A) Effect of elliptical notch parameters (R _g1 and R _g2)

Figure 4 demonstrates the effect of R _g1 (semi-major axis radius of the notch) on the simulated S ₁₁ characteristic of the proposed antenna. It can be investigated that the parameter R _g1 does not affect the impedance bandwidth of the antenna, while it has modified the position of resonance frequencies (f _r2, f _r3, f _r4, and f _r5.) and impedance matching at these frequencies. Due to change in the length of R _g1, the path length of the current is also modified. Due to this reason, resonance frequencies are shifted. The range of this parameter (R _g1) has been varied from 9 to 12 mm. It has been investigated from Fig. 4 that the value return loss has increased at resonating frequencies (f _r1, f _r2, f _r4, and f _r5) by increasing the value of R _g1. The uneven variation of return loss has been observed at resonance frequency f _r3. This parameter alters the electromagnetic coupling effect between ground and radiating patch. Due to this reason, the value of return loss has been changed at resonating frequencies. The maximum value of return loss is found at frequency f _r4 for R _g1 = 12 mm. On increasing the value of R _g1, resonating frequencies f _r2 and f _r3 have been shifted left, while resonance frequency f _r4 is shifted right. It can also be seen from Fig. 4 that impedance bandwidth of the antenna has been deteriorated for R _g1 <9 mm.

Fig. 4.

The simulated S ₁₁ characteristic of the proposed antenna for different values of R _g1.

Figure 5 illustrates the simulated S ₁₁ characteristic of the antenna with various values of R _g2 (semi-minor axis radius of the notch). It can be seen from Fig. 5 that the parameter R _g2 significantly affects the impedance bandwidth of the antenna. It alters the mutual coupling between radiating patch and ground plane. The value of R _g2 has been varied from 2 to 6 mm. The best impedance matching is found for R _g2 = 3 mm with the impedance bandwidth of 132.89%. It can also be investigated that impedance matching is worst for R _g2 <3 mm and R _g2 >5 mm. This is because of less electromagnetic coupling between ground plane and radiating patch. It has also been observed from Fig. 5 that this parameter alters the value of return loss at resonating frequencies.

Fig. 5.

The simulated S ₁₁ characteristic of proposed antenna for different values of R _g2.

B) Effect of size of parasitic slot (L _g2 and W _g3)

The size of parasitic slots affects the electromagnetic coupling between the defected ground plane and radiating element. It also alters the position of resonance frequencies and the impedance matching resonance frequencies. On increasing the size of parasitic slots, the impedance matching at resonance frequency f _r1 is reduced, while uneven impedance matching has been investigated at resonance frequencies f _r2, f _r3, f _r4, and f _r5. Figure 6 shows the simulated S ₁₁ characteristic of the proposed antenna for various sizes of the parasitic slots. For the larger size of the parasitic slots, i.e. L _g2 = 40 mm and W _g3 = 20 mm, the impedance bandwidth of the antenna has been deteriorated. It is because of less mutual coupling between radiating patch and ground. The best impedance matching has been found for size L _g2 = 25 mm and W _g3 = 17 mm. It has also been investigated from Fig. 6 that the size of the parasitic slot changes the value of return loss in the entire spectrum (1–6 GHz).

Fig. 6.

The simulated S ₁₁ characteristic of proposed antenna for different values of L _g2 and W _g3.

A) S ₁₁ characteristic

The photograph of the fabricated Scarecrow antenna is shown in Fig. 7. Firstly, the proposed antenna has been analyzed using CST Microwave Studio. After numerical analysis, the S ₁₁ characteristic of the antenna has been measured in the frequency band 1–6 GHz using vector network analyzer N9923A. The comparison between measured and simulated S ₁₁ characteristic of the proposed antenna has been depicted in Fig. 8. It can be seen that this antenna is resonating at frequencies 1.29, 1.77, 2.28, 2.94, 3.72, and 6 GHz. The proposed antenna has exhibited the impedance bandwidth of 116.32% from 1.23 to 4.65 GHz for |S ₁₁ | < − 10 dB. Another resonating band has been found from 5.7 to 6 GHz. Impedance matching has been deteriorated in the frequency band 4.65–5.7 GHz. Some mismatch between measured result and simulated result has been observed due to ground plane effect, SMA connector, and fabrication tolerance. The first resonating frequency has been found at 1.29 GHz. It has also investigated that resonating frequencies are generated because of the elements (radiating patch and ground plane) of the proposed antenna. The frequency 1.29 GHz has been generated because of the partial ground plane. The detail about resonating frequencies (1.29, 1.77, 2.28, 2.94, and 3.72 GHz) has been discussed in the current distribution section.

Fig. 7.

Fabricated Scarecrow patch antenna.

Fig. 8.

Comparison of measured and simulated return loss of Scarecrow patch antenna.

The lower cut-off frequency of the proposed monopole antenna depends on the length of the radiating patch. Lower cut-off frequency f _l of the proposed antenna is calculated by the given equation [Reference Ray and Ranga21, Reference Agrawall, Kumar and Ray22].

(1)$$f_l = {{70} \over {(L + r + p)k}}\; {\rm GHz,}$$

where L = 50 mm that is equal to the length of cylindrical monopole, r = 5 mm, which is a radius of the cylindrical monopole antenna (2πrL = πab,  where a = 25 mm and b = 20 mm) and p = 0 mm; k = 1.15 is constant. The measured and calculated values of lower cut-off frequency f _l are 1.23 and 1.14 GHz, respectively. The second resonating frequency is found at 1.77 GHz and it has been generated because of the radiating patch. This frequency or mode (TM ₁₁) can be calculated by the following equations [Reference Agrawal, Vakula and Sarma23].

(2)$$\eqalign{ a_{eff} &= a\left[ 1 + \displaystyle{{2h} \over {\pi \varepsilon _r a}}\; \left\{ {\rm ln}\left( {\displaystyle{a \over {2h}}} \right) + (1.41\varepsilon _r + 1.77)\right.\right. \cr &\left. + \displaystyle{h \over a}\; (0.268\varepsilon _r + 1.65) \right]^{1/2} \;,}$$

(3)$$\; \; f_{11}^{e,o} = \displaystyle{{15} \over {\pi ea_{eff}}} \sqrt {\displaystyle{{q_{11}^{e,o}} \over {\varepsilon _r}}}, $$

(4)$$q_{11}^e = - 0.0049e + 3.7888e^2 - 0.7278e^3 + 2.314e^4, $$

(5)$$q_{11}^o = - 0.0063e + 3.8316e^2 - 1.1351e^3 + 5.2229e^4. $$

where a is the semi-major axis, h is the height of the substrate, ε _r is the dielectric constant of the substrate, e is an eccentricity of the ellipse, and a _eff is an effective radius. The calculated and measured values of the second resonance frequency are 1.79 and 1.77 GHz. Small error of 1.11% has been found between calculated and measured second resonance frequencies. It is because of slots, notch, and effect of the ground plane.

B) Input impedance

The comparison of the measured and simulated input impedance (Z _in) of the proposed antenna has been displayed in Fig. 9. At the resonating frequencies, the reactance of the antenna is zero and antenna impedance is equal to the real part of the input impedance. In the frequency band 1.2–4.5 GHz (Fig. 9), measured input resistance fluctuate around 60 Ω, while simulated input resistance has oscillated around 50 Ω. In the frequency band 4.8–5.2 GHz, the measured input resistance is high because of the ground plane effect. At frequency 5 GHz, antenna behaves as a parallel resonator because the resistance of the antenna is quite high (near 200 Ω) and reactance is zero. Figure 9 also shows the comparison of measured and simulated imaginary part (Z _in(imaginary)) of the input impedance (Z _in). It has been investigated that the measured reactance of the antenna oscillates around 0 Ω in the frequency band 1.2–4.3 GHz, while simulated reactance has varied around 0 Ω in the entire frequency band (1–6 GHz).

Fig. 9.

Variation of real and imaginary parts of input impedance Z _in.

C) Current distribution

Simulated surface current distributions at resonance frequencies (1.29, 1.77, 2.28, 2.94, and 3.72 GHz) of the proposed antenna are illustrated in Fig. 10. By investigating the simulated current distribution, current vectors are not only distributed on the patch but also on the ground plane. It has also been found that the surface current distribution is symmetric about the y-axis at each resonating frequency. By inspecting the current patterns, the first resonance frequency f ₁ = 1.29 GHz has been produced because of the ground plane. At this frequency, the current vectors have distributed along the periphery of the modified partial ground plane. The fundamental frequency f ₁ can be determined by the following equations:

(6)$$L_{f1} = L_{g2} + W_{g3} + L_{g3} + W_{g2} + L_{slot}, $$

(7)$$L_{slot} = \displaystyle{{\pi R_{g1} R_{g2}} \over 4},$$

(8)$$\; f_1 = \displaystyle{c \over {L_{f1\;} \sqrt {\varepsilon _r}}} \;, $$

where L _f1 is the length. The calculated value of f ₁ is 1.255 GHz. The small difference between measured and calculated values of fundamental frequency has been found. It can be seen from Fig. 10 that second resonating frequency has been developed by the radiating patch. At second resonating frequency f ₂ = 1.77 GHz, one half-wavelength variation of current vectors has been investigated along the elliptical periphery of the patch. This frequency f ₂ has been determined by the resonance frequency equations (2)–(5) of ellipse [Reference Agrawal, Vakula and Sarma23]. The third resonating frequency (higher order mode) f ₃ = 2.28 GHz has been developed by the side edge of the ground plane and f ₃ is calculated by the following equations:

(9)$$L_{f3} = L_{g2} + W_{g3} + L_{g3}, $$

(10)$$\; f_3 = \displaystyle{c \over {L_{f3\;} \sqrt {\varepsilon _r}}} \;, $$

where L _f3 is the length. The calculated value of f ₃ is 2.23 GHz. The fourth resonating frequency f ₄ = 2.94 GHz has been produced because of the top edge and side edge of the ground plane. This frequency f ₄ is determined by succeeding the equation:

(11)$$L_{f4} = L_{g2} + W_{g3}, $$

(12)$$f_4 = \displaystyle{c \over {L_{f4\;} \sqrt {\varepsilon _r}}}. $$

Fig. 10.

Simulated surface current distribution of proposed antenna at frequencies f ₁ = 1.29 GHz, f ₂ = 1.77 GHz, f ₃ = 2.28 GHz, f ₄ = 2.94 GHz, and f ₅ = 3.72 GHz.

The calculated value of frequency f ₄ is 3.10 GHz. It has been observed that the harmonics of second resonating frequency is also present in the spectrum. The harmonics of f ₂ is determined by nf ₂ where n is a positive integer. The frequency f ₅ is nearly equal to 2f ₂. At fifth resonating frequency f ₅ = 3.72 GHz, two exact half-wave variation of current vectors has been found along the elliptical periphery of the patch.

D) Far-field pattern

Radiation patterns of the proposed antenna have been displayed in Fig. 11. Far-field patterns of the proposed antenna have been simulated and measured in H-plane and E-plane with 10° increment at frequencies 1.29, 1.77, 2.28, 2.94, and 3.72 GHz. At frequency 1.29 GHz, a good omnidirectional pattern has been found in H-plane, while bidirectional pattern is observed in E-plane. E-plane measured result differs from the simulated result because of measurement, fabrication error, and radiation from coaxial cable at lower frequencies. At frequencies 1.77, 2.28, 2.94, and 3.72 GHz, distorted bidirectional pattern (measured) has been found in E-plane. At frequencies 1.77, 2.28, and 2.94 GHz, quasi-omnidirectional pattern has been investigated in H-plane. At higher frequency, the measured results differ from simulated results due to the presence of higher order modes. These higher order modes have been generated by the feed line, ground plane, and patch. Finally, these modes have been mixed and produce complicated modal distribution. At frequency 3.72 GHz, omnidirectionality has been lost in H-plane. It may be because of higher order modes.

Fig. 11.

Far-field patterns (E-plane (left), H-plane (right)) of proposed antenna at frequencies (a) f ₁ = 1.29 GHz, (b) f ₂ = 1.77 GHz, (c) f ₃ = 2.28 GHz, (d) f ₄ = 2.94 GHz, and (e) f ₅ = 3.72 GHz.