Design of two-element antenna

The proposed two-element, textile, CP antenna and its fabricated prototype are represented in Figs 1(a) and 1(b).

Fig. 1. Textile two-element antenna (a) prototype design (b) hardware (c) design steps (d) |S ₁₁|/|S ₂₂| of design steps (e) |S ₁₂|/|S ₂₁| of design steps (f) axial ratio (dB) in different steps.

The copper tape of 0.057 mm thickness is used as a conducting element whereas jeans material (loss tangent (tan δ) = 0.025) with permittivity (ɛ_r) of 1.7 and thickness (t) of 1 mm is used as a substrate. The width (W_s) and length (L_s) of the substrate are 31 × 31 mm². The dimensions of the circularly polarized textile antenna are depicted in Table 1. The proposed antenna evolution is discussed in five steps, which are illustrated in Fig. 1(c). In step-1, two orthogonally placed microstrip step-feed is accomplished with the defected ground. The impedance bandwidth (|S ₁₁|) and isolation (|S ₂₁|) are achieved from 3.2 to 4.8 GHz and 11 to 15 dB respectively, as depicted in Figs 1(d) and 1(e). In step-2, width W1 is increased with symmetrically open stub on both sides in the ground, which perturbs the surface current in the ground and supports circular polarization. The axial ratio in Fig. 1(f) reveals that the width (W1) and open stub in the ground shift the axial ratio in the operating band close to 3 dB whereas it is less than 3 dB at 3.2 GHz. The impedance bandwidth (|S ₁₁|) varies from 3.6 to 4.4 GHz in this step with isolation (|S ₂₁|) more than 13.5 dB. In step-3, a diagonal stub is accomplished in the ground to diminish the surface current between ports, which improves more than 5 dB isolation (|S ₂₁|) along with axial ratio as illustrated in Fig. 1(f). The impedance bandwidth (|S ₁₁|) of the antenna is achieved from 3.3 to 4.3 GHz with isolation (|S ₂₁|) more than 18.5 dB. In step-4, two symmetric L-shaped open-stubs are incorporated between the ports to further enhance the isolation (|S ₂₁|) and axial ratio. The impedance bandwidth (|S ₁₁|) is obtained from 3.4 to 4.4 GHz with 18.4 to 21.9 dB isolation. In step-5, impedance matching is enhanced with a rectangular slot introduced in the ground behind the step feed, which improves the impedance bandwidth (|S ₁₁|) and isolation (|S ₂₁|) of the two-element antenna. The impedance bandwidth (|S ₁₁|) of the antenna varies from 3.3 to 4.3 GHz with more than 24 dB isolation and the less than 3 dB axial ratio is achieved in the entire operating band.

Table 1. Dimensions of different geometrical parameters

Analysis of circular polarization and surface current of two-element antenna

In this section, axial ratio, surface current and radiation characteristics are analyzed to understand the antenna performance. It is clear from the previous section that the defected ground with open-stubs as well as the position of feed perturbs the surface current, which enhances the isolation (|S ₂₁|) and axial ratio (dB) for achieving the CP band. The obtained 3 dB axial ratio covers the entire operating band. The vector analysis of surface current is investigated at 0°, 90°, 270° and 360° phase in 3.8 GHz as depicted in Fig. 2(a). It is revealed that the direction of the current vectors in the ground is clockwise as well as anti-clockwise which validates LHCP and RHCP in the antenna. The surface current at 3.8 GHz in port-1 is excited and is represented in Fig. 2(b), where the open-stubs mitigate the surface current moving from one port to another in a clockwise and anti-clockwise direction, which improves the isolation (|S ₂₁|), and reduce the mutual coupling of the antenna. The 3D radiation pattern at 3.8 GHz is illustrated in Fig. 2(c), the figure reveals the diversity performance of the two-port antenna and its ability to transmit and receive the EM waves from different directions.

Fig. 2. Textile two-element diversity antenna (a) CP behaviors analysis form surface current (b) surface current when port-1 is excited (c) 3D radiation pattern.

Simulated and experiment results

The CP and two-element antenna is measured on Anritsu MS2025B VNA to verify the simulation results obtained from HFSS 13.

The 10 dB impedance bandwidth (|S ₁₁|) varies from 3.3 to 4.3 GHz while the isolation (|S ₂₁|) is more than 24 dB as depicted in Fig. 3(a), where the measured results are close to simulated results that validate the results. A minor deviation is found between simulated and measured results due to some irregularities in soldering and fabrication. The realized gain (dB) of the antenna varies from 3 to 3.45 dB and less than 0.5 dB gain variation is achieved in all operating bandwidths as illustrated in Fig. 3(b). The axial ratio (3 dB) is illustrated in Fig. 3(c), which reveals that less than 3 dB measurement and simulated results are achieved. The measured and simulated E and H components (radiation pattern) of an antenna are 90 degrees out of phase at 0°, 90°, 180° and 270° degree which validates the circular polarization behavior of the antenna. The radiation efficiency (%) is achieved more than 90 in the entire operating band. In this paper, equation (1) is used to calculate the transmission efficiency [Reference Haerinia and Noghanian13] of the proposed CP antenna where the Tx antenna or Rx antenna is analyzed with a variation of 10 mm over a distance of 10 to 50 mm at 3.8 GHz. The analysis of transmission coefficient (|S ₃₁|) in dB at different distances is shown in Fig. 3(e). The Antenna-1 is the Tx antenna and Antenna-3 is the Rx antenna because the Antenna-1 and antenna-2 are accomplished in transmitter antenna where in Ref [Reference Haerinia and Noghanian13] paper, Antenna-1 is the transmitter and antenna-2 is the Rx antenna. The fractional variation of transmission efficiency from 10to 50 mm is 2.94% while for 20 to 50 mm is 11.3%, which makes it appear that the antenna can work effectively with minimum variation in transmission efficiency.

(1)$$\eta = ( {{\vert {S_{31}} \vert }^2} ) \times 100\% $$

Fig. 3. Textile two-element diversity antenna (a) simulated and measured s-parameters (b) realized gain (dB) (c) axial ratio (dB) (d) radiation efficiency (%)(e) Transmission coefficient (S ₂₁) at 3.7 GHz with Rx antenna distance varies from 10 to 50 mm (f) Measurement set-up (g) and (h) normalized radiation pattern at 3.8 GHz (i) Measurement of antenna on the human body (j) |S ₁₁| and |S ₁₂| analysis of antenna on the human body (Simulated and measured).

In design, both antenna elements are the same, therefore measurement is done in Ant-1 and another element is terminated with a matched load of 50 Ω in an anechoic chamber where the distance from reference antenna to under test antenna is 3 meter as depicted in Fig. 3(f), ridge type horn antenna used as a reference antenna with more than 11 dB gain at operating bandwidth. The LHCP and RHCP normalized radiation patterns at 3.8 GHz in the xz and yz planes are illustrated in Figs 3(g) and 3(h), where a stable radiation pattern is observed.

The two-port antenna is analyzed on the chest, wrist, thigh and shoes at 20 mm distance from the antenna where foam padding is used to separate the antenna from the human body which does not affect the s-parameter as depicted in Fig. 3(i). The simulation analysis is conducted on the human wrist phantom model. The electrical characteristics of the human wrist phantom model is taken from Ref [15] and the thickness of skin, fat, and muscle are used 2, 5, and 10 mm with 50 mm × 50 mm dimensions. The distance of the phantom model from an antenna is 20 mm to remove the coupling effect as well as a minimum effect on the isolation of the antenna. The measured and simulated results of |S ₁₁| and |S ₁₂| parameters are illustrated in Fig. 3(j), where small variation is found in impedance bandwidth (|S ₁₁|) and isolation is higher than 20 dB in all on-body measurements. The antenna elements are orthogonally placed in the design; therefore, the antenna provides the pattern diversity, which can be observed from the radiation patterns. The diversity performance of the antenna is also determined from the MIMO parameters, therefore ECC and diversity gain (DG) are analyzed with radiation pattern from equations (2) to (3) [Reference Jha, Kumar, De and Jain16] whereas total active reflection coefficient (TARC) and channel capacity loss (CCL) are extracted from the simulated and measured s-parameter and computed from equations (4) to (5) [Reference Kumar3].

(2)$$ECC = \displaystyle{{{\left| {\left( {\int_0^{2\pi } {\int_0^\pi {(XPR\; E_{\theta 1}E_{\theta 2}^* P_\theta + \; E_{\varphi 1}E_{\varphi 2}^* P_\varphi )} } } \right){\rm Sin}\theta d\theta d\varphi } \right|}^2} \over \matrix{{\int }_0^{2\pi } \int_0^\pi \left( {XPR\; E_{\theta 1}E_{\theta 1}^* P_\theta + \; E_{\varphi 1}E_{\varphi 1}^* P_\varphi } \right){\rm Sin}\theta d\theta d\varphi \times \hfill \cr \int_0^{2\pi } {\int_0^\pi {\left( {XPR\; E_{\theta 2}E_{\theta 2}^* P_\theta + \; E_{\varphi 2}E_{\varphi 2}^* P_\varphi } \right)} } {\rm Sin}\theta d\theta d\varphi \hfill} }$$

(3)$$DG = 10\sqrt {1-{\vert {ECC} \vert }^2} $$

(4)$$TARC = \displaystyle{{\sqrt {( {\vert {S_{11} + S_{12}e^{\,j\theta }} \vert }^{2{\rm \;}}} + {\vert {S_{22}e^{\,j\theta } + S_{21}} \vert }^{2{\rm \;\;}}) } \over {\sqrt 2 {\rm \;}}}$$

(5)$$CCL = {-}\log _2\det ( {\psi^R} ) $$

The ECC represents a correlation of the radiation pattern of antenna elements and its practical accepted value should be less than 0.5 where DG should be higher for practical applications. The extracted simulated and measured ECC and DG is < 0.2 and close to 10 dB respectively, therefore good diversity performance of the two-port textile antenna is achieved as depicted in Figs 4(a) and 4(b). The TARC and CCL of MIMO antenna parameters are extracted from equations (3) and (4), where the correlation matrix at receiving end is represented by ψ ^R. The results reveal that the minimum variation is observed in the measured value of the TARC at different phase angles whereas measured and simulated CCL is less than 0.35 bits/sec/Hz in the entire operating band (Figs 4(c) and 4(d)). The CCL is less than the accepted value of 0.4 bits/sec/Hz [Reference Kumar3]. Table 2 represents the comparative analysis with recently published textile or flexible single element and two-element antenna. The analysis reveals that the proposed antenna is compact, low cost and flexible with low ECC. The diversity performance and CP is achieved in the entire operating band, however, it can be a good candidate for the 5G, wearable and biomedical applications.

Fig. 4. Textile two-element diversity antenna (a) ECC (b) DG (c) TARC (d) CCL.

Table 2. Comparison analysis of proposed antenna with existing antenna

Simulated analysis of breast cancer detection

In this section, simulation analysis is investigated on the human female breast model and standard parameters of skin, fat, glandular tissue, as well as a tumor cell, are used [15]. The proposed antenna is placed in the broadside direction at a 5 mm distance from the model for maximum power penetration. The differences in the local and average SAR values are detected in the female breast model at 1W incident power on 1 g tissues with an 8 mm tumor radius. The average and local (maximum and minimum) SAR (W/Kg) of the skin, fat, glandular, and tumor cell are depicted in Table 3. The analysis focuses on two major outcomes, according to the first finding, where the difference between the local and average of SAR are achieved with and without the tumor tissue, which can be analyzed in the practical situation with the tumor-affected breast and normal breast. In the second finding, there is a significant difference in the SAR of the tumor tissue and surrounding glandular tissue, so that the tumor tissue can be easily detected from the breast. The local and average SAR with and without tumor is illustrated in Fig. 5. The analysis reveals that the proposed two-port antenna can be used in biomedical cancer detection applications.

Fig. 5. Analysis of female breast model at 1W power on glandular tissue (a) local SAR with tumor (b) local SAR without tumor (c) average SAR with tumor (d) average SAR without tumor.

Table 3. SAR(average) and SAR(Local)Values of Tissue types at 3.8 GHz