenna Arrays from National Institute of Technology, Warangal, India in 2010. She has been working as an Assistant Professor at National Institute of Technology, Warangal, India since 2006. She has published 20 papers in various International Conferences and Journals. Her areas of interest include phased array antennas, ultrawide band antennas, multiband antennas, fault diagnostics, and neural network.
