Published in IET Microwaves, Antennas & Propagation Received on 19th November 2008
Revised on 21st July 2009 doi: 10.1049/iet-map.2008.0400
ISSN 1751-8725
Microstrip-fed broadband circularly polarised
monopole antenna
J.-W. Wu
1
J.-Y. Ke
1
C.F. Jou
1
C.-J. Wang
2
1Department of Communication Engineering, National Chiao Tung University, Hsinchu, Taiwan 2Department of Electrical Engineering, National University of Tainan, Tainan, Taiwan
E-mail: cjwang@mail.nutn.edu.tw
Abstract: A microstrip-fed broadband circularly polarised (CP) monopole antenna was studied. A broad impedance bandwidth and wide axial ratio bandwidth (AR-BW) could be achieved simultaneously. This antenna used a conventional monopole architecture, except for its deforming ground plane and asymmetric-feed approach. The asymmetric-asymmetric-feed was used to provide an orthogonal component distinct from its original linear polarisation. In addition, by embedding a slit and a stub on the ground plane, this antenna could generate CP wave radiation and achieve a broad impedance bandwidth. According to the measurement results, the impedance bandwidth was 6.56 GHz for a 10 dB return loss, which covered a range of 2.32 – 8.88 GHz. The AR-BW was 1.2 GHz for a 3 dB AR, which covered a range of 3.2 – 4.4 GHz.
1
Introduction
Recent applications of circularly polarised (CP) waves have attracted much attention due to their significant superiority in resisting inclement weather as compared to linearly polarised (LP) waves. In particular, they have been employed in modern communication systems that are sensitive to atmospheric variations, such as radar tracking, navigation, satellite systems and mobile communication systems [1]. The hazard caused by misalignment can be ignored to simplify antenna mounting as well as to improve reception efficiency. Exciting a CP wave requires two conditions: first, the amplitudes of two near-degenerate orthogonal E vectors must be equal; second, the phase difference (PD) between the two orthogonal E vectors must be approximately 908. Right-hand circular polarisation (RHCP) or left-hand circular polarisation (LHCP) can be defined by a 908 phase lead or lag. Traditionally, a polariser has been required for exciting a quadrature phase contribution to produce CP. To do so, some approaches have employed couplers, dividers or phase shifters to provide a 908 PD [2]. These mechanisms have been referred to as the so-called dual-feed technique. On the other hand, some researchers established a cavity model to estimate the central frequency of CP and the polarised sense based on the physical dimensions and feeding positions of the antenna [3 – 5]. These configurations
enabled CP capability to be realised using a single-feed method, which simplified the feeding networks.
In addition to patch antennas, many types of antennas can effectively generate CP, such as slot antennas [6], helical antennas [7] and arrays [8]. In recent years, many studies have designed CP monopole antennas. In 1998, Ojiro used a monopole feed and a symmetrical loop to generate a travelling wave current and realise CP [9]. A coplanar waveguide (CPW)-fed monopole antenna with a shorting sleeve strip was used to excite a CP mode by the coupling effect between the monopole antenna and sleeve[10].
This paper proposes a microstrip-fed monopole antenna to achieve a broad impedance bandwidth and wideband CP. This antenna is composed of an asymmetric feed line, a rectangular radiator and a ground plane with an embedded slit and stub. Asymmetric feeding achieves a wide impedance bandwidth and excites an elliptically polarised (EP) wave. By modifying the shape of the ground plane, this antenna simultaneously generates a broad impedance bandwidth and wideband CP. The current work presents parametric studies of the antenna geometry, and the measured results show that this antenna excites a broad impedance bandwidth of 102.5% at a centre frequency of 5.6 GHz and a wideband CP radiation wave of 31.6% with respect to a centre frequency of 3.8 GHz.
2
Operation of circular
polarisation
Feeding structures are typically classified into two categories, central feeding and asymmetric feeding, which can cause different surface current distributions on an antenna.
Fig. 1a shows the surface current distribution for central feeding, which can be divided into vertical and horizontal currents. The distribution of the horizontal current excites two components that are 1808 out of phase. Therefore the radiation in the far field in the horizontal direction is very weak. Thus, it is very difficult for a conventional monopole antenna to excite CP. Asymmetric feeding, on the other hand, generates two orthogonal currents, which include the vertical and horizontal currents, as shown in Fig. 1b. Their amplitudes and PD cannot reach CP conditions; hence, the asymmetric feeding method can only excite an EP wave.
In general, CP is generated by two orthogonal E vectors with equal amplitudes and a 908 PD. It is defined as
E ¼ EHorþe jd
EVer (1)
where E is the instantaneous electric filed vector, EHorand
EVer, respectively, denote the electric field vectors in the
horizontal and vertical planes, and d is the PD. If the amplitudes of EHorand EVer are equal and d ¼ +908,
the polarised wave is RHCP or LHCP[7]. In addition, the value of the axial ratio (AR) can be used to represent the characteristic of the polarisation. The AR is defined by the RHCP or LHCP[11, 12], and it is expressed as
AR ¼ 20 logr þ1 r 1 (2) where r ¼ ERHCP ELHCP (3)
Generally, there are three types of polarised waves: LP, EP and CP. For a perfect CP wave, the AR value is 0 dB; for
a perfect LP wave, the AR value is infinite. EP is considered to lie between LP and CP. Because a perfect CP wave with AR ¼ 0 dB is ideal, CP is typically defined based on an AR value of less than 3 dB.
The asymmetric feeding method generates EVerand EHor,
but it excites EP. To achieve equal amplitudes and a 908 PD, a slit was embedded on the ground plane. Using this method, the amplitudes of EHorand EVerwere almost equal, and the
phase of EHor led EVer with a 908 PD, which excited an
LHCP radiation wave. Fig. 2 shows the simulated surface current. The amplitude difference and PD between EHor
and EVer were varied to generate the CP mode.
Furthermore, adding a stub on the ground plane further increased the impedance bandwidth, while retaining CP performance. Fig. 3 illustrates the configuration of the proposed monopole antenna. The proposed antenna was etched on an FR4 substrate with a relative permittivity 1r¼ 4.4, loss tangent tan d ¼ 0.024 and thickness
H ¼ 1.6 mm. The top view in Fig. 3a shows that an asymmetric microstrip-feed line of width W1and length L1
was connected to a rectangular radiator of width W2 and
length L2. As shown in the bottom view in Fig. 3b, a
L1W ground plane was etched on the bottom side of
this antenna. An S1S2rectangular slit was embedded on
the ground plane, and a D1D2 stub was loaded on the
ground plane to the right of the slit. The overall size (L W H ) of the proposed antenna was approximately 45 40 1.6 mm3. Fig. 4 shows a flowchart for the design process used for the proposed CP monopole antenna.
3
Analysis of antenna design
This section describes the design techniques used for this antenna, which include increasing the impedance bandwidth and generating circular polarisation. The
Figure 1 Simulated surface current distributions at 3 GHz
a Central feeding b Asymmetrical feeding
Figure 2 Simulated surface current distributions of embedding a slit on the ground at 3 GHz
simulations were carried out using the finite element method software ‘Ansoft High Frequency Structure Simulator’ HFSS 10.0[13].
3.1 Asymmetric feed-line
Fig. 5describes the effects of central and asymmetric feeding on the simulated return losses. The asymmetric feeding method excited a mode at 5 GHz to increase the impedance bandwidth from 40.6% for central feeding to 82% due to via a change in the current distribution. The simulated AR results of the central and asymmetric feeding at the broadside direction are shown in Fig. 6. The centrally fed antenna radiated an LP radiation wave with an AR value greater than 40 dB. The asymmetric feeding changed the polarised wave from LP to EP, reducing the
AR value to approximately 7 dB at 5 GHz. Some parasitic elements were added to the ground plane to obtain a CP radiation wave. Details of the design analysis of the parasitic elements are described in the following subsections.
3.2 Embedding a slit on the ground plane
A wide impedance bandwidth and EP were generated by asymmetric feeding. To generate CP with two orthogonal currents of equal amplitude and a 908 PD, a S1S2
rectangular slit was embedded on the ground plane. The performance of the impedance bandwidth and the CP were affected by the dimensions of the slit (S1, S2).
Figure 3 Configurations of the proposed printed monopole antenna
a Top view b Bottom view
Figure 4 Design flow chart for the proposed antenna
Figure 5 Comparison the simulated return losses of central feeding and asymmetric feeding
The simulated return losses and AR of the broadside direction results with different rectangular slit lengths (S1)
are plotted in Fig. 7. With a fixed value for S2, slits with
three different lengths – 14, 15 and 16 mm – were analysed. As seen in Fig. 7a, the return losses from 2 to 4 GHz were affected slightly by the length S1. However,
the impedance matching was strongly dependent on S1
from 4 to 9 GHz. In order to match the high-frequency impedance band from 4 to 9 GHz, the length of S1 was
chosen to be 16 mm. Fig. 7b also shows the effects of the S1 length on the AR in the broadside direction. The
findings clearly show that the CP mode frequency was not controlled by S1, but that the widest AR-BW could be
reached by properly adjusting S1. The length of S1not only
matched the high-frequency impedance band from 4 to 9 GHz but also tuned the AR-BW.
Figs. 8a and b depict the effects of the rectangular slit height (S2) on the simulated return losses and AR for the
Figure 6 Comparison the simulated AR of central feeding and asymmetric feeding
Figure 7 Simulated return losses and AR of the rectangular slit length
a Return losses b AR
Figure 8 Simulated return losses and AR of the rectangular slit height
a Return losses b AR
broadside direction results. The results inFig. 8ashows that the impedance matching was strongly dependent on S2.
When S2was 4 mm, the third mode was at 8 GHz. When
S2 was increased to 5 mm, the third mode was shifted to
7.2 GHz. These results show that the third mode was controlled by S2. Note that the first mode at 2.3 GHz and
the second mode at 5.7 GHz were shifted only slightly.
Fig. 8b shows that the CP mode frequency and AR-BW could be tuned by using different values for S2. When
S2¼ 4 mm, the centre frequency of the CP mode was
3.75 GHz and the 3 dB AR-BW was approximately 45.3%. When S2 was increased to 6 mm, the centre
frequency of the CP mode shifted to 2.95 GHz, but the 3 dB AR-BW was only approximately 28.8%. Based on this result, the characteristics of the CP mode were also controlled by S2.
The simulated results shown in Figs. 7band8b indicate that this method of embedding a rectangular slit on the ground plane excited the CP radiation wave, controlled the CP mode frequency, and achieved a wideband CP characteristic. However, a comparison of Fig. 8a with
Fig. 5 shows that this method caused impedance mismatching from 2 to 4 GHz.
Figure 9 Simulated return losses and AR of the stub height
a Return losses b AR
Table 1 Dimensions of the proposed printed monopole antenna L 45 mm W 40 mm L1 23 mm W1 3 mm L2 19 mm W2 14 mm S1 16 mm W3 1.5 mm S2 4 mm W4 1 mm D1 10.5 mm D2 2 mm
Figure 11 Simulated and measured AR of the proposed Antenna
Figure 10 Simulated and measured return losses against frequency for the proposed Antenna
3.3 Adding a stub to the ground plane
The above discussion on the rectangular slit revealed that it generated wideband CP but caused impedance mismatching from 2 to 4 GHz, thus reducing the impedance bandwidth. To achieve a broad impedance bandwidth and still retain wideband CP, a perturbation stub was added to the ground plane on the right side of the slit. Adding the stub to the left side of the slit degraded the impedance matching.
Fig. 9shows the simulated return losses and AR results at different stub heights (D2), with the other parameters fixed.
Fig. 9ashows that the mode at 4.2 GHz was excited as D2
was increased. This phenomenon improved the impedance matching of the first mode at 2.3 GHz. In addition to
impedance matching, the stub also affected the CP mode frequency. Fig. 9b shows that the CP mode frequency could be tuned by using different values for D2; however,
the AR bandwidth was independent of D2. A comparison
of Fig. 8bwithFig. 9bshows that the CP mode frequency was still mainly controlled by the slit height (S2).
Based on these simulated results for asymmetric feeding with a ground plane embedded with a slit and stub, this work concludes that using asymmetric feeding enhances the impedance bandwidth and excites the EP radiation pattern. The slit embedded on the ground plane generates CP radiation waves but causes impedance mismatching in the band from 2 to 4 GHz. Adding a stub to the ground plane excites a new mode to match the impedance from 2 to 4 GHz, and only slightly affects the CP mode Figure 12 Measured radiation pattern of the proposed Antenna in the XZ- and YZ-plane
a 3.60 GHz b 4.00 GHz
characteristic. Therefore a broad impedance bandwidth and wide AR bandwidth can be simultaneously attained by properly adjusting the sizes of the slit and stub.
4
Simulation and measurement
results
The optimised dimensions of the geometric parameters are listed in Table 1. A comparison of the simulated and measured return losses for the proposed antenna is shown in Fig. 10. The 10 dB impedance bandwidth of the measured return loss reached 6.56 GHz, which covers the range from 2.32 to 8.88 GHz, or approximately 117% with respect to the centre frequency of 5.6 GHz. There was good agreement between the simulated and measured results, with the exception that the measured results shifted to a higher frequency. The simulated and measured AR results for the broadside direction against frequency are plotted in Fig. 11. The disagreement between the simulated and measured results may be attributed to the phase variations of EHor and EVer, which in turn may be due to the
misalignment of the antenna and ground plane in the fabrication process. The phases of the electrical fields were very sensitive to the geometric dimensions. The measured 3 dB AR bandwidth was approximately 1.2 GHz, from 3.2 to 4.4 GHz, which corresponded to approximately 31.6% with respect to the centre frequency of 3.8 GHz. The measured minimum AR was 0.94 dB at 3.6 GHz. The measured results indicate that the proposed antenna demonstrated a broad impedance bandwidth and wide AR bandwidth.
The measured normalised LHCP and RHCP radiation patterns in the XZ-plane and YZ-plane for frequencies of 3.60 and 4.00 GHz are shown in Figs. 12a and b, respectively. Owing to the influence of the asymmetric feed line and ground plane, the CP radiation patterns of the antenna were not omnidirectional, especially in the YZ-plane. The measured 3 dB AR beam widths in the XZ-and YZ-planes were 1028 XZ-and 398, respectively, at 3.60 GHz, as shown in Fig. 12a. As seen in Fig. 12b, the 3 dB AR beam widths were 638 and 318 at 4.00 GHz.
Therefore the AR beam width resulted in a bigger and better performing XZ-plane as compared to the YZ-plane.
Fig. 13 shows the maximum measured gains and radiation efficiencies. The efficiencies were greater than 72% from 2.8 to 4.6 GHz, and the gain variation was less than 1.6 dBi.
5
Conclusion
In this study, we developed a novel microstrip monopole antenna that is capable of realising a broad impedance bandwidth and wide AR bandwidth. It was found that asymmetric feeding changed the radiation wave from LP to EP and increased the impedance bandwidth. A rectangular slit embedded on the ground plane excited a wideband CP and controlled the frequency of the CP mode. A rectangular stub added to the ground plane enhanced the impedance bandwidth and only slightly affected the CP characteristics. The impedance bandwidth achieved measured results of 117% from 2.32 to 8.88 GHz, and a 3 dB AR bandwidth achieved results of approximately 31.6% from 3.2 to 4.4 GHz for LHCP. The proposed antenna would provide numerous advantages for a modern wireless communication system, such as low weight, simple structure, easy fabrication, low production cost, broad impedance bandwidth, and CP radiation pattern.
6
Acknowledgment
This work was supported by National Science Council, Taiwan, under grant no.: NSC 96-2221-E-024-001 and 97-2221-E009-002. The authors are grateful to thank the National Center for High-performance Computing for supports of simulation software and facilities.
7
References
[1] HSU S.H., CHANG K.: ‘A novel reconfigurable microstrip antenna with switchable circular polarization’, IEEE Antennas Wirel. Propag. Lett., 2007, 6, pp. 160 – 162 [2] WAN K.C., XUE Q.: ‘A novel reconfigurable wideband circularly polarized transmitter implemented by indirect-controlled-phased-source’, IEEE Antennas Wirel. Propag. Lett., 2007, 6, pp. 604 – 607
[3] RICHARDS W.F.,LO Y.T.,HARRISON D.D.: ‘An improved theory for microstrip antennas and applications’, IEEE Trans. Antennas Propag., 1981, 29, (1), pp. 38 – 46
[4] LO Y.T.,ENGST B., LEE R.Q.: ‘Simple design formulas for circularly polarized microstrip antennas’, IEE Proc. Microw. Antennas Propag., 1988, 135, (3), pp. 213 – 215
[5] LEE S.K.,SAMBELL A.,KOROLKIEWICZ E.,OOI S.F.: ‘Analysis and design of a circular-polarized nearly-square-patch antenna using a cavity model’, Microw. Opt. Technol. Lett., 2005, 46, (4), pp. 406 – 410
Figure 13 Measured gain and radiation efficiency of the proposed Antenna
[6] TONG K.F.,WONG T.P.: ‘Circularly polarized U-slot antenna’, IEEE Trans. Antennas Propag., 2007, 55, (8), pp. 2382 – 2385 [7] STUTZMAN W.L.,THIELE G.A.: ‘Antenna Theory and Design’ (Wiley, New York, 1998, 2nd edn.)
[8] QIN Y.,GAO S., SMABELL A.: ‘Broadband high-efficiency circularly polarized active antenna and array for RF front-end application’, IEEE Trans. Microw. Theory Tech., 2006, 54, (7), pp. 2910 – 2916
[9] OJIRO Y., HIRAGURI T., HIRASAWA K.: ‘A monopole-fed circularly polarized loop antenna’. Proc. IEEE AP-S Int. Symp., June 1998, vol. 2, pp. 810 – 813
[10] WANG C.J.,LIN Y.C.: ‘New CPW-fed monopole antennas with both linear and circular polarisations’, IET Microw. Antennas Propag., 2008, 2, (5), pp. 466 – 472
[11] IGWE G.I.: ‘Axial ratio of antenna illuminated by an imperfectly circularly polarized source’, IEEE Trans. Antennas Propag., 1987, AP-35, (3), pp. 339 – 342
[12] TOH B.Y., CAHILL R., FUSCO V.F.: ‘Understanding and measuring circular polarization’, IEEE Trans. Edu., 2003, 46, (3), pp. 313 – 318
[13] HFSS – Ansoft Corporation: ‘High Frequency Simulation Software (HFSS) Version 10.0’ (Ansoft Corporation, USA)
