An L-shaped planar antenna is investigated numerically and validated experimentally. The effect of varying the ratio of the height of vertical segment to the length of horizontal segment is examined in relation to the impedance bandwidth, radiation pattern, and mode of operation. The study shows that the L-shaped planar antennas with different height-to-length ratios operate in different modes and feature distinct impedance and radiation characteristics.
 An inverted-L antenna may be taken as a short monopole with top loading by a horizontal wire element, which is approximately a quarter wavelength long in total. The input impedance of the inverted-L antenna was analyzed by Guertler , but the usefulness of this antenna is limited by its low input impedance and narrow impedance bandwidth. An L-shaped wire monopole antenna has been recently analyzed, and the input impedance characteristics have been investigated within the full range of parameters. It has been shown for this antenna that the ratio of the height of the vertical wire to the length of the horizontal wire determines whether the antenna acts as a monopole or as a transmission-line antenna [Chen, 2000].
 By replacing the wire element of the L-shaped antenna by a planar element, the antenna becomes an L-shaped planar monopole and the impedance bandwidth is increased. A novel feed arrangement for a wideband planar inverted-L antenna has been proposed [Chen and Chia, 2001] and dual frequency operation of the narrowband planar inverted-L antenna has been reported [Kou and Wong, 2000]. It has been shown that the use of planar elements dramatically increases the bandwidth of wire antennas [Agrawall et al., 1998; Ammann, 2000]. Much work has recently been carried out on the planar monopole antenna, with many planar geometries and feed techniques being employed to exploit the wideband properties [Agrawall et al., 1998; Ammann and Chen, 2003]. The height of the square planar monopole is typically about 20% and 43% of a free space wavelength at the frequencies corresponding to the lower and upper edge of the impedance bandwidth, respectively. By bending the planar monopole into an L-shape, its height may be reduced, thus making it more compact. The antenna ultimately operates in the transmission-line mode (patch mode) [Herscovici, 1998] as the height is decreased.
 This paper describes the investigation into the L-shaped planar antennas experimentally and numerically. By examining the VSWR and radiation patterns for the different length ratios of the horizontal and vertical arms, the operation mode and bandwidths of the L-shaped planar antennas are studied. The simulations based on a method of moments (MoM) code augmented by the uniform theory of diffraction (UTD) are verified by the measurements.
2. Antenna Structure
 The radiating element comprises a 36 × 36 mm copper plate of thickness 0.4 mm which is placed above a finite-size ground plane of dimensions 120 × 120 mm. A coaxial probe of radius 0.6 mm and length g = 2.5 mm excites the bottom of the plate through the ground plane via a 50Ω SMA connector. The feedgap, g, is optimized for maximum impedance bandwidth. The plate is then folded through 90° at various heights, h, realizing an inverted-L shape as shown in Figure 1. The ratio of the vertical height h to the horizontal length l is examined in relation to antenna parameters.
3. Results and Discussion
 The lower and upper edge frequencies (2:1 VSWR) of the 36 × 36 mm planar monopole with l = 0 mm were found to be 1.63 and 3.54 GHz, respectively, yielding a fractional impedance bandwidth of 74%. The planar element was then folded at various heights, realizing a monopole with an inverted-L shape. When the ratio (h:l) of the vertical height h to the horizontal length l is large (>3:1), the effects of the bend on the antenna parameters such as impedance bandwidth and radiation pattern are small. However, as this ratio becomes smaller (<2:1), the lower edge frequency f1 increases significantly, and the upper edge frequency f2, in general, remains relatively unchanged. Thus the impedance bandwidth decreases as the ratio becomes smaller. The measured return loss over the frequency range from 1 to 6 GHz is shown in Figure 2, for the heights from h = 36 mm to h = 0 mm in 9 mm steps. For h = 0 mm, the antenna effectively becomes an edge-fed air-spaced patch, which explains the poor return loss. (Note, for h = 0 mm the separation between the groundplane and the plate is g = 2.5 mm.) Figure 3 illustrates a plot of the upper and lower edge frequencies f1 and f2, for both 2:1 and 3:1 VSWRs, against the monopole height, h. The dependence of the lower edge frequency, f1, on the ratio (h:l), can be seen clearly from these plots. It can be seen that the VSWR remains greater than 2:1 for h = 0 mm. However, no change of antenna mode is noticeable from the VSWR data.
3.2. Radiation Patterns
 The radiation patterns were measured at 2.0 and 3.6 GHz, and the maximum gains found to be 4.3 and 5.2 dBi, respectively, for h = 36 mm. The patterns were also simulated using a hybrid MoM/UTD code. The measured and simulated patterns, which are in good agreement, are illustrated in Figures 4–6 for the heights h = 36, 20 and 0 mm. Very little variation was observed for the H-plane (ϕ, θ = 90°) patterns which were quasi-omnidirectional in shape for all ratios of (h:l) at both frequencies. Hence, for brevity, only the measured patterns with the greatest difference are shown. At 2 GHz, all the H-plane patterns were found to be purely omnidirectional (±1 dB), as was the case for small (h:l) at 3.6 GHz. As the ratio increased, at 3.6 GHz, some deviation from omnidirectional (±3 dB) was observed as shown in Figure 4. Figure 5 illustrates the E-plane (θ, ϕ = 0°) patterns taken at 2 GHz, which are in good agreement with the simulated data. The monopolar null (θ = 0°) is seen to be about 30 dB for h = 36 mm in Figure 5a. This null becomes shallower as h decreases (as (h:l) decreases) and is seen to be about 10 dB with some 10° offset from zenith for h = 20 mm and 3 dB for h = 0 mm in Figures 5b and 5c, respectively. The decrease of this null shows a change of operating mode from monopolar to the transmission-line mode (patch mode). At 3.6 GHz, the null depth is 30 dB for h = 36 mm and is negligible for h = 0 mm and h = 20 mm as shown in Figures 6a–6c. Further investigation found the null depth at 2 GHz to be 3 dB for h = 5 mm and 20 dB for h = 17 mm, and at 3.6 GHz, the 3 dB and 20 dB nulls were observed for h = 21 mm and h = 28 mm, respectively. It is more important to notice that the change of operating mode from the monopolar mode to patch mode is not abrupt and features of both modes exist over a range of (h:l) ratios. We suggest that the patch mode exists alone when the null at θ = 0° is less than 3 dB and that the monopolar mode exists alone when the null at θ = 0° is greater than 20 dB. Both modes exist for a null depth between 3 and 20 dB, which may be offset from θ = 0°. Table 1 illustrates the heights, h, which correspond to these values at both 2 and 3.6 GHz. The percentage of operating wavelength corresponding to the height of the antenna is also given. Measured cross-polarization levels are shown in the E-plane plots and also indicate the reduction and ultimate disappearance of the null at θ = 0 as (h:l) decreases.
Table 1. Values of Height, h, Corresponding to Null Ranges 0–3 dB and 3–20 dB at Both 2.0 GHz and 3.6 GHz
 The broadband L-shaped planar antennas have been investigated experimentally and numerically. The effects of varying the ratios of height of the vertical segment to length of horizontal segment on the antenna parameters have been shown to be significant, having a major effect on impedance bandwidth and radiation patterns (operating mode). The conclusion drawn from the investigations is helpful to understand the transition of the operating mode of the L-shaped antenna.