Improved Cross Polarization and Broad Impedance Bandwidth from Simple Single Element Shorted Rectangular Microstrip Patch: Theory and Experiment

1 Introduction

Rectangular microstrip patch antenna (RMPA) is the most useful antenna structure for its wide variety of applications, due to its simple design, ease of implementation, low profile, thin and conformal properties. Apart from these desirable properties, a simple RMPA suffers from some severe disadvantages like low gain, narrow impedance bandwidth and poor polarisation purity particularly in its H plane. A RMPA operated at its fundamental mode, radiates linearly polarised field called co-polarised (CO) radiation along with a few degrees of cross-polarised (XP) radiation [1], [2], which increases with frequency and becomes prominent for electrically thick substrate. This XP radiation becomes more significant in H plane than in E plane as obtained in open literature [3]. Therefore, RMPA suffers from poor CO-XP isolation (polarisation purity) mainly in its H plane. Moreover, it produces only around 500 MHz impedance bandwidth when operated in X band frequency. Thus, the concurrent improvement in XP radiation along with the impedance bandwidth for simple RMPA becomes an important issue to the researchers.

Some investigations were reported in Refs. [4]–[6], where the modifications of conventional patch structure have been employed for reduction of XP radiation. The stacked patch configuration [4] produces XP radiation below –16.5 dB over all angles in both principal planes. Similar type of structure with “mirrored pair” feeding technique has been reported in Ref. [5] for L-band array. Two-layer shorted microstrip antenna has been utilized in Ref. [6] and in all those cases, around 15–20 dB of CO-XP isolation are found.

Several groups of researcher have exploited different feeding mechanism to achieve either broad banding or dual polarization along with reduced XP radiation as reported in Refs. [9]–[13]. Meandering strip feeding structure along with curved ground plane [9] produces approximately 650 MHz impedance bandwidth in L band along with –20 dB CO-XP isolation in H plane. Folded plate pair as the differential feeding scheme [10], resonant annular-ring slot and a T-shaped microstrip feed line along with a pair of meandering strips [11] have been employed to improve impedance bandwidth and XP performance of the patch simultaneously. Another investigation [12] exploited aperture coupling to enhance the impedance bandwidth and XP radiation performance of a microstrip patch. Approximately 23 dB CO-XP isolation along with 430 MHz bandwidth at S band has been reported. However, all these structures are quite bulky and complex to design.

The application of defected ground structure (DGS) to improve XP radiation is one of the recent focus of investigations and, 10–12 dB of XP suppression compared to conventional RMPA along with low impedance bandwidth can be obtained by using dot or arc shaped DGS [7], [8]. Around 13 dB of XP suppression compared to conventional RMPA is apparent from [14], where, folded DGS have been employed with a patch in X band. It shows relatively higher impedance bandwidth of 1 GHz in X band [14]. In all these cases, the defects have been placed strategically at a close proximity of patch boundary to perturb the higher order orthogonal resonance. Nevertheless, all the DGS structures inevitably produce back radiation (higher than –20 dB) in terms of XP fields and those are clearly evident from all the structures [7]–[14]. Moreover, in all the cases, defects have not been employed beneath the patch for not to perturb the primary dominant mode field pattern. Instead, it is extended outside the patch peripheries, which produces a severe limitation to accommodate RMPA in array configurations.

Another very recent investigation [15] has dealt with XP radiation of RMPA with non-proximal symmetric DGS and around 9 dB of XP suppression in H plane compared to conventional patch is revealed. However, this structure suffers from the problem of very narrow impedance bandwidth and is very difficult to feed using commercial SMA connectors.

Some investigations [16]–[18] during the last decade have shown some completely new configurations of microstrip patch, where the patch element was shorted with the ground plane. But all those investigations have been done with a view to achieve broad impedance bandwidth and did not deal with XP radiation from a patch. An aperture coupled RMPA loaded with grounded strips has been reported in Ref. [19]. Impedance bandwidth of around 42 MHz in C band with CO-XP isolation of 20 dB in the H plane is evident from Ref. [19]. However, the structure is very complex and it enhances E plane XP of around –17 dB.

In the present investigation, a simple single element RMPA with three pairs of shorting plates placed along non radiating edges is proposed (Figure 1) to address both the issues of input and radiation performance concurrently. The shorting plates are placed in such a way that, the dominant mode CO radiation will be similar to that of conventional RMPA while it affects the XP field which is mainly due to higher order orthogonal resonances. The proposed structure is very simple and easy to manufacture. A simultaneous improvement in impedance bandwidth and very high CO-XP isolation are noted from single element RMPA. Moreover, unlike the earlier structures, the present design does not suffer from increased back radiation in terms of XP fields. It may be noted that, the present structure maintains the high polarisation purity (more than 25 dB CO-XP isolation) over the entire angular range (±180°) around the broadside direction along with 1.2 GHz impedance bandwidth in X band. The present report provides a clear visualization based understanding of the observed phenomena, which will be helpful for germination of other new structures.

Figure 1:

Schematic representation of proposed RMPA with shorting plates. (a) Top view, (b) Side view (L and W are the length and width of the patch, l₁ and d₁ are length and width of the shorting plates and ε_r is the dielectric constant of the substrate, p_c is the feed position).

2 Theory and proposed structure

2.1 Theory

RMPA with three pair of shorting plates placed at the non radiating edges modify the cavity beneath the patch. Therefore the fields within the cavity become affected and hence modify the radiation property of the RMPA. A recent investigation by the present authors [20], [21] has clearly depicted that the XP radiations are typically from the non radiating edges of the patch. In fact, the oscillations of electric field beneath the patch in a direction, orthogonal to E plane produce higher order orthogonal resonance and they are called higher order orthogonal modes. The XP radiations are typically due to those higher order orthogonal modes and the fields of those modes, located near non radiating edges results in XP radiation from patch antenna. Hence the shorting plates have been incorporated at those edges to minimize the fields at non radiating edges. This in turn reduces XP radiation from the proposed RMPA.

The electric field vectors for shorted patch in TM_np mode may be obtained from Ref. [20] as

(1a)Ex∞cosnπLy′sinpπWz′

(1b)Hy∞cosnπLy′cospπWz′

(1c)Hz∞cosnπLy′sinpπWz′

where, E_x, H_y and H_z are the electric and magnetic field components of the dominant mode. The number of half wave variations along length and width of the patch are denoted by n and p respectively.

The electric surface current over patch surface can be obtained from

(2)Js=nˆ×H

The co-sinusoidal variation in eq. (1) denotes the variation of E_x along length (L) of the patch, while the sinusoidal variation denotes the variation of E_x along orthogonal direction. Therefore, any higher order of orthogonal resonance (i.e., for any non zero value of p) leads to minimum electric field intensity when z approaches W. In fact, the intensity of the electric fields near non radiating edges due to all higher order orthogonal modes (primarily responsible for XP radiation) are forced to be minimum in order to satisfy the vanishing electric field boundary conditions. This eventually mitigates the possibility of XP radiation from non radiating edges due to higher order orthogonal modes. The electric field vector for first higher order orthogonal resonance is illustrated in Figure 2 for conventional and proposed RMPA. It reveals that the electric field intensity near non radiating edge for first higher order orthogonal mode in case of proposed RMPA is significantly low compared to that for conventional RMPA.

Figure 2:

Electric field intensity near non radiating edges for first higher order orthogonal resonance (resonating along orthogonal x-z plane): (a) conventional RMPA, (b) proposed RMPA.

The magnitude of electric field distribution for the dominant resonant mode is depicted in Figure 3(a). It clearly confirms the conjecture of minimization of electric field near non radiating edges for dominant mode in case of proposed RMPA. Hence, the orthogonal component of dominant mode is also annihilated and as a result XP performance improves. The electric field vector within the substrate for the dominant resonant mode is presented in Figure 3(b). It is apparent from the figure that, the electric field component beneath the patch in the dominant mode is exactly similar like conventional RMPA. However, unlike conventional RMPA, here, the electric fields near non radiating edges have lower intensity compared to the electric fields located at central region. This is basically due to the sinusoidal variation of electric field along the orthogonal direction. Except this, the field distribution beneath the patch at central region of radiating edge remains unchanged and therefore, the dominant mode radiation characteristics remains unaltered.

Figure 3:

(a) Magnitude of electric field over patch surface in dominant mode for proposed RMPA. (b) Electric field vector beneath the patch in dominant mode for proposed RMPA.

For further corroboration of the phenomena, the electric surface current (J⃗s) over the patch surface for both conventional and proposed RMPA is depicted in Figure 4. The XP radiation from RMPA is typically from the non radiating edges and those are due to asymmetric field distribution along the length of the patch. This asymmetry in the field distribution is basically due to the placement of feeding probe and is unavoidable for probe fed patch. This asymmetry in the field causes asymmetry in the electric surface current (J⃗s) along y direction i.e., the patch length. It is noted that, for conventional RMPA, the y component of J⃗s does not become maximum at the centre. Instead, it is slightly off centred when it is seen along x-y plane. This y component of J⃗s at non radiating edge may be attributed for high XP radiation from RMPA. In case of proposed RMPA, the fields beneath the patch are so modulated that, the electric surface current at non radiating edge does not follow the conventional profile. In fact, this J⃗s becomes insignificant near the non radiating edges, while it follows similar profile like conventional RMPA at the central region of the patch. Therefore, XP radiations are mitigated keeping co-polarized radiation unaltered.

Figure 4:

Electric surface current distribution over patch at dominant mode: (a) conventional RMPA, (b) proposed RMPA.

Here in, we may concentrate on input impedance bandwidth of the proposed RMPA. Three pairs of thin shorting plates with dimension (l₁×d₁) not only modify the radiation property but also regulate the input characteristics of the RMPA. In fact, these pairs of thin shorting plates act as small dipoles of non circular cross section which are placed along non zero electric field lines. Therefore, the present RMPA is loaded with short dipoles and this short dipole loaded RMPA eventually improves the input parameter (impedance bandwidth) of the proposed RMPA. The length (l₁) of thin short plate is essentially same as substrate thickness h and it is in the order of λ/10 for the present antenna. Therefore,

(3)l1=λgr10

(4)λgr=λrεr

(5)λr=cfr

where f_r is dominant mode resonant frequency, λ_gr is the resonant wavelength within dielectric and ε_r is the dielectric constant of substrate material.

The width of the short (d₁) is considered to be very thin and is chosen in such a way that it will not hamper the field structure beneath the patch. Each short dipole (l₁×d₁) produces the reactive impedance [22] as

(6)Xs=30{2Si(kl1)+cos(kl1)[2Si(kl1)−Si(2kl1)]−sin(kl1)[2Ci(kl1)−Ci(2kl1)−Ci(2kd′12l1)]}

which comes parallel to patch input impedance Z_p. Here d1′ is the equivalent circular radius of dipole of width d₁. For the dipole of non circular cross section, d1′may be obtained from Ref. [23];

(7)d1′=0.25d1

Now, using eqs (3)–(5);

(8)kl1=2πλgλgr10=0.628λgrλg=0.628cfrεrcfεr=0.628ffr

And

(9)2kd′12l1=7.85ffrd12εrc2

Now, using eqs (6)–(9);

(10)Xs=30{2Si(0.628ffr)+cos(0.628ffr)[2Si(0.628ffr)−Si(2×0.628ffr)]−sin(0.628ffr)[2Ci(0.628ffr)−Ci(2×0.628ffr)−Ci(7.85ffrd12εrc2)]}

Now, the electrical equivalent of a conventional probe fed RMPA is a simple parallel tuned circuit. Hence, the input impedance of conventional probe fed RMPA can be written as

(11)Zp=1(1Rr)+jωC+(1jωL)

where, R_r is the resonant resistance of the patch at particular feed position. C and L are capacitance and inductance respectively and can be obtained from Ref. [3]. When this conventional probe fed RMPA is loaded with short dipoles, the dipole reactance (X_s) will come in parallel to patch and the resultant input impedance of dipole loaded patch or present antenna becomes,

(12)Zdp=1(1Rr)+jωC+(1jωL)+jXs

Putting the values of C and L from Ref. [3],

(13)Zdp=1(1Rr)+j[fQTfrRr−frQTfRr+Xs]

where, Q_T is the total quality factor and may be obtained from Ref. [24]

Hence,

(14)Re(Zdp)=1Rr(1Rr)2+[fQTfrRr−frQTfRr+Xs]2

(15)Im(Zdp)=Xf−[fQTfrRr−frQTfRr+Xs](1Rr)2+[fQTfrRr−frQTfRr+Xs]2

where, X_f is the feed reactance and can be obtained from Ref. [23].

The sharp variation of input reactance limits the operating bandwidth of RMPA. When the shorting plates are placed near non radiating edges, the structure becomes thick dipole loaded (dipole length to diameter ratio ~4.2 near resonant frequency) RMPA. The reactance of thick dipole slowly varies with the frequency and as it is in parallel to patch reactance, the resultant reactance of the proposed patch varies slowly with frequency. The variation of the resultant input reactance of the proposed RMPA is shown in Figure 5, which clearly indicates the fact. Consequently, the impedance bandwidth of the proposed antenna is enhanced compared to conventional RMPA.

Figure 5:

Variations of imaginary part of input impedance (input reactance) as a function of frequency for conventional and proposed RMPA. Total quality factor Q_T=0.22, resonant resistance R_r=50 Ω and can be obtained from Ref. [24] for RMPA with L=8 mm, W= 12 mm, h= 1.575 mm, ε_r=2.33. Feed reactance X_f=15 Ω and can be obtained from Ref. [23].

2.2 Proposed structure

A RMPA is designed using thin copper strip of thickness 0.1 mm, length L=8 mm and W=12 mm to operate around X band. Taconic’s TLY-3-0620 PTFE material (ε_r=2.33) with thickness h=1.575 mm is utilized as substrate. The feed location (2.4 mm from centre of the patch) has been optimized based on good impedance matching. The ground plane dimensions are considered to be 80×80 mm². Three pairs of thin copper strips of thickness 0.1 mm with height of h=l₁=1.575 mm and width d₁=1.5 mm are placed through the groove cut at the substrate at the position of non radiating edges as shown in Figure 1, which in fact gives birth to shorted plates at non radiating edges. The fabricated patch utilized for the measurement is presented in Figure 6.

Figure 6:

Photograph of proposed fabricated patch.

3 Results and discussion

A pair of prototypes with above mentioned dimensions (section II B) have been fabricated and utilized for measurements. Both the patches are excited by a PE4128 SMA connector and Agilent’s E8363B network analyzer has been used to obtain the reflection coefficient profiles for the prototypes. We use Agilent’s E4413A CW power sensor (50 MHz to 26.5 GHz) with Agilent’s probe to pick up the received signal. The simulated [25] and measured results obtained for the conventional and the proposed structure are presented and compared in the following section. Both the structures have been designed to resonate near X band frequency.

The simulation results for CO-XP isolation and the input impedance bandwidth (less than –10 dB bandwidth) for the proposed antenna for different width (d₁) of shorting plate is presented in Table 1. It shows that as we increase the value of d₁, minimum CO-XP isolation as well as the input impedance bandwidth improves. The improvements in both the parameters continues up to d₁=1.5 mm. Further increase in d₁, does not show significant improvement in minimum CO-XP isolation while reduces the input impedance bandwidth of the present antenna. Hence we refrain from increasing the value of d₁, as it may hamper the dominant mode characteristics. No significant variation in peak CO gain is apparent from all the structures.

Reflection coefficient profile of the conventional and proposed RMPA is shown in Figure 7. Simulation and measured results are plotted in same figure for both the structures. It reveals the close mutual agreement between simulation and measurements for individual antenna. It may be noted that, the resonant frequency shifts to the higher side of the spectrum for proposed RMPA compared to conventional one. This is because of the shorted plates placed along non radiating edges in present structure as clearly depicted in Ref. [20]. In fact, due to the shorted plates, one sinusoidal variation of field component along the orthogonal direction is evident from eq. (1) and also obtained from Ref. [20]. Therefore, in case of present structure, the dominant mode frequency is shifted by 30% toward the higher side of the spectrum. It may be noted that, this will not perturb the dominant mode radiation pattern at all. The insightful analysis reported in Ref. [20] by present authors beautifully shows this up shift of dominant mode frequency (f_r) without hampering dominant mode radiation. The input impedance bandwidth (less than –10 dB bandwidth) for the conventional antenna is only 500 MHz and the same for present structure becomes 1.32 GHz near X band frequency as is clear from the Figure 7. In the case of present structure, the electric fields are forced to be minimum because of shorting plates at non radiating edges, and therefore magnetic fields increase near non radiating edges. This inevitably produces high interaction of electric currents with the shorts. Therefore, conduction loss increases and lowers the cavity Q factor, which in turn increases the impedance bandwidth.

Figure 7:

Simulation and measured reflection coefficient profile for conventional and proposed RMPA.

Figure 8 shows the complete radiation patterns for both the prototypes at the centre frequency. It shows that the co polarized radiation pattern (CO) is similar for both the structures in the principal E and H planes. It is observed from the Figure 8(a) that, for conventional RMPA, H plane XP level becomes appreciable (higher than –20 dB) from ±25° o to ±120° around the broadside with the peak XP level of –10 dB near ±50°. Therefore, in case of conventional structure, CO-XP isolation is very poor and it is in the order of 10 dB only in its H plane.

Figure 8:

Comparison of simulated and measured radiation patterns for conventional (f=10.05 GHz) and proposed RMPA (f=12.9 GHz) at fundamental resonant mode. (a) H plane, (b) E plane.

Now, when the shorting plates are placed; a remarkable change in XP radiation is observed in its H plane radiation pattern. The H plane XP pattern is suppressed by 15–25 dB compared to conventional structure as is exposed from the same figure (Figure 8(a)). It results in high CO-XP isolation of more than 25 dB for the present antenna for whole ±180° angular range around broadside. It may be noted that, the back radiation in terms of XP fields are insignificant in case of present structure and it is in the order of –35 dB at ±180°. The E plane radiation pattern is depicted in Figure 8(b). It shows that the E plane XP level in conventional RMPA is significantly low and it is below –35 dB in the total ±180° angular range around broadside. However, approximately 4–5 dB suppression in E plane XP pattern is also evident from proposed structure as is clear from the figure (Figure 8(b)). Nevertheless, this suppression is not significant as the E plane XP pattern remains always insignificant and less than 30 dB compared to CO radiation. Hence we refrain from presenting the E plane pattern of the structure when the investigations are carried out for other frequencies in the operating band of proposed RMPA. The H plane radiation pattern of proposed RMPA in other two frequencies in the operating band is presented in Figure 9. A consistent suppression of H plane XP level is revealed in those frequencies while the CO pattern is kept unaltered. Both the Figure 9(a) and (b) shows that the present antenna maintains its excellent CO-XP isolation of more than 26 dB in other two frequencies also. The back radiation in terms of XP fields are more than 30 dB down than peak CO gain for the present antenna for all those frequencies which may be helpful in frequency reuse technique in array applications. The measured CO gain of the proposed antenna varies from 6.1 dBi to 6.27 dBi in its entire operating frequency band. Therefore, the antenna may be utilized for the applications where polarization purity is the major issue over entire angular range in setting up a wireless communication link for broad impedance bandwidth.

Figure 9:

Comparison of simulated and measured radiation patterns for proposed RMPA at different frequencies in its operating band: (a) f=12.4 GHz (b) f=13.4 GHz.

4 Conclusion

Simple single element rectangular microstrip antenna with shorted plate is proposed for broader impedance bandwidth and significantly improved CP to XP isolation. The proposed RMPA finds potential applications in the fields where high XP is a major limitation over wide frequency range.