I. INTRODUCTION
The cylindrical shape dielectric resonator (DR) is the first basic geometry used as an antenna element for radiation and termed as cylindrical DR antenna (CDRA) [Reference Petosa1–Reference Kajfez and Guillon4]. The cylindrical DRs are available with the following lowest order modes such as TE01δ , TM01δ , HE11δ , EH11δ , TE011+δ etc. [Reference Kajfez and Guillon4]. For each mode, it has its own radiation pattern which can be cast-off as per required application. It offers wide bandwidth and high efficiency compared with microstrip antenna, due to the absence of surface wave and also it radiates through the entire surface [Reference Mongia and Ittipiboon5, Reference Saed and Yadla6].
One of the basic techniques being reported by some researchers to enhance the bandwidth of the DRAs is sectoring and splitting of the elementary structure [Reference Kishk and Glisson7–10]. Through the sectoring and splitting, radiation area as well as volume can be reduced, which is useful to design low weight antenna. Compact cylindrical sector DRA has been developed with a 75% reduction in volume compared with conventional CDRA with the support of fractional modes and allowance of dual frequency/dual feed operation [Reference Mongia, Ittipiboon and Antar8]. A slot-coupled half-split CDRA placed on a metallic plane excited with magnetic dipole mode provides 10% enhancement in bandwidth with monopole radiation pattern [Reference Mongia9]. The volume of the sectored antenna can be reduced by using metalized faces of the DRA [Reference Matthew, Tam and Ross3, Reference Cormos, Laisne, Gillard, Bolzer and Nicolas10].
Monopole radiation pattern with different shapes are reported in literature [Reference Mongia, Ittipiboon, Bhartia and Cuhaci11–Reference Kishk, Glisson, Junker and Kajfez19]. For small and compact electric monopole antenna, dielectric ring resonator antenna has been introduced by Mongia et al. [Reference Mongia, Ittipiboon, Bhartia and Cuhaci11]. Four element CDRA is introduced by Guha and Antar [Reference Guha and Antar14], which shows monopole-like radiation pattern with 29% impedance bandwidth. Half-split hemispherical DRA and segmented quarter split hemispherical DRAs have been investigated as broadband variants of low-profile monopole type DRAs and provide 35% of impedance bandwidth [Reference Guha, Gupta, Kumar and Antar13, Reference Guha and Antar14]. Four-element rectangular and triangular DRAs excited with coaxial probe provide around 35% bandwidth with monopole radiation patterns [Reference Gangwar, Ranjan and Aigal15, Reference Gangwar, Singh and Kumar16].
This paper presents the design and development of composite forms of q-CDRA with performance comparison of single, two, and four elements. The simulation studies of the proposed antennas are carried out using Ansoft high-frequency structure simulator (HFSS) simulation software. The proposed four-element composite q-CDRA shows almost double improvement in bandwidth as compared with the published structures [Reference Guha, Gupta, Kumar and Antar13, Reference Guha and Antar14]. Some elementary simulated results for the proposed four-element composite q-CDRA have been reported in [Reference Ranjan and Gangwar17]. The prototypes of all three antennas have been fabricated and tested. The measured results have been compared with simulated one and found good agreement between each other. Proposed four-element composite q-CDRA provides 58.15% bandwidth (|S 11| ≤ −10 dB) for uniform monopole-like radiation pattern over the entire band, which has much larger bandwidth compared with single as well as two elements q-CDRA.
II. ANTENNA GEOMETRY
The cylindrical sector DRA has been proposed by cutting a part with a certain angle from the mold cylinder. The cylindrical sector DRA has been shown in Fig. 1, which consists of a cylindrical DRA with the sector of an arbitrary angle β on a ground plane. For q-CDRA, the angle β is 90°, i.e. angle between face 1 and face 2. The prototype for formation of q-CDRA has been shown in Fig. 2, i.e. by splitting the basic CDRA into four equal parts as shown in Fig. 2. The geometry and fabricated structure of q-CDRA has been shown in Fig. 3. Its sector is face left open so that approximately magnetic wall boundary conditions exist on them.
Geometry of cylindrical sector CDRA.
Geometry for formation of q-CDRA through CDRA.
(a) Geometry and (b) fabricated structure of q-CDRA.
By assuming a magnetic wall approximation for complete surface, as it is open in air. The field exists inside the cylindrical sector DRA in terms of TM mnp modes are [Reference Long, Stuart, Mark and Shen2, Reference Matthew, Tam and Ross3]
$$\eqalign{E_z^{mnp} & = \left[ {A{J_m}\left({{k_r}r} \right) + B{Y_m}\left( {{k_r}r} \right)} \right] \cr & \quad.\left[ {C\,{\rm sin}\left( {v\emptyset} \right) + D\,{\rm cos}\left({v\emptyset} \right)} \right]{\rm cos}\left( {{k_z}z} \right).} $$
Here A, B, C, and D are arbitrary constants, which depend on the placement and feed of the DRA. J m , Ym denote first and second kind mth-order Bessel functions, respectively, where m is the positive real number which depends upon boundary conditions and n, p are positive integers. For simplification of excitation, coaxial probe feed has been used at the center of cylinder.
The resonant frequency of the TM mnp mode:
$$k_r^2 + k_z^2 = {k^2} = {\epsilon _r}{\left( {\displaystyle{{2\pi f} \over c}} \right)^2},$$
where k r = X′ mn /a, k z = (2p + 1)/2h are wavenumbers.
X′ mn is the root satisfying J′ m (X′ mn ) = 0.
Thus, the final equation for q-CDRA calculating the resonant frequency for the TM mnp mode is [Reference Matthew, Tam and Ross3]
$${f_{mnp}} = \displaystyle{c \over {2\pi a\sqrt {{\mu _{r\;}}{{\rm \epsilon }_r}} }}{\rm }\sqrt {X{{_{mn}^{\prime} }^2} + {{\left[ {\displaystyle{{\pi a} \over {2h}}\left( {2p + 1} \right)} \right]}^2}}. $$
Here for q-CDRA, we have chosen β = π/4. As we are leaving open both the faces of the sector, it will form approximate magnetic walls. The magnetic walls on faces 1 and 2 have the boundary conditions
$$\eqalign{\displaystyle{{\partial {E_z}} \over {\partial \phi}} & = 0\; \;{\rm at}\;\phi = 0\; \;{\rm and}\;\phi = \beta, \quad n = \displaystyle{{v\pi} \over \beta}, \cr & \quad \; \quad {\rm where}\;0 \lt \beta \le 2\pi,} $$
where v is a non-zero positive integer to satisfy the boundary conditions at faces 1 and 2, respectively. The solutions for both faces are open, required C = 0. The resonant frequency depends on radius and height of the cylindrical sector DRA for a particular mode. The resonant frequency for TM01δ mode of the q-CDRA has been calculated using equation (3) for 0 < δ < 1, which will be between 7.2 and 10.93 GHz, depends on δ value [Reference Matthew, Tam and Ross3]. A two elements composite q-CDRA has been designed and fabricated as shown in Fig. 4. The proposed four-element composite q-CDRA has been introduced by combining four similar quarter split sectors of the CDRA. Four identical quarter sectors are kept in a compact way such that they form complete cylinder with an equal air gap as shown in Fig. 5(a). The introduction of air gap between the elements is responsible for the enhancement in bandwidth [Reference Kishk, Glisson, Junker and Kajfez19]. The proposed four-element composite q-CDRA has been designed using Ansoft HFSS simulation software and fabricated for validation. The antenna geometry and fabricated structure of the proposed four-element composite q-CDRA is shown in Figs 5(a) and 5(b), respectively. Each element of the proposed DRA structure has dimensions of radius a = 4 mm and height h = 13 mm. The fabricated proposed antenna element has relative permittivity (εr) of 9.8 (alumina ceramic, 99%). The DR elements are packed together in a compact way on a metallic ground plane of dimensions 50 × 50 × 3 mm3, which is fed through 50 Ω coaxial probe. Depending upon the size of ground plane, the effects like the ripple in radiation pattern by edge diffraction, reduction in gain due to radiation spillover, and change in input impedance due to extra current reflection at the edge have occurred [Reference Kajfez and Guillon4, 20]. The size of the ground plane has been obtained through parametric analysis for minimum return loss and maximum bandwidth. From Fig. 6, it can be observed that 50 × 50 mm2 is providing the best result, and beyond this it is showing the same bandwidth. As the dimension of the DRA is very less as compared with the ground plane size, so it can be assumed that the optimized dimension acts as infinite ground plane for the proposed DRA. Optimization of ground plane size of the proposed antenna has been performed with the help of Ansoft HFSS simulation software and shown in Fig. 6. The copper plate (99%) has been used as the ground plane for the fabrication of the antenna. Since the dimension of DR element is very less compared with the ground plane, we can assume the ground plane to behave like infinite ground plane for the DRA. The optimization of probe height for proposed composite q-CDRAs have been performed though extensive simulations for minimum return loss. The height of the probe is found to be 10.6, 12, and 12.4 mm from the bottom of the ground plane for q-CDRA, two elements composite and four elements composite q-CDRA, respectively.
(a) Geometry and (b) fabricated structure of two elements composite q-CDRA.
(a) Geometry and (b) fabricated structure of the proposed four elements composite q-CDRA.
Variation in |S 11| curve with frequency for different ground plane sizes of four-element composite q-CDRA.
III. RESULTS AND DISCUSSION
The performance of the proposed antenna has been compared with complete CDRA. The |S 11| versus frequency curve has been shown in Fig. 7 for the proposed antenna and complete CDRA. From Fig 7, it can be observed that the operating bandwidth is high for four-element q-CDRA compared with CDRA. This is due to decrease in effective dielectric constant of the proposed DRA, as the air gap is introducing between q-CDRAs.
|S 11| versus frequency curves of complete CDRA and four-element composite q-CDRA.
A) q-CDRA
The input characteristics and far field of the proposed q-CDRA have been investigated using Ansoft HFSS simulation software and compared with the measured result. The input characteristics measurement has been performed using Rohde & Schwarz Vector Network Analyzer (Model No. ZVM, 10 MHz–20 GHz). The |S 11| versus frequency curves of the proposed q-CDRA have been shown in Fig. 8. The simulated as well as measured resonant frequency, operating bandwidth, and percentage bandwidth has been extracted from Fig. 8 and compared in Table 1. The measured resonant frequency for q-CDRA is found to be 10.00 GHz which is coming in between theoretically calculated resonant frequency range (7.2–10.93 GHz) using equation (3). The simulated resonant frequency is found to be 10.55 GHz, which is approximately near to the theoretical and measured resonant frequency. From Fig. 8 and Table 1, it can be observed that there is a slight mismatch found between the simulated and measured data, which may be due to the fabrication error.
|S 11| versus frequency curves of q-CDRA.
Resonant frequency and return loss performance of proposed composite q-CDRAs.
The far field for q-CDRA has been studied through Ansoft HFSS simulation software and compared with the measured one. The radiation pattern measurement has been taken inside anechoic chamber. The radiation pattern has been shown in Fig. 9 at resonant frequency for XY-plane (θ = 90°), YZ-plane (ϕ = 90°), and ZX-plane (ϕ = 0°). The q-CDRA shows nearly omni-directional pattern for all three planes.
Radiation pattern of q-CDRA (a) in the XY-plane, (b) in the YZ-plane and (c) in the ZX-plane at resonant frequency.
B) Two-element composite q-CDRA
The characteristics of two elements composite q-CDRA have been investigated using Ansoft HFSS simulation software and compared with the measured one. The |S 11| versus frequency curves of the proposed q-CDRA and two elements composite q-CDRAs have been compared and shown in Fig. 10. The two elements composite q-CDRA has much wide bandwidth as compared with q-CDRA due to more radiation area. The operational bandwidth, resonant frequency, and percentage bandwidth have been extracted from Fig. 10 and tabulated in Table 1. The measured resonant frequency is found to be 6.5 GHz, which is very close to the simulated resonant frequency of 6.8 GHz. From Fig. 10 and Table 1, it can be observed that there is a slight mismatch between the simulated and measured data, which may be due to the fabrication error.
Comparison of |S 11| versus frequency curves of proposed single element and two elements composite q-CDRA.
The radiation pattern of the proposed composite q-CDRA has been studied through Ansoft HFSS simulation software and compared with the measured one at resonant frequency, which is shown in Fig. 11. The radiation pattern measurement has been taken inside anechoic chamber for the XY-plane (θ = 90°), YZ-plane (ϕ = 90°), and ZX-plane (ϕ = 0°). The composite antenna shows nearly monopole-like radiation pattern and omni-directional radiation pattern in the XY- and YZ-planes.
Radiation pattern of two-element composite q-CDRA (a) in the XY–plane, (b) in the YZ-plane and (c) in the ZX-plane at resonant frequency.
C) Four-element composite q-CDRA
The |S 11| versus frequency curves of the proposed four elements composite q-CDRA has been investigated using Ansoft HFSS simulation software. Simulated results are compared with the measured results. Measurement has been performed using Rohde & Schwarz Vector Network Analyzer (Model No. ZVM, 10 MHz–20 GHz). The |S 11| versus frequency curves of the proposed four-element composite q-CDRAs have been shown in Fig. 12. From Fig. 12, the resonant frequency, operating frequency range, and the percentage bandwidth of the proposed structures are extracted and shown in Table 1. From Fig. 12 and Table 1, it can be observed that the proposed four elements composite q-CDRA provides wide bandwidth (~58.15%) and covers complete 5.0 GHz WLAN and WiMAX band. There are slight mismatches found between the simulated and measured resonant frequency and percentage bandwidth of the proposed antenna, which may be due to the fabrication error. It can be also observed that the proposed four elements composite q-CDRA having lower resonant frequency as compared with single element and two elements composite q-CDRA. This is due to the increase in volume-to-surface area of the proposed structure as compared with single element and two elements composite q-CDRAs. The performance comparison between other shapes of DRA (previous published article) with the proposed four-element composite q-CDRA antenna has been shown in Table 2. From Table 2, it can be observed that the proposed four-element composite q-CDRA has better percentage bandwidth compared with other published DRAs. It can also be observed from Table 2 that the proposed four-element composite q-CDRA takes very less radiation area for same performance or even better than other shapes.
Comparison of |S 11| versus frequency curves of proposed two elements and four elements composite q-CDRA.
Comparison of the proposed four-element composite q-CDRA performance with similar type of antenna.
The input impedance versus frequency characteristic of the proposed four-element composite q-CDRA has been investigated using Ansoft HFSS simulation software. Simulated results have been compared with the measured ones and shown in Fig. 13. Measurement has been performed using Rohde & Schwarz Vector Network Analyzer (Model No. ZVM, 10 MHz–20 GHz). From Fig. 13, the measured and simulated value of input resistance at the resonant frequency of the proposed four-element composite q-CDRA has been found to be 51.50 and 50.08 Ω, respectively. The measured value of input resistance of the proposed four-element composite q-CDRA is in close agreement with the simulated one, which shows good impedance match with 50Ω coaxial probe feed.
Variation of input impedance versus frequency curve of the proposed four-element composite q-CDRA.
The near-field distributions of the proposed composite q-CDRA have been performed at resonant frequency through Ansoft HFSS simulation software. The distributions of E- and H-fields of the proposed composite q-CDRA are shown in Figs 14(a) and 14(b), respectively. It is apparent from Fig. 14 that the distributions reveal TM01δ -like mode in the DR elements of the proposed composite q-CDRA. The resonant frequency occurs at around 7 GHz that exactly corroborates the dominant mode resonance in the proposed antenna. It is also observed that the electric field components face their counter vectors in the XY-plane and thus cause no radiation along the broadside direction. The resulting electric fields are polarized along the z-axis and thus lead to a vertically polarized radiation surrounding the radiating structure like a quarter wave magnetic monopole [Reference Petosa1].
The far-field patterns of the proposed composite q-CDRA at resonant frequency are studied through simulation and measurement. The simulation has been done using Ansoft HFSS simulation software and measurement of the proposed antenna has been performed inside the anechoic chamber. The simulated and measured radiation patterns of the proposed composite q-CDRA at the resonant frequency are shown in Figs 15(a)–15(c) in XY (θ = 90°), YZ (ϕ = 90°), and ZX (ϕ = 0°) planes, respectively. The simulated and measured radiation patterns being observed at 5.5 and 6 GHz in all three planes are shown in Fig. 15 and Fig. 16, respectively. From Figs 15, 16 and Fig. 17, it can be observed that the radiation pattern of the proposed four-element composite q-CDRA shows good agreement in the simulated and measured patterns. It is also observed that the radiation pattern of the proposed composite q-CDRA has omni-directional pattern in the XY-plane. This may be due to the fact that no radiation along broadside of the proposed composite q-CDRA is in conformity with the counteracting E-field distributions within the elements of the proposed composite q-CDRA [Reference Kajfez and Guillon4]. The ground plane should act as infinite ground plane to reduce the effect of back lobe [Reference Balanis20]. In Figs 15–17, it is shown that the proposed four elements composite q-CDRA has monopole-like radiation pattern at all three frequencies. The XY-plane shows omni-directional radiation pattern and in the YZ- and ZX-planes null at the broadside (θ = 0°) due to TM01δ mode. For the proposed four-element composite q-CDRA, the cross polarization level is found to be 20 dB down with respect to copolarization level in all three planes at different measured frequencies.
(a) E-field and (b) H-field distribution in the proposed four-element composite q-CDRA.
Radiation pattern of proposed four-element composite q-CDRA (a) in XY-plane, (b) in YZ-plane, (c) in ZX-plane at 5.5 GHz.
Radiation pattern of the proposed four-element composite q-CDRA (a) in the XY-plane, (b) in the YZ-plane, and (c) in the ZX-plane at 6 GHz.
Radiation pattern of the proposed four-element composite q-CDRA (a) in the XY-plane, (b) in the YZ-plane, and (c) in the ZX-plane at resonant frequency 7.05 GHz.
The simulated and measured gains versus frequency curves of the proposed four-element composite q-CDRA are shown in Fig. 18. Three-antenna method has been used for gain measurement [Reference Balanis20]. From Fig. 18, it can be observed that the gain has highest value (4.70 dBi) at the resonant frequency and nearly constant gain in complete operating bandwidth. It is also depicted in Fig. 18 that there is close agreement with the measured and simulated gains of the proposed antenna. The value of simulated radiation efficiency and peak directivity at the resonant frequency of the proposed antenna is found 98% and 4.80 dB, respectively.
Variation of gain versus frequency curve for the proposed four-element composite q-CDRA.
IV. CONCLUSION
A probe feed novel q-CDRA and multi-element composite q-CDRAs have been designed and fabricated. The return loss characteristics of the q-CDRA and multi-element composite q-CDRAs have been compared with each other. The four-element composite q-CDRA provides wider bandwidth (~58.15%) and cover whole 5.0 GHz WLAN and WiMAX band with uniform monopole type radiation pattern over the complete matching bandwidth. The proposed four-element composite q-CDRA performance has been also compared with similar type of antenna and found better in performance. Designed composite q-CDRA has found a suitable application in WLAN and WiMAX band.
ACKNOWLEDGEMENTS
Author, Ravi Kumar Gangwar wishes to acknowledge Directorate of Extramural Research & Intellectual Property Rights (ER&IPR), DRDO, India for the financial support of this work through Extramural Research Grant project No. ERIP/ER/1303145/M/01/1549. The authors also would like to acknowledge Prof. Dharmendra Singh, Department of Electronics and Communication Engineering, IIT Roorkee, India, for his guidance and also for providing his antenna measurement facility at IIT Roorkee.
Pinku Ranjan was born in Nalanda (Bihar), India, in 1988. He received his B.Tech in Electronics and Communication Engineering from Jawaharlal Nehru Technological University (JNTU) Hyderabad, India, in 2010. He is presently pursuing his Ph.D. as Senior Research Fellow in Department of Electronics Engineering from Indian School of Mines, Dhanbad, India. He served as Junior Research Fellow in the Department of Electronics Engineering, Indian School of Mines, Dhanbad, India, during 2012–2014. He has authored or co-authored for more than eight research papers in international/national journal/conference proceedings. His research interests include dielectric resonator antenna and bio-electromagnetics. He is a student member of IEEE society.
Ravi Kumar Gangwar was born in Bareilly (Uttar Pradesh), India, in 1983. He received his B.Tech. degree in Electronics and Communication Engineering from U.P. Technical University, Lucknow, India, in 2006, and Ph.D. degree in Electronics Engineering in 2011 from the Indian Institute of Technology (Banaras Hindu University), Varanasi, India. He is presently Assistant Professor in Department of Electronics Engineering, Indian School of Mines, Dhanbad. He served as Senior Research Fellow/Junior Research Fellow in the Department of Electronics Engineering, IIT-BHU during 2007–2011. He has authored or co-authored over 48 research papers in international/national journals/conference proceedings. His research interest includes dielectric resonator antenna, metamaterial and bio-electromagnetics. He is member of IEEE APS society. He is a reviewer of PIER, International Journal of Electronics, International Journal of Microwave and Wireless Technology, Wireless Personnel Communication international journals.
