Configuration and analysis of the proposed antenna

The configuration of the proposed self-diplexing cavity-backed antenna is illustrated in Fig. 1. The antenna is constructed using QMSIW, a slot, and two 50Ω independent feed lines. The slot is engraved on the top side of the QMSIW to form two EMCRs. These EMCRs are excited by two independent 50Ω feed lines for radiating at two distinct frequencies. The antenna is realized to operate at 2.6 and 4.9 GHz for long-term evolution (LTE) and public safety band (PSB) applications, respectively. This antenna is realized on 0.787 mm thick 5870 RT/Duriod substrate with ε_r = 2.33 and tanδ = 0.002.

Fig. 1. Schematic of QMSIW cavity-backed slot antenna. Final dimensions: L ₁ = 22, W ₁ = 18, L ₂ = 6.0, L ₃ = 12.5, L ₄ = 24, L _a = 5.0, W _a = 0.5, L _b = 3.0, W _b = 0.5, L _c = 3.0, W _c = 1.5, L _d = 20.95, W _d = 1.8, d = 1.0, p = 2.0 (units: millimeters).

Initially, a full-mode SIW (FMSIW) cavity ($W_{eff}^{FMSIW}$ × $L_{eff}^{FMSIW}$) is designed to operate TE ₁₁₀ mode at 3.4 GHz following the design equations in [Reference Barik, Cheng, Dash, Pradhan and Karthikeyan14]. This FMSIW is divided into two equal parts, each part is called as half-mode SIW (HMSIW). Then the HMSIW is divided into two halves to produce QMSIW. The resonant frequencies of each SIW cavity is determined by following [Reference Jin, Li, Alphones and Bao3]:

(1)$$f_{mn0} ^{FMSIW} = {1\over {2\pi \sqrt {\varepsilon _r } }}\sqrt {\left({{{m\pi }\over {W_{eff} ^{FMSIW} }}} \right)^2 + \left({{{n\pi }\over {L_{eff} ^{FMSIW} }}} \right)^2 }$$

(2)$$W_{eff} ^{FMSIW} = W_{SIW} - 1.08{{d^2 }\over p} + 0.1{{d^2 }\over {W_{SIW} }}$$

(3)$$L_{eff} ^{FMSIW} = L_{SIW} - 1.08{{d^2 }\over p} + 0.1{{d^2 }\over {L_{SIW} }}$$

(4)$$W_{eff} ^{HMSIW} = {{W_{eff} ^{FMSIW} }\over 2} + \Delta W_{add}$$

(5)$$\matrix{ \Delta W_{add} = hs\left({0.05 + {{0.3}\over {\varepsilon _r }}} \right)\times \cr \ln \Bigg({0.79{{W_{eff} ^{FMSIW} }\over {4hs}} + {{52W_{eff} ^{FMSIW} - 261}\over {hs^2 }} + {{38}\over {hs}} + 2.77} \Bigg)\cr }$$

(6)$$f_{mn0} ^{QMSIW} = {1\over {2\pi \sqrt {\varepsilon _r } }}\sqrt {\left({{{m\pi }\over {W_{eff} ^{QMSIW} }}} \right)^2 + \left({{{n\pi }\over {L_{eff} ^{QMSIW} }}} \right)^2 }$$

(7)$$W_{eff} ^{QMSIW} = W_{eff} ^{HMSIW}$$

(8)$$L_{eff} ^{QMSIW} = L_{eff} ^{HMSIW}$$

The proposed antenna is designed using a QMSIW as illustrated in Fig. 2. Figure 3 shows the electric field distributions for SIW, HMSIW, and QMSIW resonators at 3.4, 6, and 3.35 GHz, corresponding to the $TE_{110}^{FMSIW}$, $TE_{110}^{HMSIW}$, and $TE_{110}^{QMSIW}$, respectively. The resonating frequencies of HMSIW and QMSIW resonators slightly differ with respect to FMSIW due to the fringing field at the open edges. The electric field distribution of the QMSIW without slot in TE ₁₁₀ mode at 3.35 GHz and the frequency response are depicted in Fig. 4. Next, a slot is introduced on the top side of the QMSIW cavity to obtain two EMCRs for radiating at two different frequency bands. The resonating frequencies of top-side and bottom-side EMCRs can be determined by:

(9)$$f_{mn0}^{Top EMCR} = {1\over {2\pi \sqrt { \varepsilon_r } }}\sqrt {\left({{{m\pi }\over {2W_{1} }}} \right)^2 + \left({{{n\pi }\over {2L_{4} }}} \right)^2}$$

(10)$$f_{mn0}^{Bottom EMCR} = {1\over {2\pi \sqrt { \varepsilon_r } }}\sqrt {\left({{{m\pi }\over {2L_{1} }}} \right)^2 + \left({{{n\pi }\over {2L_{3} }}} \right)^2}$$

Fig. 2. Conversion of SIW to QMSIW cavity.

Fig. 3. E-field distribution of SIW, HMSIW, and QMSIW cavities.

Fig. 4. (a) E-field distribution of the QMSIW without slot in TE ₁₁₀ mode at 3.35 GHz and (b) its frequency response.

Where m = n = 1, 2,…, and ε_r be the relative permittivity of the substrate. The size of the EMCRs are made different to achieve distinct frequency bands. The top-side and bottom-side EMCRs are responsible for radiation at 2.6 and 4.9 GHz. These frequency bands can be designed independently by varying the dimensions of the EMCRs. The full-wave performances of the proposed antenna are depicted in Fig. 5. The antenna achieves better than −20 dB return loss and more than 34 dB port isolation. High isolation is achieved due to the weak cross-coupling path between two 50Ω orthogonal feed lines. The electric and magnetic field distributions are illustrated in Figs 6 and 7, respectively. It is seen that the maximum E-fields are observed at the outer-edge of the top side and bottom side of the EMCRs.

Fig. 5. Simulation responses of the proposed QMSIW antenna.

Fig. 6. Electric-field distributions. (a) At 2.6 GHz. (b) At 4.9 GHz.

Fig. 7. Magnetic-field distributions. (a) At 2.6 GHz. (b) At 4.9 GHz.

The variation of radiating bands due to the parameters L ₂, L ₃, and L ₄ are studied. To reconstruct the radiating bands (f ₁ and f ₂), the parameters L ₂, L ₃, and L ₄ can be varied individually/simultaneously as per the application requirements. Due to the variation of dimension of the parameters L ₂, L ₃, and L ₄, the capacitive loading on the cavity is modified which results in tuning of radiating frequencies. The tuning of first radiating band is determined by varying the parameters L ₂ and L ₄ as shown in Fig. 8. It is observed that the first radiating band is shifted toward left when the parameters L ₂ and L ₄ increase without altering the second radiating band. The first radiating band can be tuned in the frequency range of 2.56–2.78 when L ₂ (L ₄) varied from 6.0 to 4.0 mm (23.9 to 21.5 mm). Similarly, the designing of second radiating band is realized by varying the parameter L ₃ as depicted in Fig. 9. From the figure, it is seen that the second radiating band is shifted toward right when the parameter L ₃ decreases without any effect on the first radiating band. The second operating band can be tuned in the frequency range of 4.87–5.13 when the parameter L ₃ varied from 12.3 to 10.7 mm. Therefore, the radiating frequency bands can be designed independently/simultaneously by varying the dimensions of the EMCRs as per applications.

Fig. 8. Tuning of first resonating frequency. (a) Due to L ₂. (b) Due to L ₄.

Fig. 9. Tuning of second resonating frequency.

Based on the above studies, a simple design approach has been suggested.

1. (1) Select the dimension of the FMSIW cavity by following the design equations for resonant frequency (1)–(3).

2. (2) Produce HMSIW by symmetrical cutting of the FMSIW. This HMSIW is then cut symmetrically to form QMSIW.

3. (3) The resonant frequencies of the HMSIW and QMSIW are determined by using the equations (4)–(8).

4. (4) Insert a slot on the top surface of the QMSIW to construct two EMCRs which exhibit the self-diplexing characteristic.

5. (5) Select the dimension of the top EMCR responsible for first radiating band as 0.22λ_g × 0.29λ_g.

6. (6) Select the dimension of the top EMCR responsible for second radiating band as 0.28λ_g × 0.50λ_g.

7. (7) Optimize the parameters L _a and L _b to achieve proper impedance matching.

8. (8) Vary the parameters L ₂, L ₃, and L ₄ for designing of radiating frequency bands independently.

9. (9) Repeat last two steps to obtain the desired antenna characteristics.

Fabrication, measurement, and comparative analysis

To support the theoretical analysis, the proposed antenna is fabricated and demonstrated for LTE and PSB applications. Figure 10 shows the photograph of the fabricated antenna. All the experiments are carried out by employing excitation at one port and terminating the other port by 50Ω load. The input matching and isolation are measured using Rohde and Schwarz vector network analyzer. Figure 11 depicts the performances of the full-wave simulated and measured reflection coefficients (|S ₁₁| and |S ₂₂|) and isolation (|S ₁₂|). The far-field performances of the fabricated prototype in two-orthogonal planes of Φ = 0^° and Φ = 90^° at 2.6 and 4.9 GHz are demonstrated inside an anechoic chamber. The full-wave simulated and measured peak gains and radiation efficiency are illustrated in Figs 12 and 13, respectively. The full-wave simulated and measured radiation patterns of the fabricated antenna at 2.6 and 4.9 GHz are depicted in Figs 14 and 15, respectively. The measured and full-wave simulated performances are very consistent as expected. However, very small deviation is seen between full-wave simulated and measured performances due to fabrication tolerance and connector loss. The performances of the fabricated prototype are summarized as follows:

1. (1) Simulation

   1. (a) The input return losses |S ₁₁| and |S ₂₂| are better −22 dB.

   2. (b) The port isolations are 36.07 and 36.13 dB at 2.6 and 4.9 GHz, respectively.

   3. (c) The full-wave peak gains are 5.34 and 5.56 dBi at 2.6 and 4.9 GHz, respectively.

   4. (d) The radiation efficiencies of the proposed antenna are 89.1and 95.7% at 2.6 and 4.9 GHz, respectively.

   5. (e) The FTBR and co-to-cross polarization level are better than 22 and 20 dB, respectively.

2. (2) Measurement

   1. (a) The return losses |S ₁₁| and |S ₂₂| of the fabricated antenna are > −21 dB at all the radiating frequencies.

   2. (b) The port isolations (|S ₁₂|) of the fabricated antenna are found to be 36.69 and 35.23 dB at 2.6 and 4.9 GHz, respectively.

   3. (c) The measured peak gains are 5.12 and 5.59 dBi at 2.6 and 4.9 GHz, respectively.

   4. (d) The measured radiation efficiencies of the fabricated antenna are 90.2 and 95.4% at 2.6 and 4.9 GHz, respectively.

   5. (e) The FTBR and co-to-cross polarization level are very consistent to the full-wave simulated performances.

Fig. 10. Fabricated prototype. (a) Front view. (b) Back view.

Fig. 11. Full-wave simulated and measured magnitude responses of the suggested SDA.

Fig. 12. Full-wave simulated and measured gain.

Fig. 13. Full-wave simulated and measured efficiency.

Fig. 14. Full-wave simulated and measured radiation pattern at 2.6 GHz. (a) E-plane (ϕ = 0^°). (b) H-plane (ϕ = 90^°).

Fig. 15. Full-wave simulated and measured radiation pattern at 4.9 GHz. (a) H-plane (ϕ = 0^°). (b) E-plane (ϕ = 90^°).

A comparative analysis is discussed on performance indicators (size, isolation, peak gain, return loss and FTBR) between the proposed antenna and previously reported works. Table 1 illustrates the performance comparison of state-of-the-art dual-frequency antennas. Compared to [Reference Zhang, Hong, Zhang and Wu8–Reference Barik, Cheng, Dash, Pradhan and Karthikeyan15], the proposed prototype is ultra-compact. The size of the manufactured antenna is 74.3% smaller than the most compact self-diplexing antenna reported in [Reference Barik, Cheng, Dash, Pradhan and Karthikeyan14]. The proposed antenna achieves highest port isolation when compared with [Reference Zhang, Hong, Zhang and Wu8–Reference Barik, Cheng, Dash, Pradhan and Karthikeyan15]. The antenna exhibits good peak gain, FTBR, and co-to-cross polarization level when compared with existing works in [Reference Zhang, Hong, Zhang and Wu8–Reference Barik, Cheng, Dash, Pradhan and Karthikeyan15]. Therefore, the proposed self-diplexing cavity-backed QMSIW slot antenna is suitable for dual-frequency compact communication systems.

Table 1. Performance comparison between proposed and previously reported self-diplexing antennas