1 Introduction

Radiofrequency micro electro mechanical systems (RF MEMS) Technology is a promising technology for communication system because of low loss and high performance. In present years, intense research in MEMS technology devices and its fabrication has been attracted many researchers, academicians, and industrialists towards communication field. There is a great demand for smaller and faster communication systems which need less area and low power consumption [1]. Radio frequency micro electro mechanical systems (RF MEMS) have been dominating and replaced the existing conventional solid-state semiconductor devices such as pin diodes, field effect transistors, etc. These RF MEMS devices are innovative especially in switch design, where the design size varies from 1 to 100 µm. Katehi et al. [2] RF MEMS switches are classified as follows, based on the actuation mechanism: (1) electrostatic, (2) magneto static, (3) piezoelectric, (4) thermal. Depends on the type of contact with the signal line: resistive type or capacitive type, In-circuit configuration perspective: Series type and Shunt type, based on the mechanical movement of the beam either in vertical or lateral movement: Fixed Fixed Beam and Cantilever Beam.

These switches are mainly used in reconfigurable antennas to change its electrical length and frequency characteristics by activating and deactivating the antenna elements. MEMS switches can be used for high-frequency applications in which switch function is based on mechanical movement of the beam. The actuation of MEMS switches has investigated by many types of research since. The first MEMS device switch is fabricated in 1979 by Petersen. The eminent research work brought a rapid development in RF MEMS and it can be efficiently used for Ka-band frequency of (27–40 GHz) for 5G, Satellite communications, and military aircraft applications [3].

Roh et al. reported, the development in the field of telecommunication through improvising the research, particularly in Long Term Evolution (LTE) enhancement techniques like multiple input multiple outputs (MIMO), co-ordinate multipoint (CoMP), heterogeneous networks (HetNets) has been able to manage the traffic requirement for the next few years. The telecommunication industry and the academic institutions have been delivering significant outcomes in fourth generation (4G) LTE to deliver high data rates to the consumers by increasing the number of base stations, merging of spectra, etc. But none of the processes has been able to solve the hurdle of managing the traffic urge estimated for 2020. Hence 5G technology has come into existence. For this 5G technology, wide bandwidths method is proposed. In this patch antenna is take as 3 GHz and at 30 GHz an array antenna is used at communications [4]. Keller et al. proposed a method for decreasing the weight of the total system and minimize the complexity of wafer-scale fabrication. In this, a coplanar waveguide fed with a patch antenna for satellite communication purpose. It is designed to operate at ka-band (27–40 GHz) frequency band having a return loss of 10 dB [5]. Rahim et al. proposed A new ideology for reconfigurable antennas has been implemented with two different configurations i.e. narrowband to narrowband and wideband to narrowband in microstrip antenna and slot type antenna for wireless communication mainly for millimeter wave 5G technology [6]. Hong et al. mentioned the first-time practical solution design considerations for effective results are achieved by the phased-array mesh-grid antenna with dual-polarized beam switching for human cellular handsets. This technology can lead to the experience of an mm-Wave 5G radios designed in a Smart Phone, it is designed at a frequency of 60 GHz to integrated with reconfigurable antenna [7]. Demirel et al. [9] presented the breakdown of the bridge type RF-MEMS switches occur due to the deformation caused by the thermal expansion. In this switch having an actuation voltage of 25 V and isolation less than 20 dB and can be operated at 28–40 GHz frequency band. Hence, it is minimized by the inclusion of anchors to the device design. Korany et al. presented in detail about the dual-band circularly polarized phase antenna was designed for mobile communication which can be operated at 28–38 GHz having a reflection coefficient of − 15 dB and gain greater than 7 dB and an axial ratio less than 3 dB. This can deliver radiation performance such that it does not affect the head-hand proximity of the user. The designed antenna structure can be able to fit in the sleek mobile device and it may be able to use the design in millimeter wave 5G applications [8].

To overcome the drawbacks and to improve the functionality we have proposed an optimization process for the switch structure based on the frequency of application which improves the performance of the switch, in terms of high isolation and low insertion loss and less pull in voltage. The proposed process is step by step from down to top approach. For the proposed method firstly, we extract the dimensions of the switch and done some electromechanical analysis and RF performance of the switch and compared our results with the existing fabricated measured results. During the comparison we have observed that our extracted optimized switch performance is approximately same and it requires less pull in voltage. Thus, if anyone want to go for fabrication directly can through optimization process and extract the required dimensions, then these dimensions can be used for fabrication which indirectly helps in cost effectiveness. The impendence matching depends on the substrate height and material for the substrate and coplanar waveguide. The coplanar waveguide dimensions are fixed based on the impendence matching of 50Ω. The signal line should possess less transmission loss to transmit the RF input signal, so that isolation will be increased and loss of signal transmitting also decreased (return and insertion loss). The pull-in voltage depends on the actuation area and spring constant. The spring constant is directly related to the dimensions of the beam which are length, width, and thickness of the beam. The overlapping area also depends on the dimensions of the beam. By increasing the overlapping area pull in voltage will be reduced. To extract the dimensions of the beam and dielectric an optimization step by step process is followed to get well suitable dimensions. By this, the proposed switch can be operated at high frequencies applications and it can be integrated with an antenna for wireless communications in 5G technology.

The rest of the Paper organized in different sections. In Sect. 2 optimization applied for an existing switch, the proposed switch complete design, optimization process, device working, and dimensions are presented. In Sect. 3 the electromechanical analysis in that capacitance and pull in voltage and spring constant, stress analysis and C–V characteristics, Eigen frequency, RF performance of the switch are discussed. In Sect. 4 the comparison of the performance of new switch with the base paper.

2 Device Optimization and Its Specifications

2.1 Optimization Process

The proposed RF MEMS capacitive shunt switch is designed for 5G wireless applications. The device dimensions are fixed based on the parameters Such as impedance matching, Up capacitance and resonant frequency. The switch is designed using FEM tools and performance is enhanced to work at ka-band frequency range. The RF MEMS capacitive shunt switch mainly contains substrate, coplanar waveguide, the dielectric layer, actuation electrodes, and bridge membrane. Initially, to design a switch at frequency there should be an impedance matching of 50 Ω to allow the RF input signal through the coplanar waveguide. The impedance matching Z0 depends on the spacing between the ground plane and width of the signal line, height of the substrate and dielectric constant of the substrate material (Fig. 1).

Fig. 1
figure 1

A schematic view of substrate and CPW

Full size image

The height of the substrate can be estimated condition is

$$\frac{h}{\lambda } > 0.33$$
(1)

where h represents the height of the substrate, \(\lambda\) is wavelength. The wavelength of the radiating signal line is obtained by

$$\lambda = \frac{C}{{F_{C} }}$$
(2)

C is the velocity of light in vacuum C = 3×108 m/s and FC is the cut off frequency.

The maximum height of the substrate can be calculated from Eq. 1. The cutoff frequency of the waves given by the formula below

$$F_{C} \le \frac{nC}{{4h\sqrt {\varepsilon_{r} - 1} }}$$
(3)

where n is the order of transmission mode and C is the velocity of light in a vacuum \(\varepsilon_{r}\) is the relative dielectric constant of the substrate material and H is the height of the substrate and fc is the cut off frequency.

The height of the substrate is taken from the above calculation and CPW contains a signal line and ground planes on either side of the signal line. The spacing (S) between the ground and signal line and the width of the signal line is evaluated for impedance matching [17]. The pull-in voltage depends on the actuation area of the bridge. The actuation or overlapping area of the bridge can be calculated as A = W × w, such that W and w is the width of the signal line and bridge respectively. The width of the bridge is determined from the upstate capacitance and the resonant frequency of the switch [16]

$$F_{r} = \frac{1}{{2\Pi \sqrt {LC_{U} } }}$$
(4)

where Fr is the resonant frequency which is 38 GHz and L = 0.4 nH developed by the CPW transmission line and CU is the upstate capacitance.

The up capacitance of the switch is, [19]

$$C_{U} = \frac{{\varepsilon_{^\circ } A}}{{g + \frac{{t_{d} }}{{\varepsilon_{r} }}}}$$
(5)

where \(\varepsilon_{^\circ }\) = 8.825 × 10−12 F/m, A = W × w, g is a gap between the beam and dielectric at actuation area and ts is the thickness of the dielectric and \(\varepsilon_{r}\) is a relative dielectric constant of beam material. By solving the above equation w is the width of the beam is obtained and the length of the beam is four times of its width i.e. L = 3 or 4 × w.

The buckling effect depends on the thickness of the bridge and can be estimated through the formula

$$g > \frac{4\sqrt 2 }{3}t$$
(6)

Here g is the gap between the bridge and CPW, based on the gap there will be a high capacitance ratio and better displacement of the beam. The beam displacement highly depends on its thickness (Fig. 2).

Fig. 2
figure 2

Capacitance in Upstate position of the proposed switch

Full size image

2.2 Optimization for 35 GHz

The optimization process is applied to the existing switch proposed by Chu et al. and the enhancement in its performance by optimization process is analyzed for validation. Chu et al. proposed a fixed–fixed switch at the 35 GHz frequency by T matching techniques for impedance matching of 50ohms and the switch dimensions, performance electromechanical analysis and fabricated switch measured values. Here we applied our optimization process to the same fixed-fixed switch and enhanced performance characteristics are compared with the Chu et al. switch. By using this optimization process follows the design of RF MEMS switch on CPW with impedance matching of 50 Ω. The dimensions of the switch are changed with respective of length, width, thickness, and gap. The pull-in voltage is reduced through this optimization and performance of the switch is also approximately increased. The pull voltage is reduced to 9.67 V from 36 V and the isolation is increased to − 29 dB from − 20 dB at 35 GHz. The optimized dimensions for 35 GHz are shown in Table 1 and its performance analysis graphs are presented from Figs. 3, 4 and 5. The same optimization process is applied for 38 GHz operation frequency such that electromechanical analysis and S-parameter analysis are carried out for understanding switch performance.

Table 1 Specific dimensions of MEMS capacitive switch
Full size table
Fig. 3
figure 3

Return loss comparison of the proposed switch with an existing switch

Full size image
Fig. 4
figure 4

Insertion loss comparison of the proposed switch with an existing switch

Full size image
Fig. 5
figure 5

Isolation loss comparison of the proposed switch with an existing switch

Full size image

Return loss is nothing but reflection loss of the RF input signal at the input terminal, it should be less. In Fig. 3, we compared our simulation results with the existing experimental results to show the performance of our proposed optimization process. We have observed at 10 GHz it is having an approximately equal return loss of − 42.2883 dB. At the operating frequency 35 GHz our simulated result shows − 28 dB whereas experimental is having − 20.4 dB.

Insertion loss is nothing but the loss of RF input signal at the input terminal while the signal is inserting into the signal line, it should be in between 0 and − 1 dB. From Fig. 4, we can observe our optimized design model simulation exhibiting better results when compared with the existing experimental results. At operating frequency 35 GHz it is having > 1 dB, in some cases, both are having the same insertion loss. So we can suggest our proposed process is exhibiting better results.

Isolation loss is nothing but blocking of RF input signal between the input and output ports, it should be high. We compared our simulation results with the existing experimental results which are graphitized in Fig. 5 it shows the performance of our proposed optimization process. For simulation result, it is having isolation of − 29 dB at 35 GHz when compared with experimental results it exhibits − 25 d at 35 GHz (Table 2).

Table 2 Comparison of the proposed switch with Existing switch
Full size table

2.3 Optimization for 38 GHz

The proposed RF MEMS switch is fixed fixed type switch such that both ends of the beam are fixed and the electrostatic actuation is applied at the center of the beam. The capacitance is developed between the beam and a signal line. Sio2 layer is deposited over the substrate. The coplanar waveguide is constructed having two ground planes and a signal line. The dielectric layer is taken over the signal line to develop capacitance between the signal line and Beam Bridge of 0.2 µm thick dielectric is deposited over the signal line. The perforations are created inside the beam to reduce stiction problem, easy movement of the bridge and to reduce the pull-in voltage. Two actuation electrodes are constructed under the beam, over the Sio2 layer provides an electrostatic force to deflect the beam downwards.

Initially, the switch is in ON state which transmits the signal without any deflection in the beam. When a potential difference is applied to the beam, it gets deflected toward down so that it breaks down on dielectric layer enabling OFF condition. The voltage at which the beam breaks down when it undergoes 2/3rd displacement is called pull-in voltage. The switch is modeled by using electro-mechanics physics in COMSOL Multiphysics tool such that the behavior of the switch to the applied voltage and capacitance is generated between the electrodes is studied by observing the displacement of the bridge in the vertical z-direction; it shows the switch is in ON state. The main motto of RF MEMS switches is required less pull in voltage and better performance. Solid mechanics physics is applied to simulate the switch and understand the buckling effect. The robustness of the switch and the tensile strength of the material used for the beam can be analyzed. The performance of the switch is verified by using HFSS tool (Fig. 6; Table 3).

Fig. 6
figure 6

Schematic view of the proposed switch. a Side view, b top view

Full size image
Table 3 Specific dimensions of MEMS Capacitive switch
Full size table

2.4 Material Selection

In designing a device, material selection plays a key role in the functioning and performance of the device and the cost for fabrication of the device. RF MEMS switch performance turned depends on the material selection. From the past decades, different strategies have proposed based on their properties. Rao proposed a selection methodology for materials in graph and matrix form, but it is more complex in computational. R. Roth suggested a multi-attribute utility analysis (MAUA) which is more time-consuming. Ashby’s furnished less computational material selection strategy. For RF MEMS switch Ashby approach is extensively in practice for material selection of beam, dielectric, and substrate [10]. Noted the all possible materials which are suitable for the component. Extract all properties limits and material indices and then screened the materials based on its ranking. Plot the graph based on material indices between the properties unbiased graph is developed, from that graph the left most material is suitable for the switch compared to other considered materials.

Here we considered all possible materials suitable for substrate and scattered the graph against Poisson’s ratio versus young’s modulus. From Fig. 7, the left and topmost material Si (Silicon) shows high Young’s modulus and low Poisson’s ratio is taken as the substrate material. Material with high Young’s modulus possess high stiffness layer and lightly placed. Impedance matching depends on the dielectric constant of the material of substrate which improves the performance of the switch. Silicon is chosen as the material for the substrate because of its high resistivity, less cost and having phase velocity of silicon is 3.44. The transmitting of signal loss is low due to its high dielectric constant εr = 11.9. Secondly, the most important is co-planar waveguide dimensions which plays a prominent role in RF signal transmission.

Fig. 7
figure 7

Material selection for the substrate of switch

Full size image

All possible materials which are suitable for the dielectric layer are taken and plotted the graph over variation between the dielectric constant and electrical resistivity. The material with a high dielectric constant with moderate resistivity is taken as a dielectric layer. From Fig. 8 HfO2 (Hafnium oxide) is chosen as the dielectric layer, due to high dielectric constant. The pull-down capacitance will be increased with a dielectric constant of the material which directly depends on the capacitance ratio of the switch, with increasing the capacitance ratio and pull-down capacitance and the charging and stability of the switch will be maintained.

Fig. 8
figure 8

Material selection for the dielectric of switch

Full size image

All possible materials suitable for beam material are considered and plot the graph with their indices variation between Young’s modulus and Poisson’s ratio. The material with low Poisson’s ratio and moderate young’s modulus is taken as beam material. From Fig. 9 among the materials plotted above Al is having Young’s modulus not too high but not too low, so that the stiffness in the beam is flexible but not too hard. Then in the movement or bending of the beam is flexible. So, it is considered as best beam material. The Al (Aluminum) is less costly compared to Au (Gold).so here we took Al as the beam material (Table 4).

Fig. 9
figure 9

Material selection for a beam of switch

Full size image
Table 4 Material selection for proposed RF MEMS switch
Full size table

3 Results and Analysis for 38 GHz

3.1 Electromechanical Analysis

3.1.1 Pull in Voltage

The pull-in voltage or actuation voltage is defined as the voltage required by the beam to pull down towards the electrode and shows minimum 2/3rd displacement in the gap between the beam and down electrode. The pull-in voltage of the beam can be calculated by using [11]

$$V_{P} = \sqrt {\frac{{8Kg^{3} }}{{27\varepsilon_{{^\circ }} A}}} \,{\text{Volts}}$$
(7)

Here k is the spring constant [20]

$$K = \frac{{32Ewt^{3} }}{{L^{3} }}$$
(8)

where g is a gap between the electrode and beam, ε0 permittivity in free space; an actuation area i.e. is the width of the signal line and w is the width of the beam. A = W × w. The pull-in voltage is measured as 11.97 V and the spring constant is 2.406 N/m (Fig. 10).

Fig. 10
figure 10

Proposed beam structure

Full size image

From Fig. 11 we can observe that there is a constant displacement of the beam after the pull in voltage of 11.97 V through Comsol simulations. Here in this, all simulation date which is performed through the finite element method is plotted the graph over voltage versus displacement. From Fig. 12 the maximum displacement is occurred at the center of the beam. The pull voltage is directly proportional to power consumption, with less pull in voltage there will be less power consumption and fast switching time.

Fig. 11
figure 11

Voltage versus displacement

Full size image
Fig. 12
figure 12

Displacement of a beam for voltage

Full size image

3.1.2 Switching Time Analysis

The switching time is defined as the time taken to change its position from one state to another state, that time taken by the device is called switching time or switching speed. The switching speed of the switch would be low then it can be work fast. The switching time of the RF MEMS shunt switch can be given as [12]

$$T_{s} = \frac{{3.67V_{p} }}{{V_{s} \omega_{0} }}\,{\text{Seconds}}$$
(9)

where VP—pull-in voltage, Vs—supply voltage, Vs = 1.4VP. \(\omega_{0}\) represent resonant frequency. Then switching time required by the switch is ts = 0.19 ms.

The resonant frequency can be calculated from the effective mass and spring constant

$$\begin{aligned} \omega_{0} = \frac{1}{2\varPi }\sqrt {\frac{K}{m}} \,{\text{Hertz}} \hfill \\ \omega_{0} = 13.70 \,{\text{kHz}} .\hfill \\ \end{aligned}$$
(10)

From Fig. 13 it illustrates the speed of switch, how the beam is actuating for applied source voltage and the graph plotted against the switching time and source voltage. For pull in voltage 11.97 V it is having switching time of 0.19 ms.

Fig. 13
figure 13

Source voltage versus switching time

Full size image

3.1.3 ON or UP State Capacitance (CON)

In capacitive shunt switch, it is initially in ON state when the voltage is applied to the beam the membrane gets displaced downwards and capacitance is developed between the electrode and beam. It is ON capacitance or Upstate capacitance. The Upstate Capacitance Con is calculated as [13]

$$\begin{aligned} C_{ON} & = \frac{{\varepsilon_{o} \varepsilon_{r} xy}}{{g + \frac{{t_{d} }}{{\varepsilon_{r} }}}}\,{\text{F}} \\ C_{ON} & = 31\,{\text{fF}} \\ \end{aligned}$$
(11)

x—width of the beam, y—length of the beam, g—gap between dielectric and beam, td—the thickness of the dielectric, \(\varepsilon_{r}\)—relative permittivity of beam material (Table 5).

Table 5 ON State Capacitance
Full size table

Here graph is plotted between the simulated and theoretical upstate capacitance of the switch. The upstate capacitance mainly depends on the gap between the electrodes where as gap is inversely proportional to the upstate capacitance, with increase in gap the capacitance will be decreased. The upstate capacitance is also depends on the dielectric thickness. Here we observed from the Fig. 14 that for various gaps its is having different capacitances in On condition. We considered 3 µm gap because in point of fabrication and also to reduce the stiction problem so the upstate capacitance obtained is 31 fF through finite element method. We observed approximately same for theoretical and simulated values in COMSOL tool.

Fig. 14
figure 14

Simulated and theoretical at UP state

Full size image

3.1.4 Down or OFF State Capacitance (COFF)

The capacitance developed by the switch when the gap vanishes due to a displacement of the beam during actuation is called down state capacitance. The obtained capacitance is termed as downstate or off state capacitance which offers a high impedance to transmit the signal to the output terminal.

Coff is the downstate capacitance it can be calculated as

$$C_{OFF} = \frac{{\varepsilon_{o} \varepsilon_{r} xy}}{{t_{d} }}\,{\text{Farad}}$$
(12)

\(C_{OFF}\) = 0.152 pF.

In this, from Fig. 15 there is little deviation between the simulated and theoretical. The down state capacitance mainly depends on the dielectric thickness of the switch, which is inversely proportional to the downstate capacitance. Here graph is plotted between the simulated and theoretical downstate capacitance of the switch. Here we considered 0.2 µm as the dielectric thickness and downstate capacitance obtained is 0.152 pF through finite element method (Table 6).

Fig. 15
figure 15

Simulated and theoretical at DOWN state

Full size image
Table 6 OFF state capacitance
Full size table

By observing the above graph Fig. 16, if the voltage increases the capacitance developed between the bottom electrode and beam membrane also increases gradually, there is a uniform increase in capacitance up to some supply voltage but when it reaches the pull-in voltage, the beam gets 2/3rd deflection over the gap between them. Then capacitance developed is known as upstate capacitance.

Fig. 16
figure 16

Capacitance versus voltage

Full size image

3.1.5 Capacitance Ratio (\(C_{ratio}\))

It is defined as the ratio of the Downstate capacitance (OFF state) and Upstate capacitance (ON state) of the proposed switch.

$$C_{ratio} = \frac{{C_{OFF} }}{{C_{ON} }} = 104.09$$
(13)

3.1.6 Stress Analysis

Finite element modeling tools are used to observe and analyze the performance of the switch in respective aspects. In COMSOL Multiphysics, solid mechanics is used to observing the stress distribution over the beam membrane of the switch for an applied force. Here aluminum is taken as the beam material. The aluminum can exhibit 70 MPa of stress, for maximum force 12.97E −4 N. The switched beam can be affording maximum force based on its tensile strength of a material. The electrostatic force is induced on the beam when the voltage is applied. The electrostatic force is given by [21]

$$F = \frac{QE}{2}$$
(14)

Hence charge Q = CV

$$F = \frac{CVE}{2}$$
(15)

An electric field \(E = \frac{V}{g}\), g is the gap between the beam bridge and down electrode (Fig. 17).

Fig. 17
figure 17

Voltage versus magnetic force

Full size image
$$F = \frac{{CV^{2} }}{2g}$$
(16)

Considering the Parallel plate capacitance (Fig. 18), the capacitance is

Fig. 18
figure 18

Stress distribution

Full size image
$$C = \frac{{\varepsilon_{^\circ } \varepsilon_{r} A}}{g} = \frac{{\varepsilon_{^\circ } \varepsilon_{r} Ww}}{g}$$
(17)
$$F = \frac{{\varepsilon_{o} \varepsilon_{r} V^{2} A}}{{2g^{2} }}\,{\text{Newton}}$$
(18)

3.1.7 Eigen Frequency Analysis

Eigen frequency analysis is observed by using FEM tools. In this analysis, the vibration of the membrane in a particular frequency and motion of the membrane is in a vertical direction (Fig. 19) [18].

Fig. 19
figure 19

Vertical vibration for Eigen frequency

Full size image

3.1.8 Power Handling

Power handling is an important factor in MEMS technology, the MEMS devices consume less power compared to pin diodes and semi conductor’s switches. The RF MEMS shunt switch required very less power to work; mainly it depends on actuation voltage of the switch. The power consumption of shunt switch can be given by [15].

$$P = \frac{{V_{P}^{2} }}{{Z_{o} }} = 2.38\,\upmu{\text{w}}$$
(19)

3.2 RF Performance Analysis of Shunt Switch

RF performance of the switch can be analyzed by using HFSS tool and operating frequency can be analyzed [14]. Return loss is the loss of RF signal during transmission reflected back at the input terminal due to impedance mismatching of the switch. The reflected back signal is called a reflection coefficient. The return loss or reflection coefficient can be observed when the switch is in ON state. The reflection coefficient should be < − 10 dB. It can be denoted as S11. The return loss at 38 GHz frequency ka band is − 13.086 dB (Fig. 20).

Fig. 20
figure 20

Return loss of proposed switch in HFSS

Full size image
$$\begin{aligned} & S_{11} = - 20\log \left[ {\frac{VSWR - 1}{VSWR + 1}} \right] \\ & VSWR = \frac{1 + \gamma }{1 - \gamma } \\ & \gamma = \frac{{Z_{0} - Z_{L} }}{{Z_{0} + Z_{L} }} \\ \end{aligned}$$
(20)

Insertion loss has occurred during RF signal is transmitting. It is due to resistive loss in switch or contact of a beam with metal. Insertion loss either the switch is in ON state or OFF state (Fig. 21).

Fig. 21
figure 21

The insertion loss of proposed switch in HFSS

Full size image
$$S_{12} = 20\log \left[ {\frac{{Z_{A} + Z_{B} + Z_{C} }}{{Z_{A} + Z_{B} }}} \right]$$
(21)

Isolation should be high for the high-frequency application it is nothing but blocking of the signal between the input and output signal terminals of the switch. It can occur when capacitive coupling happens between the beam and electrode. It should be less than \(\left| { - 10\,{\text{dB}}} \right|\). It can be denoted as S21. The isolation of proposed at 38 GHz is 31 dB (Fig. 22; Table 7).

Fig. 22
figure 22

Isolation of proposed switch in HFSS

Full size image
Table 7 Proposed RF MEMS Switch Optimization Analysis
Full size table
$$S_{21} = 20\log \left| {\frac{{2R_{b} }}{{2R_{b} + Z_{0} }}} \right|$$
(22)

4 Conclusion

In this paper, A novel optimization model is proposed and is applied to the switch at 35 GHz which is fabricated by Chu et al. The dimensions of the fabricated switch are altered using the optimization process and performance results are compared with the existing results. The optimized is exhibits high-performance characteristics such as low pull in voltage 9.36 V and isolation − 29 dB then the fabricated switch. Hence the optimization process is performed to design a fixed–fixed type switch at operating frequency of 38 GHz. The switch produces high isolation of − 31 dB actuating at the pull-in voltage of 11.97 V with a spring constant of 2.38 N/m which is obtained by the optimization process. The electromechanical analysis such as Upstate, Downstate capacitances, C–V characteristics and stress analysis have been carried out. The performance of the switch is analyzed by taking appropriate materials selected by Ashby’s approach. These optimized dimensions are feasible to fabricate. These types of switches can be efficiently used as the switch element between the patches of the antenna for reconfigurability at high frequencies and can be utilized in future 5G communication applications.