Abstract
A compact high isolation ultrawideband (UWB) multiple-input-multiple-output (MIMO) antenna is designed. The proposed MIMO antenna consists of a rectangular monopole antenna and a slot antenna fed by two microstrip lines, respectively. To improve the impedance matching, a circular coupling structure is designed to feed the tapered slot antenna. The parasitic resonance introduced by the ground stub helps to extend the impedance bandwidth of monopole antenna at the upper UWB band. Commonly used complex decoupling or coupling structures are eliminated that endow the proposed antenna minimized foot print, which is preferred in mobile handset. Although without decoupling structure, high isolation is obtained between two antenna elements. Simulation and measurement verify the antenna’s desirable performance, showing a broad impedance bandwidth of 3.1–10.6 GHz with |S11| < −10 dB and |S21| < −20 dB over 3.4–10.6 GHz, and |S21| < −18 dB from 3.1–3.4 GHz.
1. Introduction
Wireless communication systems rely on ultrawideband (UWB) and multi-input-multi-output (MIMO) technologies for high-speed data streaming [1]. Multiple antennas are employed in MIMO systems to overcome multipath fading, to improve channel capacity, and to ensure link quality [2].
As important devices of UWB MIMO systems, MIMO antennas have been investigated widely in recent years. Various decoupling structures were used to improve isolation, which is an important parameter of MIMO antennas. A compact UWB MIMO antenna with dual band-notched 3.30–3.70 GHz and 5.15–5.85 GHz was reported in [3]. Several metal strips and perturb ground plane, as decoupling structure, lead to good isolation. In [4], a couple of slots are etched on the ground plane to improve the isolation of the antennas. In [5], inverted L-shaped branches and a polygon slot function as the decoupling structure to control mutual coupling. Two inverted C-shaped ground metal strips are placed between the two radiating elements to achieve high isolation [6]. A protruded ground plane was used in [7] to improve the isolation of the MIMO antenna. Instead of introducing decoupling structure, a coupling element was introduced to enhance the isolation between two antennas in [8]. Though the abovementioned decoupling and coupling structures show their effectiveness in isolation improvement, the enlarged foot print is not preferred in a mobile handset.
A compact UWB MIMO antenna with polarization diversity was proposed in [9], with a size of 25 × 40 mm2. The minimized foot print is obtained at the cost of narrowed bandwidth. Moreover, the two antenna elements did not have a common ground plane, which makes it difficult to be used in practice. Dual-orthogonal polarization was used in [10] for high isolation, but it requires a complex feed network to excite the two polarizations. A diversity UWB antenna was proposed in [11] with limited bandwidth. In [12], polarization diversity was also used to improve the isolation, but antenna has a large size. A carbon black film was used in [13] to improve the isolation of two antenna elements, but it will increase the cost of systems.
A UWB MIMO antenna with high isolation and compact size is presented in this paper. The proposed MIMO antenna is composed of a monopole antenna and a slot antenna. A ground stub introduces parasitic resonance that helps to extend the antenna’s impedance bandwidth at the upper UWB band. Good isolation is achieved without additional coupling or decoupling structures. The proposed antenna features impedance bandwidth 3.1–10.6 GHz. The mutual coupling is less than −18 dB from 3.1–3.4 GHz and less than −20 dB from 3.4–10.6 GHz.
2. Antenna Design
2.1. Geometry of UWB MIMO Antenna
The geometry of the proposed UWB MIMO antenna is shown in Figure 1. The slot antenna and the rectangular monopole antenna, denoted as element 1 and element 2, respectively, are fed by two 50 Ω microstrip lines. Element 1 consists of two tapered stepped slots. Element 2 is a monopole antenna. The microstrip feeding lines and the monopole are on the top of the substrate and the ground plane and the slot antenna are on the bottom. Port 1 and port 2 are located at the opposite flank of the substrate which helps improve the isolation. Both two elements are designed of functional bandwidth from 3.1 to 10.6 GHz. Dimensions of the antenna are 25 × 35 × 0.8 mm3. FR4 () is used as substrate.
As for a monopole, the gap between the radiator and the ground plane is a dominant factor on its impedance bandwidth. Besides tuning this gap, we add a parasitic stub between the monopole and slot antenna for bandwidth extension. For the initial monopole antenna structure, the fundamental resonant frequency can be calculated by [14] where l1 and l2 are the length of the ground plane and the monopole antenna patch and g is the gap between them. A1 and A2 are the area of the ground plane and monopole antenna patch. l1, l2, and g are in centimetres. is the effective dielectric constant.
The slot antenna is designed based on [15, 16]. A linear tapered slot combined with stepped slot as a radiator and a microstrip to slot transition also help to achieve the desirable bandwidth of the slot antenna.
The design values of the antenna parameters are as follows:
w1 = 4.5 mm, w2 = 1.6 mm, w3 = 4.9 mm, w4 = 5 mm, w5 = 7.4 mm, ws = 0.3 mm, w7 = 1 mm, w8 = 11 mm, w9 = 1.6 mm, w10 = 8.4 mm, w11 = 2.27 mm, w12 = 1.86 mm, w13 = 3.54 mm, L1 = 8 mm, L2 = 13 mm, L3 = 14 mm, L4 = 12 mm, L5 = 15.4 mm, L6 = 1 mm, L7 = 0.34 mm, L8 = 1.5 mm, L9 = 10 mm, L10 = 2.66 mm, L11 = 5 mm, R1 = 1.9 mm, R2 = 1.5 mm, Ls = 5 mm, w = 35 mm, L = 25 mm, and g = 1 mm. The photographs of the antenna are shown in Figure 2.
(a)
(b)
2.2. Effects of the Parasitic Stub
The influence of the parasitic stub on the antenna’s input impedance is analysed in Figure 3. It shows that the stub extend the impedance of element 2 toward the upper band. That is because of the multiple parasitic resonances brought about by the stub from 6 GHz to 11 GHz. The fundamental resonance of the monopole at 4 GHz is not much affected by the parasitic stub.
Moreover, current distributions of the MIMO antennas with and without the stub in Figure 4, indicating the performance of the proposed antenna at low frequency, are not affected by the stub. For example, as shown in Figure 4(a), the current distribution of the antenna at 3.5 GHz is stable. At 9 GHz, intense current is observed on the stub in Figure 4(b), which proves the resonance brought about by the stub at the upper UWB band.
(a)
(b)
Figure 5 shows the simulated |S22| of the MIMO antenna with varying Ls (length of stub) and Ws (width of stub). After the parameter optimization, Ls is chosen to be 5 mm. As shown in Figure 5(a), when increasing or decreasing Ls, the |S22| became deteriorated for upper UWB frequencies. In Figure 5(b), |S22| varies dramatically from 5 GHz to 11 GHz with different Ws. Ws is chosen to be 0.3 mm for a slightly broader impedance bandwidth.
(a)
(b)
2.3. Parameter Study of R1 and R2
The coupling between the feeding structures is influential on the antenna’s impedance bandwidth. The proposed slot antenna is fed by coupling a circular patch (radius R1) with a circular slot (radius R2). The effects of variations in different R1 and R2 are shown in Figures 6 and 7.
Figure 6 shows the slot’s reflection coefficient against different R1. When R1 is increased from 0.9 mm to 2.4 mm, the |S11| curve is nearly unaffected from 7.5 GHz to 11 GHz. Two parasitic resonances appear at 3.5 and 5.5 GHz when R1 is 1.9 mm. When R1 is 1.4 mm, the two parasitic resonances merge into one, resulting in a narrow bandwidth. If R1 increases to 2.4 mm, the two resonances diverge toward the lower and upper side of the UWB band, and then a notch band of 3.7–5 GHz appears. In order to cover UWB band, R1 is chosen to be 1.9 mm.
R2 is the other influential parameter, as shown in Figure 7. It is found that the impedance bandwidth of the antenna is stable versus R2 from 8–11 GHz, except for resonance at the lower side band. Different values of R1 and R2 lead to varying coupling coefficient, which changes the impedance matching of element 1. Finally, R2 is chosen to be 1.5 mm.
3. Results and Discussions
3.1. Impedance Bandwidth
The antenna is simulated by the full-wave electromagnetic High Frequency Structure Simulator (HFSS) and measured with an Agilent E5071C network analyser. The simulated and measured |S11|, |S22|, and |S12| of the proposed antenna are shown in Figure 8. Figures 8(a) and 8(b) show the proposed UWB MIMO antenna has an impedance bandwidth with |S11| < −10 dB from 3.1 GHz to more than 11 GHz. As shown in Figure 8(c), the mutual coupling (in terms of |S12|) is below −18 dB for 3.1–3.4 GHz and below −20 dB over 3.4–10.6 GHz. In general, the simulated and measured results are in good agreement. Discrepancies between measured and simulated results come principally from fabrication tolerances.
(a)
(b)
(c)
3.2. Radiation Patterns
The simulated and measured 2D radiation patterns at 4 GHz and 10 GHz are shown in Figure 9, when port 1 is excited, the port 2 is terminated with a 50 Ω load and vice versa. Figure 9 shows that the measured data agree with simulated results.
3.3. MIMO Characteristics
The envelope correlation coefficient between the two elements can be calculated using [17].
Based on the simulated and measured results, the ECC are calculated and shown in Figure 10. All the values are smaller than 0.005 for the whole UWB band, which guarantees the UWB MIMO antenna has good diversity performance.
4. Conclusions
A novel UWB MIMO antenna has been presented in this paper. Without decoupling structure, high isolation has been achieved by using different types of antenna elements. The UWB MIMO antenna has a compact size and a simple structure. The ground stub between the two antenna elements and the slot line transition were introduced to improve impedance matching of the two elements. Simulation and measurement results show the proposed antenna has an impedance bandwidth that covers the whole UWB band with isolation below −18 dB.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
This work is supported in part by the Natural Science Foundation of China under grant 61401106 and in part by the Youth Foundation of Guangdong University of Technology (15ZK0038).