DESIGN OF MIMO ANTENNA FOR ISM BAND APPLICATION by Ramandeep Kaur
In recent years, increasing data transmission capacity and reducing multipath fading has attracted the
attention of researchers. With MIMO antenna system, it is being achieved by using spatial diversity and multiplexing techniques. In presented work, a new two element pentagon fractal MIMO antenna system is proposed. To design MIMO system, initially a new geometry of Pentagon Fractal Antenna is designed and explored using IE3D. Using the concept of Koch curve, pentagons of reduced dimensions are iterated and attached to the boundary of base geometry to construct a fractal boundary. The proposed antenna is designed on RT-Duroid substrate having thickness of 3.175 mm. Geometry of second iteration gives higher gain and good impedance matching than base and first iterated geometry. Then, two element system of second iteration pentagon fractal antenna is simulated for MIMO application to compare its output characteristics with single element antenna system. Then, to find the design parameters of proposed geometry for resonant frequency of 2.45 GHz, an ANN is developed using backpropagation algorithm. Using the values of design parameters, a two parallel placed pentagon fractal antenna system is designed for resonant frequency of 2.45 GHz using IE3D. Then, inter-element distance between two antennas is varied and its effect on mutual coupling is observed. To reduce the mutual coupling as well as inter-element distance between two elements, mutual orientation is varied at different angles and results are analysed. Two element pentagon fractal MIMO system having orthogonal orientation and inter-element distance of 10 mm gives isolation of 33.86 dB and ECC of 3.16 x 10-6 which are quite less than threshold values of 15 dB and 0.5 respectively. Another technique of isolation enhancement is also used by inserting parasitic element between two antennas. The prototype of two orthogonally placed pentagon fractal antenna system is fabricated and experimentally verified using Vector Network Analyser (VNA). Measured results show good matching with simulated results.