Ultrawideband (UWB) wireless communications systems operating from 3.1 to 10.6 GHz offer low-power, high-speed links for a variety of applications. However, they must often compete with interfering signals from a number of existing transmitters that may fall within the wide UWB operating range. Of course, stop-band filters can be added to a system where needed to remove interference, but adding filters increases system complexity and cost. A better approach is to implement the desired signal rejection as part of the UWB antenna. To reject these unwanted signals, a compact printed rectangular monopole antenna with dual stop-band characteristics was developed measuring just 29 x 31 mm. The antenna is created by etching a complementary split-ring resonator (CSRR) inside the patch of the monopole antenna.

By varying the CSRR’s parameters, such as the width of the outer and inner split-ring, the stop bands can be adjusted. Using this approach, an UWB antenna was designed with an impedance bandwidth of 3.1 to 12.0 GHz with a voltage standing wave ratio (VSWR) of less than 2.0:1 except across the two stop bands of 3.23 to 3.76 GHz and 5.43 to 5.78 GHz. The antenna provides peak gain of 3.98 dBi with an omnidirectional H-plane radiation pattern.

UWB technology is a lowpower method of transmitting large amounts of data across short distances, at speeds exceeding 100 Mb/s.¹ The United States Federal Communications Commission (FCC) has allocated a wide range of frequencies, 3.1 to 10.6 GHz, for low-power UWB applications.² Within the UWB frequency range, however, numerous narrowband wireless applications already exist, including IEEE 806.16 WiMAX (3.3 to 3.6 GHz) and IEEE 802.11a wirelesslocal- area-network (WLAN) systems (5.15 to 5.825 GHz), which may generate electromagnetic interference (EMI) that could potentially disrupt the operation of a lowpower UWB system. To minimize interference between UWB systems and existing wireless communications and other wireless applications, it is necessary to filter any interference signals radiating from an UWB system by means of several band-reject filters connected to the UWB antenna.

Of course, adding two filters to an UWB system does increase its complexity. A simpler way to reduce the effects of interference from an UWB system is to design an UWB antenna with band-stop characteristics at the required frequencies. Several such designs have already been proposed.^3-10 However, these band-stop antenna designs have generally focused on only one rejection band, between 5 and 6 GHz, and have not addressed the complexity of implementing two stop-band frequencies within a single antenna design.

One of the ways to accomplish two stop bands in an UWB antenna is by etching an CSRR inside the patch of a planar antenna, as discussed in ref. 11. As detailed in refs. 11-13, the CSRR resonator structure can be considered as an electrically small resonator with very high quality factor (Q). The high Q capability of the resonator is well suited for constructing filters with sharp and deep notch characteristics for effective rejection of unwanted signals. In addition, the size of the CSRR is one-tenth or less than the guided wavelength at the resonant frequency,^14-16 making CSRRs resonator well suited for implementation within extremely compact microstrip antennas.

To explore the possibilities of incorporating dual stop bands in an miniature UWB antenna, it was necessary to develop a prototype so that simulated performance parameters could be compared with actual measured performance. For this purpose, a novel rectangular patch microstrip monopole antenna was designed for the UWB frequency range with notches centered at 3.53 and 5.66 GHz. The dual-band notch characteristics are achieved by adjusting the size of the CSRR resonator etched inside the patch of the proposed monopole antenna. A prototype antenna was designed and fabricated, with a full set of measurements made on the prototype for comparison with simulated data.

Figure 1 shows the geometry of the proposed antenna and geometry of the CSRR resonator. The proposed antenna has dimensions of 29.0 x 31.0 mm and is fabricated on inexpensive reinforced-Teflon/ duroid (RT/duroid) substrate with relative permittivity, e_r , of 3.38 and thickness of 0.813 mm. The proposed antenna originates from a conventional rectangular monopole antenna with two notches cut at the bottom of the patch antenna. A coplanar-waveguide (CPW) transmission line, with a signal strip thickness of Wf, is used as the antenna feed. A partial ground plane with dimensions of length L_g and width W was printed on the back of the substrate. The spacing between the rectangular patch and partial ground plane in the back of the substrate is denoted as g. By properly selecting the parameters for the two notches and the partial ground plane numerically and experimentally, the antenna can be fabricated for full coverage of the UWB frequency range. With experimentation, it is also possible to

reduce the antenna’s dimensions. The antenna’s various geometric parameters were adjusted and optimized by means of the Microwave Studio suite of computer-aided-engineering (CAE) software tools from CST.¹⁷ Optimal values for the various antenna parameters are listed in the table. Figure 2 shows the fabricated prototype of the proposed UWB antenna.

Figure 3 shows the simulated and measured VSWR versus frequency for the designed UWB antenna. The antenna’s performance was simulated using Microwave Studio and the VSWR was measured with a model PNA microwave vector network analyzer (VNA) from Agilent Technologies. It can be seen that the antenna has two stop bands, at 3.32 to 3.67 GHz and at 5.14 to 5.72 GHz according to the simulated results, and at 3.23 to 3.76 GHz and 5.43 to 5.78 GHz according to the measured results. These stop bands were achieved while maintaining the UWB passband antenna characteristics with a VSWR of less than 2.0:1 from 3.1 to 12.0 GHz.

The discrepancies between the simulations and the measured results can be most likely attributed to errors in fabrication and variations in fabricating the microstrip lines in the laboratory, compared to the dimensions used for the simulation. Substrate thickness and dielectric-constant tolerances, as well as manufacturing tolerances, can also contribute to the discrepancies between the simulations and the measured results.

Simulated and measured radiation patterns for the dual-notch UWB antenna are shown in Figs 4, 5, and 6 for three different frequencies at 3, 6, and 9 GHz, respectively. The simulated and measured radiation patterns reveal an omnidirectional radiation pattern in the H-plane and a dipolelike radiation pattern in the E-plane for the compact UWB antenna. Since the measurements were not performed in free space, the transmitted energy from the antenna may have been absorbed in part by the test instruments in the laboratory; this is the reason that the measured radiation patterns are not as good as the simulated radiation patterns. Significant fluctuations appear in the radiation patterns

at higher frequencies (Fig. 6). This performance deterioration at higher frequencies can be attributed to the fact that the equivalent working area of the antenna circuit must radiate over a wide frequency range, with some compromises necessary in the antenna’s dimensions in order to cover that wide frequency range. In addition, it is likely that the larger magnitude of the higher-order modes and unequal phase distributions at the higher frequencies of the coverage range also contribute to the degradation of the radiation patterns at higher frequencies.¹⁹

Continue to page 2