**2.6 Fractal antennas**

The fractal technique is helpful to reduce the antenna size owing to the self-similar and space filling properties, thus it is used to design the MIMO antenna for the compact purpose.

In [83, 86, 88], the Koch fractal technique were adopted to design the MIMO antenna. The iteration process of [83] is shown in **Figure 10**. The initial shape is the octagon of **Figure 10a**, the fractal shape after first iteration is shown in **Figure 10b**, and the second iteration has satisfied the requirement of the UWB band. The corresponding MIMO antenna is shown in **Figure 11**. The C-shape slot is sliced on the

#### **Figure 9.**

*Antenna configuration (left: (a) Top view and (b) cross-sectional view) and current distributions at different frequencies (right: (a) Port 1 excited at 3.4 GHz, (b) Port 2 excited at 3.4 GHz, (c) Port 1 excited at 3.8 GHz, and (d) Port 2 excited at 3.8 GHz) of [30].*

**Figure 10.** *Iteration process of Koch fractal [83]: (a) initiator, (b) first iteration, and (c) second iteration.*

*Techniques for Compact Planar MIMO Antennas DOI: http://dx.doi.org/10.5772/intechopen.112040*

**Figure 11.** *Fractal MIMO antenna of [83].*

fractal octagon to realize the rejection band, the orthogonal arrangement and a L-shape stub connected with the ground plane are used to increase the isolation.

The hybrid fractal technique is also used to design the UWB antenna [85], where both the Sierpinski and Koch fractal were applied, and a U-shape slot etched in the radiating element to notch the WLAN band. For the MIMO antenna design, the two antennas are put parallel, but the isolation is increased by both the stepped ground plane and the reflecting ground stub in between.

Using the similar idea as in [85], the authors in [87] combine the Koch and the Minkowski to get the dual-band MIMO antenna. The corresponding hybrid fractal idea and the MIMO antenna are shown in **Figure 12**. There is an initiator and a generator for the Minkowski fractal technique as shown in the left subfigure of **Figure 13**. This structure meets the dual-band requirements of 1.65–1.90 GHz and 2.68–6.25 GHz, and the high isolation is achieved with the help of the T-shape stub connected to the ground plane.

In contrast, the hybrid Quadric–Koch fractal antenna was designed in [89] for the multi-band requirement where the circular polarizations were obtained for the bands of 3.66–3.7 GHz and 5.93–6.13 GHz and the rest five bands are linear polarization. No additional structure was used in the MIMO antenna, where two elements are put symmetrically, and the isolation is better than 17 dB over the entire bands.

**Figure 12.** *Hybrid fractal MIMO antenna of [85].*

**Figure 13.** *The elements of MIMO antennas in [109] (left) and [110] (right).*

The Hilbert fractal technique was employed in [90] to realize the dual-band property, the correlation coefficients lower than 0.1 when the two orthogonal-arranged antennas are put close.

#### **2.7 Radiator-cutting antennas**

Though the fractal technique reduces the antenna size significantly in physical, there is another way to reduce the antenna size physically, that is, by cutting the original antenna into a small piece. This method is suitable for the antenna with the symmetrical structure or the cavity antenna who has the symmetrical modes.

In the works of [109, 110], the radiator is cut into two pieces and leave one for the radiation. The corresponding element structures of the MIMO antennas are shown in **Figure 13**. It is clear that the antennas have no complete structures. It is the quasi-selfcomplementary monopole in [109] owing to the monopole has the semi-circle patch while the slot in the ground is not complete half-complementary structure. This cutting structure keeps the UWB property as that of the complete monopole. Through the symmetrical arrangement of two elements, the MIMO antenna has high isolation without any additional structure due to the asymmetry structure.While in [110], the rectangular monopole is not only cut into two pieces, but also a semi-circle slot is etched in the half-monopole to improve the optical transparency. However, the ground plane is modified by etching a slot and adding staircase stubs resulting in the improvement of the impedance bandwidth.

The non-integer order mode cutting technique is also adopted to reduce the antenna size physically. It is the same as the radiator-cutting method. That is because the complete structure has the integer mode, if we want to use its non-integer mode we have to cut the structure. In other words, the structure cutting means the mode cutting.

The non-integer mode structure of rectangular cavity in [80] has been shown in Section 2.1, where the length of the patch is half of the complete one so that the TM1*<sup>=</sup>*2,0 mode etc. form. The same idea appears in [82], but one 1/8 mode is used for the circular cavity. While the semi-taper slot is etched on the top lay of the cavity in [111] in order to radiate the related power, where the four CPW-fed MIMO antenna is shown in **Figure 14**. The metallic vias are set to form five cavities, the outer four with the semi-taper slots are responsible for the radiation. The half mode of TM110 will generate by the proper feeding, and this MIMO antenna has high isolation owing to the orthogonal arrangement and the existence of the metallic vias.

*Techniques for Compact Planar MIMO Antennas DOI: http://dx.doi.org/10.5772/intechopen.112040*

**Figure 14.** *MIMO antenna of [111]: (A) geometry and (B) unit.*
