Antenna design and development

Geometry

The proposed dual-band quasi-Yagi radiator is illustrated in Fig. 1. The antenna design is attempted to address the requirements of IoMT applications, where reliable on-body communication is critical. The proposed quasi-yagi radiator is designed on a polyimide material. The dielectric constant is 3.5, while the power loss tangent is 0.0023. The substrate thickness of 50 µm ensures excellent flexibility and conformal properties suitable for on-body antennas. The choice of polyimide material for the proposed conformal antenna is driven by its exceptional flexibility, low dielectric loss, and mechanical durability, making it ideal for wearable and body-mounted applications in IoMT and WBAN environments. The polyimide material is copper-clad on both sides, and the copper layer is patterned to make it frequency dependent while the rear side is patterned to offer the excitation line. The antenna during the developmental stages is detailed in Fig. 2. The corresponding S-parameters at various stages of development are described in Fig. 3. The simulations are performed in the frequency domain solver of Computer Simulation Technology Software with accuracy set to 10⁻⁶.

Figure 1.

Proposed wedge-shaped dual band antenna (a–c) different views.

The design begins with a dual-radiating arm design, as illustrated in Fig. 2(a). These radiating arms are wedge-shaped, with the total length of the radiators (2l1 + g1) chosen to be approximately half a wavelength at 3.4 GHz. The radiator on the top is fed by an L-shaped feedline, which transfers microwave energy through a narrow slot line, as shown in Fig. 2(a). The initial design’s bandwidth performance is constrained by the antenna’s profile thickness. To address this limitation, several strategies are implemented to enhance bandwidth properties.

Figure 2.

Developmental stages of the antenna: (a) basic radiator, (b) feedline terminated with square patch, (c) ground modified radiator, (d) modified radiating arms, and (e) proposed antenna.

First, an L-shaped feed transmission line is used as a feed to improve the antenna’s reflection coefficient bandwidth. The antenna in Fig. 2(a) exhibits resonant characteristics, as depicted in Fig. 3. Impedance matching at the spot frequency is further enhanced by introducing a square patch termination in the L-shaped feedline as in Fig. 2(b). Although this alteration strengthens the impedance matching, the antenna’s performance remains below acceptable levels. To overcome this limitation, modifications are made to the ground as in Fig. 2(c). The ground conductor is shaped and blended with a radius “R” to achieve optimal matching conditions. This adjustment produces a resonance around 3.15 GHz with |S₁₁|_min ≤ −35 dB. The primary research objective is to achieve dual-band resonance. To this end, the number of wedge-shaped resonators is increased progressively from one to two and then to three, as shown in Fig. 3. Adding two wedge-shaped radiators, as depicted in Fig. 2(d), introduces dual-band resonance characteristics, albeit with a weak resonant mode at the upper frequency. The impedance matching is further improved by incorporating a third pair of wedge-shaped radiating arms and a semicircular director, as shown in Fig. 2(e). This final antenna geometry successfully achieves the desired dual-frequency characteristics, providing FBWs of 8.16% and 8.59%, respectively. These dual-frequency bands meet the bandwidth requirements for the transmission of medical data. Table 1 summarizes the optimized geometrical parameters of the wedge shaped quasi-Yagi antenna.

Figure 3.

|S₁₁| characteristics of the wedge shaped quasi-Yagi antenna.

Table 1.

Optimized antenna parameters (unit millimeter)

The operation of the proposed dual-band antenna is further illustrated with a help of an equivalent circuit diagram. The equivalent circuit of the proposed antenna described in Fig. 4 consists of two parallel resonant circuits (R₁-L₁-C₁ and R₂-L₂-C₂) indicating the resonant frequencies 2.45/3.4 GHz of the Yagi-Uda antenna with wedge shaped radiators. The losses associated with the antenna are represented by R _loss. The antenna is connected to an excitation source as shown in Fig. 4. By effectively adjusting the values of L₁-C₁ and L₂-C₂, the resonant frequencies of the proposed dual-band antenna can be controlled. The inductance (L _n) in the equivalent circuit is primarily due to the effective current path length of the wedge-shaped radiator, which influences the antenna’s resonant frequency. A longer path increases L _n, lowering the resonant frequency, while a shorter path reduces L _n, shifting the resonance higher. The capacitance (C _n) arises from the gap and spacing between the radiator and the ground plane, as well as the coupling between the parasitic director elements in the Yagi configuration. A smaller gap results in higher capacitance, which helps in impedance matching and bandwidth control. Together, L _n and C _n govern the dual-band resonance, ensuring efficient radiation and frequency selectivity for 2.45 and 3.4 GHz operations.

Figure 4.

Equivalent circuit model for the proposed dual-band antenna.

Development of MIMO antenna

Building upon the proposed antenna, a multi-element scheme is developed to enhance communication performance, particularly for applications in on-body communications and IoMT. The MIMO configuration shown in Fig. 5 replicates the same wedge-shaped radiators and structural modifications discussed earlier, with 10 mm edge-to-edge separation. The total area of the proposed eight-element scheme is 125 mm × 115 mm, ensuring a compact footprint suitable for wearable and body-mounted devices. The antenna elements are distributed along the edges to form an octagonal arrangement, which optimizes spatial diversity and minimizes mutual coupling. The design features a connected ground scheme, which inherently provides mutual coupling less than −15 dB between elements as shown in Fig. 6, eliminating the need for additional electromagnetic structures to suppress mutual coupling. The central portion of the antenna structure is designed as an octagonal free space, allowing for the routing of cables or other components without interfering with the antenna’s radiation performance. This geometric distribution not only achieves the desired dual-band operation but also supports the reliable transmission of medical data across multiple channels. The compact size, dual-band characteristics, and robust performance in close-proximity scenarios make this design highly relevant for IoMT applications for real-time medical data transmission.

Figure 5.

Eight-element MIMO antenna (a and b) different views. The red lines in (b) indicate the shape in the front side.

Figure 6.

Mutual coupling characteristics with Port 1 as reference.

The surface current contour, as illustrated in Fig. 7, demonstrates the independent operation of each antenna element, ensuring minimal interference and strong performance. When a specific antenna element is excited, the surface current is highly concentrated along its radiating arms, feedline, and ground plane. This localized current concentration ensures optimal radiation and impedance matching for the excited antenna. In contrast, the neighboring antenna elements exhibit negligible current strength, which confirms the low mutual coupling achieved by the design. This behavior is primarily due to the connected ground scheme and the strategic octagonal distribution of the elements, which effectively decouples the radiating fields of adjacent antennas. The minimal interaction between antenna elements ensures that each operates independently, preserving the integrity of the MIMO scheme and enabling reliable dual-band performance for IoMT applications. This intrinsic isolation is critical for maintaining consistent signal quality and efficient data transmission in multichannel communication systems.

Figure 7.

Surface current distribution during the excitation of different ports in the MIMO scheme at 2.45 GHz. (a–h) Excitation of Ports 1–8.

Fabrication and measurement

The prototype MIMO antenna is fabricated using the conventional printed circuit board (PCB) manufacturing process. The performance of the antenna in real-time is verified using real-time measurements. The measurement is conducted using Keysight E5071C Vector Network Analyzer. N4432D electronic calibration kit from Keysight Technologies is used for fine calibration. The prototype antenna is manufactured using photo-etching method. The developed antenna in the measurement environment is shown in Fig. 8. The fabricated prototype pictures are described in the inset of Fig. 8. The measurement S-parameters of the proposed eight-element antenna is illustrated in Fig. 9(a) and (b). The measured reflection and transmission coefficients are in close correlation with the simulation results as evidenced through Fig. 9. The measured FBW of the antenna is more than 8% in both the lower frequency and upper frequency band centered at 2.45 and 3.4 GHz, respectively. The mutual coupling characteristics described in Fig. 9(b) shows a strong decoupling between the ports of the eight-element antenna. In all the port combinations, the inter-port isolation is greater than 15 dB in both frequency bands.

Figure 8.

Measurement setup showing the fabricated prototype and the measurement window.

Figure 9.

(a) Reflection coefficient characteristics. (b) Mutual coupling characteristics.

The prototype antenna is also measured in the presence of human body to validate its performance for WBAN applications. The simulated and measured on-body performance of the antenna is illustrated in Fig. 10. During simulations, a three-layer human body model is used for the simulations. The layer consists of skin (ε_r = 38; σ = 1.46 S/m), fat (ε_r = 10.8; σ = 0.268 S/m) and muscle (ε_r = 52.7; σ = 1.74 S/m) as shown in Fig. 10(a). The thickness of each layer is fixed as 1 mm, 5 mm and 100 mm respectively and the simulation analysis is carried out. The simulated and measured reflection coefficient characteristics are presented in Fig. 10(b). In order to retain the original characteristics of the antenna, the antenna and the human body are separated by a distance of 1 cm using foam material. In reality, the antenna will be positioned on a platform whose base is at least 1 cm thick holding the PCB with necessary electronics. This arrangement reduces the direct interaction of the human body with the antenna. In line with this, the spacing between the antenna and the human body is chosen as 1 cm. Further, for the sake of experiment, a female subject of age 37 is chosen. The antenna is located on the abdomen region where the antenna is maximally flat and in the arm region where the antenna undergoes some bending effects. In both the cases, the antenna exhibited dual-band response with a slight frequency shift. However, under flat and bent conditions, the required bandwidth specification to perform medical information exchange is satisfied making the antenna suitable for WBAN communications.

Figure 10.

(a) On-body simulation setup. (b) Reflection coefficient characteristics.

The antenna radiation pattern is measured in an anechoic chamber and the estimated radiation patterns in simulation and measurement are depicted in Fig. 11. The proposed eight-element antenna array exhibits unidirectional radiation pattern along specific direction. The port pairs producing oppositely directed radiation patterns are Ports 1–5, Ports 2–6, Ports 3–7, and Ports 4–8. Thus, the proposed scheme offers pattern diversity connecting to different sensor elements deployed along the surface of the human body. The peak measured gain of the antenna is 5.72 dBi at 2.45 GHz and 6.13 dBi at 3.4 GHz. The simulated peak antenna efficiency is greater than 80% in both the realized frequency bands. A plot of antenna gain, efficiency versus the frequency is described in Fig. 12.

Figure 11.

Normalized radiation pattern: (a)–(d) At 2.45 GHz. (e)–(h) At 3.4 GHz.

Figure 12.

Gain/efficiency versus frequency behavior of the MIMO radiator.

Evaluation of diversity metrics

The MIMO/diversity metrics is analyzed to validate its suitability for IoMT and 5G applications. Standard equations, as referenced in paper [Reference Sharawi23], are employed to calculate key parameters of the MIMO scheme. These calculations show the MIMO scheme’s capability to deliver high performance in terms of gain, diversity, and reliability, ensuring efficient communication for on-body sensor networks.

(1)\begin{equation}{\rho _e} = \frac{{|\iint {\left( {\mathop {\mathop F\nolimits_1 }\limits^ \to (\theta ,\phi ) \bullet \mathop {\mathop F\nolimits_2 }\limits^ \to (\theta ,\phi )} \right)d\Omega }{|^2}}}{{\iint {|\mathop {\mathop F\nolimits_1 }\limits^ \to (\theta ,\phi ){|^2}d\Omega \iint {|\mathop {\mathop F\nolimits_2 }\limits^ \to (\theta ,\phi ){|^2}d\Omega }}}}\end{equation}

In Equation (1), the field components (F ₁ and F ₂) along the 3D planes are utilized to assess the Envelope Correlation Coefficient (ECC) of the diversity scheme for different port combinations in the eight-element MIMO antenna. The Apparent Diversity Gain (ADG) is evaluated using the ECC through Equation (2).

(2)\begin{equation}{G_{app}} = 10 \times \sqrt {1 - \left| {{\rho _e}} \right|} \end{equation}

To evaluate the actual diversity performance of the eight-element dual-band MIMO radiator, the losses associated with the radiation are factored into the ADG as outlined in Equation (3) and specified as Effective Diversity Gain (EDG).

(3)\begin{equation}{G_{eff}} = {\eta _{total}} \times {G_{app}} = {\eta _{total}} \times 10 \times \sqrt {1 - \left| {{\rho _e}} \right|} \end{equation}

The Mean Effective Gain (MEG) determines the overall gain characteristics in the context of a multi-antenna system. The evaluation of MEG takes into account the cross-polarization ratio (XPR) as given in Equation (4). For isotropic conditions, the XPR is assumed to be 0 dB, implying ideal conditions with no polarization discrimination. In more realistic settings, the XPR is adjusted to 1 dB for indoor environments, where walls and other obstacles can affect signal polarization, and 5 dB for outdoor scenarios, where environmental factors such as buildings and foliage can introduce more significant polarization effects. Additionally, normal power distribution, with µ and σ of 10° are assumed during the calculation. This assumption accounts for the typical variation in signal propagation and antenna alignment, reflecting realistic conditions in both indoor and outdoor environments. By incorporating these factors, the MEG calculation ensures a comprehensive and accurate assessment of multi-antenna system performance for effective design suitable for real-world applications.

(4)\begin{equation}MEG = \int\limits_0^{2\pi } {\int\limits_0^\pi {\left[ \begin{gathered} \frac{{XPR}}{{1 + XPR}}{G_\theta }(\theta ,\phi ){P_\theta }(\theta ,\phi ) \hfill \\ + \frac{1}{{1 + XPR}}{G_\phi }(\theta ,\phi ){P_\phi }(\theta ,\phi ) \hfill \\ \end{gathered} \right]} } \sin \theta d\theta d\phi \end{equation}

The estimated MIMO metrics for different scenarios (Isotropic (XPR = 0 dB); Indoor (XPR = 5 dB) and Outdoor (XPR = 1 dB)) are presented in Table 2. The far-field based ECC is significantly below 0.25 ensuring minimal correlation between antenna elements and excellent isolation. The ADG and EDG values are above 7.4, demonstrating the antenna’s ability to maintain high diversity performance under varying conditions. Additionally, the ratio of MEG is substantially lesser than −0.6 dB, further confirming the efficient gain characteristics of the proposed antenna system. These results highlight the effectiveness of the design in supporting reliable and robust MIMO performance, making it desirable for high throughput communications such as IoMT and 5G wireless communication.

Table 2.

MIMO performance parameters at 2.45 and 3.4 GHz using measurement data

State-of-the-art comparison

The proposed dual-band conformal antenna is compared with the existing dual-band MIMO antennas particular used from MIMO applications. Table 3 presents the detailed comparison of the quasi-Yagi antenna.

Table 3.

Comparison with the state-of-the-art

The following are the inferences derived from the Table 3:

1. 1. The proposed antenna has an element size of 0.38λ_eff × 0.52λ_eff, which is smaller than many other designs, such as those in papers [Reference Musa, Shah, Majid, Mahadi, Rahim, Yahya and Abidin7, Reference Althuwayb, Alibakhshikenari, Virdee, Rashid, Kaaniche, Atitallah, Armghan, Elhamrawy, See and Falcone9, Reference Sid, Cresson, Joly, Braud and Lasri10, Reference Adam, Rahim, Yasin and Nasrol21, Reference Sharawi23, Reference Saeidi, Al-Gburi and Karamzadeh24].

2. 2. The proposed antenna has an extremely low profile height of 0.00076λ_eff, much lower than other designs reported in papers [Reference Musa, Shah, Majid, Mahadi, Rahim, Yahya and Abidin7–Reference Noghanian11, Reference Rajawat, Mohta, Sharma and Gupta20–Reference Saeidi, Al-Gburi and Karamzadeh24]. This makes the proposed antenna highly suitable for wearable and space-constrained applications.

3. 3. The antenna reported in this research operates at 2.45 and 3.4 GHz, with FBW of 8.16% and 8.59%, which is higher than many other antennas reported in papers [Reference Musa, Shah, Majid, Mahadi, Rahim, Yahya and Abidin7, Reference Sid, Cresson, Joly, Braud and Lasri10, Reference Adam, Rahim, Yasin and Nasrol21–Reference Sharawi23]. Although paper [Reference Althuwayb, Alibakhshikenari, Virdee, Rashid, Kaaniche, Atitallah, Armghan, Elhamrawy, See and Falcone9] offers higher FBW, the proposed work maintains a balance between FBW and compactness.

4. 4. The reported antenna provides 5.72 and 6.13 dBi of gain, which is comparable to papers [Reference Althuwayb, Alibakhshikenari, Virdee, Rashid, Kaaniche, Atitallah, Armghan, Elhamrawy, See and Falcone9] and [Reference Adam, Rahim, Yasin and Nasrol21]. While paper [Reference Althuwayb, Alibakhshikenari, Virdee, Rashid, Kaaniche, Atitallah, Armghan, Elhamrawy, See and Falcone9] offers a higher peak gain, the proposed antenna achieves a good balance between gain and efficiency, which is higher than many others antennas [Reference Althuwayb, Alibakhshikenari, Virdee, Rashid, Kaaniche, Atitallah, Armghan, Elhamrawy, See and Falcone9, Reference Sid, Cresson, Joly, Braud and Lasri10, Reference Adam, Rahim, Yasin and Nasrol21, Reference Kang, Tak and Choi22, Reference Saeidi, Al-Gburi and Karamzadeh24].

5. 5. The proposed antenna is replicated to implement an 8 elements for MIMO, which is the highest number of elements in this comparison, outperforming papers [Reference Althuwayb, Alibakhshikenari, Virdee, Rashid, Kaaniche, Atitallah, Armghan, Elhamrawy, See and Falcone9] and [Reference Noghanian11]. Thus, the proposed solution has a significant advantage in terms of data throughput and communication reliability.

In summary, the proposed dual-band eight-element MIMO antenna has attractive properties such as small element size and highly compresses profile height, large FBW, good gain and efficiency with enhanced channel capacity. These aspects make it highly suitable for modern applications like 5G, WBANs, and wearable devices, especially when compared to other designs that may prioritize larger size or fewer MIMO elements.