Introduction
The inherent property of wide bandwidth and directional radiation pattern has triggered the development of tapered slot antennas (TSAs) for medical imaging and through-wall imaging applications. TSA was first reported in the year 1979 by Gibson [Reference Gibson1] which was also referred to as exponentially TSA or Vivaldi antenna. Thereafter, the development of Vivaldi antennas with different tapering profiles [Reference Yngresson, Schaubert, Korzeniowski, Kollberg, Thungren and Johansson2] was explored and reported. However, exponentially tapered slot Vivaldi antenna is mostly used owing to its least sidelobe levels as compared to linear TSA and constant width slot antennas [Reference Yngresson, Schaubert, Korzeniowski, Kollberg, Thungren and Johansson2]. Traditional TSAs have difficulty in broadband impedance matching of microstrip line and slot line resulting in the reduction of operating bandwidth. The difficulty in wideband impedance matching was vanquished by utilizing a tapered microstrip transition, through parallel strip lines, to symmetric double-sided slot line as proposed by Gazit [Reference Gazit3]. It has directional radiation characteristics with maximum radiation in the end-fire direction. For invoking maximum radiation in the end-fire direction, the phases of the traveling wave current in either arm must be in the opposite phase. Thereafter the advantages of antipodal Vivaldi antenna (AVA) came to recognize in the wideband antenna design paradigm. Some earlier development of AVA included significant works reported in [Reference Greenberg, Virga and Hammod4] and [Reference Siddiqui, Antar, Freundorfer, Smith, Morin and Thayaparan5]. Attempt to reduce the size of the antenna and thereby exploring bandwidth enhancement techniques have been reported in [Reference Bai, Shi and Prather6, Reference Fei, Jiao, Hu and Zhang7]. In [Reference Bai, Shi and Prather6], both circular loading and slot-loading techniques are adopted for improvement in impedance matching while in [Reference Fei, Jiao, Hu and Zhang7], exponential slot loading is done for improvement in impedance matching. One of the designs presents corrugation in the radiating arms for improving directivity and bandwidth [Reference Pandey and Meshram8]. An improvement in impedance bandwidth has been achieved by using a slot edge with exponential opening forming a palm tree-like structure [Reference De Oliveira9]. A similar design has been reported in [Reference Mahmud, Islam, Samsuzzaman, Kibria and Misran10] where a side slotted Vivaldi antenna is reported to improve the impedance matching. In [Reference Natarajan, George, Kanagabasai, Lawrance, Moorthy, Rajendran and Alsath11], a comparison of single petal structured Vivaldi antenna and dual petal structured Vivaldi antenna are compared. It is observed that a dual petal structured AVA offers more impedance bandwidth than a single petal structured AVA. In [Reference Malakooti, Moosazadeh, Ranasinghe and Fumeaux12], it is shown that elliptical corrugations on the outer edge of AVA with optimal shaping of the inner edges of the antenna result in good impedance matching. Moreover, a patch compensated wideband Vivaldi antenna is reported in [Reference Yadav, Kumar and Rajveer13].
In this work, novel impedance matching techniques like binomial tapering of the outer edge, binomial slit loading, and protruded binomial stub loading are incorporated in the conventional antipodal Vivaldi antenna (CAVA) to increase the electrical length and thereby reducing the lower operating frequency. The proposed antenna has a fractional bandwidth of 164% and is compact compared to other related works [Reference Pandey and Meshram8–Reference Yadav, Kumar and Rajveer13]. The organization of the paper is as follows. The design methodology is described in the section “Antenna design methodology”. Experimental results are compared with its predicted values and are presented in the section “Results and discussions” which is followed by a conclusion in the section “Conclusion”.
Antenna design methodology
The diagrammatic representation of the proposed antenna is depicted in Fig. 1. The proposed AVA is binomially tapered, loaded with binomial slits, and binomial tapered stubs. Rigorous preliminary design alternatives and corresponding simulation data revealed that the binomial tapering is found to be more effective in impedance matching at lower frequencies, unlike conventional exponential tapering. However, for brevity, those have not been given here. All the simulations have been carried out using commercially available electromagnetic simulation software CST microwave studio™. The designed AVA has a compact size of 40.4 mm × 38 mm, etched on a low-cost FR4 substrate with permittivity of 4.4 and thickness 1.59 mm. The stages of development of the proposed antenna, described in this paper, are depicted in Fig. 2 and are detailed under the following subsections. The preliminary design analysis discloses that not all design tapering types can be applied brute force in achieving compact as well as wide impedance bandwidth. Some tapering may provide exceptionally suitable impedance bandwidth, but the overall size is large. So, in presenting the design in this work, the suitable alternative that satisfies impedance bandwidth, consistent radiation characteristics, as well as compact overall dimension was found to be offered by a Binomial tapering and a similar profile in the realization of the slit and stubs.
(a) Layout of the proposed antenna, (b) details of various tapering profiles. Lsub = 38, Wsub = 40.4, Ws = 15.8, Lt = 12.08, Ls = 36, lf = 7, wf = 1.47, g = 1, tsl = 3.47, wstb = 7, lb = 17.22, ls = 13, wsl = 6.8, lsl = 7(in mm).
Evolution of the proposed AVA.
Design of the CAVA
The exponential inner tapering of CAVA is obtained from equation (1) [Reference Greenberg, Virga and Hammod4].
In (1), a is the coefficient that relates the slot widening and ki is half of the total slot width at the feed interface of the antenna. Similarly, the outer edge of CAVA is obtained by using equation (2) [Reference Greenberg, Virga and Hammod4].
where k0 is the cumulative value of one of the conductors at the feed and ki. The coefficient b relates to the shape of the outer edge taper. The maximum separation between the arms of CAVA is set at λg/2 at the lowest frequency. The optimized values of ki, ai, k0, a0, and b are 0.13, 0.15, 1.6, 0.01, and 2 respectively. As depicted in Fig. 2, the bandwidth obtained by CAVA ranges from 3 to 20 GHz. For the CAVA to have wide operational bandwidth, the phases of the traveling wave currents on each arm have to be 1800 apart. However, the limiting factors of the bandwidth are tapering of the slot radiator, the width of the feed line, and transition from the feed line to the radiator.
Effect of binomial tapering of the outer edge
Binomial tapering [Reference Ling, Lo, Yan and Chung14] of the outer edge of CAVA is done according to equation (3).
In CAVA, the flare length in terms of wavelength is much shorter in lower frequencies, resulting in less efficient radiation of current on the flare. Binomial tapering of outer edge results in an increase in electrical length along the outer edge, which shifts the first resonant frequency in the lower frequency region. As depicted in Fig. 2, the bandwidth obtained by binomial tapered antipodal Vivaldi antenna (BTAVA) ranges from 2.8 to 20 GHz. A parametric study with different values of the order of binomial function N was carried out and is depicted in Fig. 3. It is revealed that, when N = 1, there is an impedance mismatch around 9 and 11 GHz. Moreover, binomial tapering in the order of N = 3 also results in impedance mismatch around 5 GHz. The best result of |S 11| characteristics is obtained when N = 2.
Effect of variation of |S 11| characteristics for different values of N.
Effect of binomial slits in the outer edge
Further improvement in input impedance matching in low-frequency region can be accomplished by etching binomial slits along the outer edges (Fig. 1). Binomial slits are designed according to the design principle proposed in [Reference Ling, Lo, Yan and Chung14]. The slit near to the feed, in either conductor, is created by employing (4) and (5).
where z | = zcosθ and y | = ycosθ, and θ is defined by (6).
The other slits are constructed similarly and are separated by lsl and the value of N| is taken to be 2. The effect of variation of the order of binomial function N| is shown in Fig. 4. The observation reveals that when N| increases from 1 to 2, the first resonant frequency shifts to lower frequency region, thereby increasing the impedance matching at lower frequencies. However, no significant change in |S 11| characteristics is observed on further increase in N|.
Effect of variation of |S 11| characteristics for different values of N|.
Moreover, the impedance matching around 6–8 GHz is affected by the variation in length l sl as shown in Fig. 5. It is observed that the best impedance matching is obtained when the value l sl is 7 mm.
Effect of variation of |S 11| characteristics for different values of lsl.
The parametric study of tsl is done and is reported in Fig. 6. It is observed that, increase in tsl shifts the first resonant frequency in the lower frequency region. However, the best result is achieved when the value of tsl is 3.47 mm. Further increase in tsl does not have a significant effect on |S 11| characteristics.
Effect of variation of |S 11| characteristics for different values of tsl.
Effect of binomial stubs in the outer edge
A binomial tapered stub is appended near to the feed to introduce some additional modes and thereby further increasing the input impedance matching in lower frequencies. As shown in Fig. 2, the impedance bandwidth of the protruded stub BTAVA exhibits an impedance bandwidth that ranges from 2 to 20 GHz. The design of the protruded stub is governed by the following equations.
The values of the orders N|| and N||| of binomial functions are 3 and 4, respectively.
Results and discussions
The fabricated prototype of the proposed antenna is shown in the inset of Fig. 7. The affirmation of the simulated S-parameter results of the proposed antenna is done using Rhode and Schwarz ZVA 40 VNA and is portrayed in Fig. 7. A small discrepancy between simulated and measured |S 11| dB characteristics is due to inaccuracy in the soldering of the SMA connectors. The measured impedance bandwidth of the proposed antenna ranges from 2 to 20 GHz. The current distribution plot of the proposed Vivaldi antenna at frequencies 2.5, 10, and 18 GHz are shown in Fig. 8. It is observed that the current density is significant along the edges of the binomial slots and in the binomial tapered slots at lower frequencies, which indicates an effective enhancement of the current path resulting in an increase in lower-end bandwidth.
Comparison of simulated and measured |S 11| characteristics.
Current distribution at (a) 2.5 GHz, (b) 10 GHz, and (c) 18 GHz.
The simulated and measured radiation patterns at different frequencies are shown in Fig. 9. End fire radiation properties are observed throughout the operational bandwidth. Higher-order modes are generated at higher frequencies resulting in some ripples in the radiation patterns which is observed in Fig. 9. Peak realized gain of the proposed antenna is more than 8 dBi as depicted in Fig. 10. Moreover, the simulated and measured radiation efficiency of the proposed antenna is also depicted in Fig. 10. It is observed that the radiation efficiency is more than 90% for the proposed AVA.
Simulated and proposed radiation pattern of the proposed AVA in xz plane and yz plane at (a) 2 GHz, (b) 7 GHz, (c) 12 GHz, and (d) 20 GHz.
Simulated and measured gain and efficiency of the proposed antenna.
Phase linearity within the operational bandwidth is an important aspect of wideband antenna design. For this, the maximum acceptable group delay must be within Dt = (1/2fs) [Reference Perez-Martinez, Burgos-Garcia and Asensio-Lopez15]. It is observed from Fig. 11 that the group delay response is almost flat, which indicates distortion-less transmission. Moreover, the pulse fidelity factor is also an important performance metric of the antenna in the time domain. The fidelity factor shows the amount of similarity between the transmitted and received pulses. The fidelity factor is determined by utilizing equation (9).
where st(t) and sr(t) are input signal and received signal, respectively. The transmitting antenna and receiving antenna are kept at a far-field distance at the lower frequency of the operational bandwidth [Reference Yang, Chu and Zheng16]. Fidelity factors for different orientations of transmit-receive antenna systems in xz plane and yz plane are obtained and tabulated in Table 1. It is revealed that as the signals approach orthogonality, the fidelity factor decreases.
Group delay profile of the proposed AVA.
Fidelity factors of the proposed AVA in two principle planes
A comparison of the proposed Vivaldi antenna with the previously reported Vivaldi antennas in [Reference Pandey and Meshram8–Reference Yadav, Kumar and Rajveer13] is tabulated in Table 2. It is revealed, that the proposed antenna is compact and has the highest percentage bandwidth of 164%. Moreover, the proposed Vivaldi antenna has an efficiency of 90%.
Comparison of the existing literature with the proposed antenna
Conclusion
In this work, binomial tapering of outer edges of antipodal Vivaldi antenna, loaded with binomial slits and binomial tapered stubs have been found suitable for improving impedance matching in the lower frequency region and thereby achieves super wideband characteristics. It has been observed that the antenna exhibits a percentage bandwidth of 164% with satisfactory end-fire radiation characteristics. The conventional trend of increase in directivity with frequency is observed for the proposed antenna. The flat group delay profile of the antenna confirms the non-dispersive nature of the proposed antenna and thereby prevents the generation of time-varying non-linearity translating to higher-order resonances. Moreover, the antenna also shows a high peak realized gain of more than 8 dBi. Assembling all these properties of the designed antenna, make it suitable for through-wall imaging and medical imaging applications.
Abhik Gorai received his M.Tech from National Institute of Technology, Durgapur. He started his career as an engineer in Huawei Telecommunication and thereafter, he served in Alkatel Lucent. He is presently an Assistant Professor in the School of Electronics engineering of KIIT Deemed University. His interest lies in fractal antennas, metamaterials, MIMO Antennas, and SIW Antennas. He is a reviewer of IEEE, IET, IETE, and Elsevier journals.
Rowdra Ghatak received his M.Tech (Microwave Engineering) from The University of Burdwan and Ph.D. (Engg) from Jadavpur University. He initiated his career in microwave engineering as a trainee at CEERI Pilani in fabrication and testing of S-band magnetrons. Thereafter he served at the National Institute of Science and Technology Berhampur, Odisha and at the University of Burdwan. He is presently a Professor in Electronics and Communication Engineering Department of National Institute of Technology Durgapur. He is a recipient of the URSI Young Scientist Award in 2005. He received support under DST Young Scientist scheme for development of UWB antennas for imaging RADAR. He has more than 250 publications in various National/International journals and conferences. His research interest lies in the areas of the fractal antenna, metamaterials, application of evolutionary algorithms to electromagnetic optimization problems, RFID, computational electromagnetic and microwave passive and active circuit design. He is a reviewer of IEEE, IET, Elsevier, John Wiley, and Springer journals. He has organized workshops in the capacity of workshop chair/convener on Microwave Circuits and Antenna and has been a member of technical program committee membership of numerous IEEE conferences in the Asia Pacific region. He was a student paper contest co-Chair in the IEEE AEMC 2013 and Technical program co-chair at IEEE MicroCom 2016 and conducted numerous technical lectures.
