New circularly polarized microstrip patch antenna array design for energy harvesting applications

,


Introduction
The energy harvesting has taken much interest in recent decades [1], especially wireless power transmission (WPT) thanks to its autonomy without using cables or interconnect wires, its independence of the main socket [1], its low cost in the manufacturing phase [2], and its generation of energy without harming the environment.The WPT field motivates researchers to develop new strategies to generate electricity in a clean way.The most popular WPT system is the rectenna system also called rectifying antenna.The rectenna blocks [3] are the receive antenna, the input matching circuit, the radio frequency-direct current (RF-DC) rectifier, the output matching circuit, and the load.
This paper focuses on a receive antenna design which is the first and essential block of the rectenna, and it permits to increase in the efficiency of the RF-DC conversion system by increasing the received power at the input of the rectenna.There are a lot of studies about the antenna such as the choice of the operating frequency at 5.8 GHz [4], 35 GHz [5], etc., the choice of the substrate, the choice of the polarization: linear polarization [6] or circular polarization [7]- [8], and the choice of the radiator element shape as shown in Figure 1.The main contribution of this paper is to develop a new design of a circularly polarized (CP) microstrip patch antenna (MPA) array operating at 2.45 GHz in the industrial, scientific, and medical (ISM) band.The choice of the radiator element shape is a square shape.The radiator element uses both techniques the slot technique at the center of the radiator element and the slit technique at the four corners of the radiator element.The benefits of these two techniques [9] are to achieve a CP, reduce the overall surface of the MPA array, and achieve an adequate adaptation to the excitation port.This proposed MPA array uses also the defected ground structure (DGS) method.The DGS method aims to reduce the overall size of the MPA array and to enhance the bandwidth [10].The effectiveness of this developed MPA array design is proved by simulation results obtained by Computer Simulation Technology Microwave Studio (CST MWS) software, validated by another solver High-Frequency Structure Simulator (HFSS), and compared with other works in the literature.
This work was divided into four parts.The second section describes the specifications of the MPA array design.The third section is concerned with the simulated results followed by their comprehensive discussions using CST MWS and HFSS.Finally, a conclusion is given in the last section to summarize this paper.

Specifications of the design
The first step concerning the design of the MPA is to specify the essential parameters such as the operating frequency value, and the specifications of the used material for the substrate (i.e., its relative permittivity, its thickness, its tangential loss, etc.).
The second step is concerned with the selection of the patch shape, its conductor type, and its dimensions by using the equations [8]- [11] which aim to determine the dimensions of the patch.These equations are given below from (1) to (4).

𝑊 =
(1) Where: W is the patch's width as in (1),  is the dielectric effective permittivity as in (2), L is the patch's length as in (4) which is extended by a distance ∆ as in (3) due to fringe phenomena,  is the wavelength and  is the resonance frequency.
The selection of the substrate is FR4-epoxy.The specifications of the design proposed in this work are listed in Table 1.The last step is focused on the simulation of the proposed MPA structure, the optimization of the dimensions, and the modification of the MPA structure by using some known techniques in the design of MPA.The optimization of all dimensions of the ground, the substrate, and the patch is repeated until the simulation results meet the intended objectives [8].
This present work investigates a new proposed design of a CP MPA that operates at 2.45 GHz in the ISM band.The selection of the patch shape is a square based on two techniques and the DGS method.The patch is inserted into a rectangular slot in the center in order to obtain CP because CP requires rectenna systems to conserve a constant direct current (DC) voltage or power across the load when either the transmitter or the receiver changes direction in space and to minimize the size of the square patch antenna.Additionally, the patch is inserted by V-slits at its four corners to propagate the CP, improve its quality, and increase the bandwidth around the operating frequency at 2.45 GHz.This proposed MPA design is feeding by a microstrip line method using a single excitation port.The DGS method aims to increase the bandwidth and reduce the overall size of the MPA.To enhance the MPA's performance, the 1 × 4 (i.e., one row and four columns) MPA array design as illustrated in Figure 2 is more developed based on the basic element with 4 identical basic elements that have the same dimensions.
The idea of developing the MPA array design is to increase the MPA's performance such as the gain, the directivity, and the efficiency.Figure 3a shows the top view of the MPA array.Figure 3b shows the bottom view of the MPA array.The benefits of this proposed design are:  The ISM band is chosen thanks to its utility in various applications in the three fields of industrial, scientific, and medical;  The choice of circular polarization (CP) is more efficient thanks to its greater mobility and its capability to preserve a constant DC power for the rectenna system;  The choice of the square shape feeding by a microstrip line method using a single excitation port based on the slit and slot techniques is more efficient in order to achieve a CP,  The DGS method is used to reduce the overall size of the MPA array structure and to facilitate the integration with the other devices.

Simulated results and discussions
The simulation results of the important parameters can help us to explain the MPA's performance, discuss the results of the proposed design in this work, and compare these results with other works in the literature to give a brief conclusion about the performances.All the simulated results are provided by the simulator of the full−wave which is named Computer Simulation Technology Microwave Studio (CST MWS).Another solver named High-Frequency Structure Simulator (HFSS) is used to confirm the results obtained by CST MWS.
Figure 4 shows the evolution of the reflection coefficient also called the return loss (S11) as a function of the frequency of the CP MPA array.This work concerns the study of the influence of various values of length of the ground noted by  and fixing the width of the ground noted by  as illustrated in Figure 4a.When the value of  increases, the return loss tends to lowfrequencies (LF) as shown in Figure 4a.When the value of  reduces, the return loss tends to high-frequencies (HF) as shown in Figure 4a.This work concerns the study of the influence of various values of width of the ground noted by  and fixing the length of the ground noted by  as illustrated in Figure 4b.When the value of  increases, the return loss tends to lowfrequencies (LF) as shown in Figure 4b.When the value of  reduces, the return loss tends to high-frequencies (HF) as shown in Figure 4b.From the analysis of Figure 4a and Figure 4b, the optimized values of the length of the ground and the width of the ground are 67 [mm] and 22 [mm] respectively.Hence, the quantity of the losses is small from this antenna patch, and we will have a better energy transfer through the antenna thanks to a good adaptation.The bandwidth of the proposed CP MPA array design is around 120.44 [MHz] as illustrated in Figure 5a.
Figure 5b shows the variation of the voltage standing wave ratio (VSWR) as a function of the frequency of the proposed CP MPA array.At the 2.45 GHz resonance frequency, the VSWR's value of this MPA is about 1.019 and 1.091 by CST MWS and HFSS respectively.These values are close to 1.Then, we deduce that this proposed design is well-suited and offers maximum power with a small quantity of losses.
Figure 5c shows the variation of the input impedance including both the real part and the imaginary part.As shown in Figure 5c, the real part equals to 50Ω and the imaginary part equals to 0 Ω.So, this proposed design is well-adapted to the excitation port of value 50Ω.
Figure 5d illustrates the variation of axial ratio.At 2.45 GHz, the value of the axial ratio is strictly inferior to 3 [dB] is around 0.91 [dB].This value proves that the proposed 1×4 MPA array in this paper is operating with a CP at 2.45 GHz in the ISM band.
Figure 5e shows the variations of the various powers.At 2.45 [GHz], the accepted power is about 0.5 [W], the radiated power is about 0.47 [W], and the reflected power is about 0.0437 [mW].These values of the powers proves that the proposed design can transfer the maximum of the energy with a minimum of losses.
Figure 5f shows the efficiency of the proposed design.The value of the efficiency at 2.45 [GHz] is around 94.03 %.    3, we observe that our proposed network (i.e., 1×4 CP MPA array) is more efficient in terms of bandwidth and directivity compared to the other works cited.However, the 1×4 network of our work is better in terms of directivity especially since this network contains only four patches, it works at 2.45 GHz in the ISM band with a CP, and it is well adapted.Our design will be able to work in all applications that are based on WPT as wireless sensor networks, aerospace applications, automobile applications in electric vehicles for charging batteries, brain-machine interface systems, neural recording systems in medical areas, wireless mobile chargers, etc.

Conclusion
This present work gives a novel developed design of a circularly polarized (CP) microstrip patch antenna (MPA) network operating at 2.45 GHz resonant frequency in the industrial, scientific, and medical (ISM) band for various radio frequency energy harvesting (RFEH) applications.The proposed design is described as a square shape of the radiator element using both techniques slot and slit techniques.This developed CP MPA used the defected ground structure (DGS) method.This proposed CP MPA array is designed, simulated using CST MWS, and validated successfully by HFSS.Simulation results prove the effectiveness of this design in achieving maximum energy with low losses, adequate adaptation to the excitation port, high gain is about 9.14 [dBi], high directivity is about 9.41 [dBi], large bandwidth is about 120.44 [MHz], and smallest surface equals 135.67  .This novel design is considered a better solution for various RFEH especially rectenna systems thanks to its self-powered electronic devices.For future research, this CP MPA array can be extended further for manufacturing in order to test its prototypes.

Figure 1 .
Figure 1.The most used geometric forms of the patch

Figure 2 .
Figure 2. Three-dimensional view of the Proposed MPA array design

Figure 4 .
Figure 4.Return loss vs frequency, (a): Various lengths, and (b): Various widths From Figure 5a, at the resonance frequency of 2.45 GHz,  = −40.58[dB] by CST MWS and  = −27.14[dB] by HFSS.The use of different solvers explains these different values of return loss because each solver has its own numerical method to calculate the value of return loss.Based on Figure 5a, these return loss values are too smaller than −10 [dB] at 2.45 GHz.Hence, the quantity of the losses is small from this antenna patch, and we will have a better energy transfer through the antenna thanks to a good adaptation.The bandwidth of the proposed CP MPA array design is around 120.44 [MHz] as illustrated in Figure5a.Figure5bshows the variation of the voltage standing wave ratio (VSWR) as a function of the frequency of the proposed CP MPA array.At the 2.45 GHz resonance frequency, the VSWR's value of this MPA is about 1.019 and 1.091 by CST MWS and HFSS respectively.These values are close to 1.Then, we deduce that this proposed design is well-suited and offers maximum power with a small quantity of losses.Figure5cshows the variation of the input impedance including both the real part and the imaginary part.As shown in Figure5c, the real part equals to 50Ω and the imaginary part equals to 0 Ω.So, this proposed design is well-adapted to the excitation port of value 50Ω.Figure5dillustrates the variation of axial ratio.At 2.45 GHz, the value of the axial ratio is strictly inferior to 3 [dB] is around 0.91 [dB].This value proves that the proposed 1×4 MPA array in this paper is operating with a CP at 2.45 GHz in the ISM band.Figure5eshows the variations of the various powers.At 2.45 [GHz], the accepted power is about 0.5 [W], the radiated power is about 0.47 [W], and the reflected power is about 0.0437[mW].These values of the powers proves that the proposed design can transfer the maximum of the energy with a minimum of losses.Figure5fshows the efficiency of the proposed design.The value of the efficiency at 2.45 [GHz] is around 94.03 %.Figure5gshows the three-dimensional radiation pattern at 2.45 [GHz].The gain is about 9.14 [dBi] and the directivity is about 9.41 [dBi].
Figure 5g shows the three-dimensional radiation pattern at 2.45 [GHz].The gain is about 9.14 [dBi] and the directivity is about 9.41 [dBi].

Table 1 .
Specifications of the design.

Table 2
show the optimized values of the proposed CP MPA array design.

Table 2 .
Optimized dimensions of the MPA array.

Table 3 .
Comparison of the simulated results.