I. Introduction
Patch antennas are receiving increasing interest in various mobile communication systems since they can provide advantages over traditional whip and helix antennas in terms of high efficiency, low EM coupling to the human head, and increased mechanical reliability [1]. In many applications, the requirements on both bandwidth and physical size are quite stringent. Typically, the antenna is required to have a bandwidth exceeding 10%, a resonant length of about , and thickness not exceeding in one of the bands, where is free space wavelength. During the last two decades, many investigators have dedicated their efforts to creating new design or variations to the original antenna that produce either wider bandwidths or multiplefrequency operation in a single element [2]–[18]. Regarding the bandwidth enhancement of patch antennas, several techniques have been proposed, including the use of an impedance matching network [2], the use of multiple resonators [3], and the use of thick substrates [4], [5]. For an electrically thick substrate patch antenna, coaxial feed is typically used. However, the increased inductance introduced by the longer probe will limit the achievable bandwidth to less than 10% of the resonant frequency. For this reason, several other methods [6] [7] [8] have been proposed to solve this problem, including etching a small circular slot [6], cutting a U-slot [7]on a patch, and the use of an L-probe feed [8]. Unlike most of the bandwidth-widening methods in the literature, the U-slot patch antenna proposed in [7] retains the major advantage of a single patch on one layer. Subsequently, many published results have confirmed the broadband characteristics of the U-slot patch antenna or its variations [9] [10]–[13]. Both the U-slot patch and the L-probe feed patch can attain over 30% bandwidth. However, the substrate thickness typically is about . Meanwhile, many solutions to achieve multiple-frequency operation were carried out [14]–[18], such as the multilayer stacked-patch antenna [14], the reactive-loading patch antennas by adding
Geometry of the antenna with .
Measured SWR of the antenna in Fig. 1.
shorting pins and/or etching slots on a patch [15]–[17], or more recently proposed fractal antennas [18].