Design of Dual-Band MIMO Antenna with High Isolation for WLAN Mobile Terminal Jung-Nam Lee, Kwang-Chun Lee, Nam-Hoon Park, and Jong-Kweon Park In this paper, we propose a dual-band multiple-input multiple-output (MIMO) antenna with high isolation for WLAN applications (2.45 GHz and 5.2 GHz). The proposed antenna is composed of a mobile communication terminal board, eight radiators, a coaxial feed line, and slots for isolation. The measured –10 dB impedance bandwidths are 10.1% (2.35 GHz to 2.6 GHz) and 3.85% (5.1 GHz to 5.3 GHz) at each frequency band. The proposed four-element MIMO antenna has an isolation of better than 35 dB at 2.45 GHz and 45 dB at 5.2 GHz between each element. The antenna gain is 3.2 dBi at 2.45 GHz and 4.2 dBi at 5.2 GHz. Keywords: MIMO, dual-band antenna, isolation, WLAN, PIFA. Manuscript received Apr. 23, 2012; revised Sept. 4, 2012; accepted Sept. 13, 2012. This work was supported by the IT R&D program of MKE/KEIT, Korea (10038765, Development of B4G Mobile Communication Technologies for Smart Mobile Services). Jung-Nam Lee (phone: +82 42 860 1884, jnlee77@etri.re.kr), Kwang-Chun Lee (kclee@etri.re.kr), and Nam-Hoon Park (nhpark@etri.re.kr) are with the Communications Internet Research Laboratory, ETRI, Daejeon, Rep. of Korea. Jong-Kweon Park (ingpark@hanbat.ac.kr) is with the Department of Radio Wave Engineering, Hanbat National University, Daejeon, Rep. of Korea. http://dx.doi.org/10.4218/etrij.13.0112.0250 ETRI Journal, Volume 35, Number 2, April 2013 © 2013 I. Introduction A technology for innovatively increasing the transfer rate is required to provide an improved quality of service for wireless transmissions. Various kinds of multimedia services are required in a wireless environment. The amount of data to be transmitted has increased, as has the speed of data transmissions. Therefore, we conduct a study on a method for efficiently using limited frequencies. As part of the study, research on a multiple-input multiple-output (MIMO) system that uses a channel in the spatial domain is actively conducted. MIMO technology uses multiple antennas at both the transmitter and receiver to simultaneously transmit multiple signals using the same wireless channel. MIMO technology increases the channel capacity within limited frequency resources and provides a high data transmission rate. Further, MIMO technology can increase the capacity of wireless data tenfold without using additional frequencies and can transmit a wide range of data with high reliability. The capacity of a MIMO system is reduced owing to connections between receiver signals. The connections between signals received from different antenna devices are very important parameters in a MIMO system [1]-[8], such as compact dual-printed inverted-F antenna diversity systems for portable wireless devices [1], a modified planar inverted-F antenna (PIFA) and its array for MIMO terminals [2], an integrated MIMO antenna with a high-isolation characteristic [3], a compact four-element diversity-antenna array for personal digital assistant (PDA) terminals in a MIMO system [4], a three-antenna MIMO system for WLAN operation in a PDA phone [5], a novel miniaturized tri-band PIFA for MIMO applications [6], design considerations for a low antenna correlation and mutual coupling reduction in multi-antenna terminals [7], and a Jung-Nam Lee et al. 177 Antenna 2 50-Ω coaxial (2.45-GHz radiator) feed cable 5.2-GHz radiator Bending line Antenna 1 (2.45-GHz radiator) [mm] 5.2-GHz radiator 6.2 Ground plane (50 mm × 100 mm) of a mobile phone 6 4.5 2 9.5 4.5 5.5 PW2 (4.2 mm) Antenna 4 (2.45-GHz radiator) 5.2-GHz radiator 5.2-GHz radiator Antenna 3 (2.45-GHz radiator) UH:8 mm compact antenna array for MIMO applications at 1,800 MHz and 2,450 MHz [8]. Since a large number of antenna devices are used in a MIMO system, when the antennas are mounted to a mobile terminal, the distance between the antennas becomes very short, which may create stronger connections between them. Since the antennas are connected to each other, a relatively low gain is obtained. In this paper, we propose a dual-band MIMO antenna with high isolation for WLAN applications (2.45 GHz and 5.2 GHz). The proposed antenna is composed of a first radiator with one end connected to a feed line and receiving a signal within the first frequency band (2.45 GHz) and a second radiator having one end connected to a ground plane and receiving a signal within the second frequency band (5.2 GHz). To improve the isolation of the proposed antenna, narrow rectangular slots are inserted on the ground plane [1]. The measured –10 dB impedance bandwidths are 10.1% (2.35 GHz to 2.6 GHz) and 3.85% (5.1 GHz to 5.3 GHz) at each frequency band. The proposed four-element MIMO antenna has an isolation of better than 35 dB at 2.45 GHz and 45 dB at 5.2 GHz between each element. The antenna gain is 3.2 dBi at 2.45 GHz and 4.2 dBi at 5.2 GHz. A B A: feeding point (1.5 mm) Uw:4.2mm B: shorting point (1.0 mm) 1.5 mm (a) HS1 LS1 HS1 HS2 LS2 HS2 5.2-GHz slot 2.45-GHz slot 2.45-GHz slot II. MIMO Antenna Design An important aspect of the proposed antenna is a configuration in which the frequency signals in different bands do not affect each other when a plurality of small antennas are formed in a mobile communication device. In this configuration, the mobile communication device can achieve increased isolation among the plurality of antennas. The proposed antenna is composed of four radiators, an isolation improving slot, a feed line, and a mobile communication terminal board. Figure 1 shows the geometry of the proposed MIMO antenna. The ground plane has a dimension of 50 mm × 100 mm. A dielectric substrate with a height of 0.8 mm and a relative dielectric constant of 4.5 is used. The antenna is excited using a 50-Ω coaxial feed line. The optimal parameters can be chosen as Hs1 = 10 mm, Hs2 = 9 mm, Ls1 = 18 mm, and Ls2 = 21.5 mm based on an extensive simulation using Ansys HFSS [9]. The first radiator (2.45-GHz radiator) has one end connected to a power feeding line and receives a signal within the first frequency band (2.45 GHz). The second radiator (5.2-GHz radiator) has one end connected to a ground plane and receives a signal within the second frequency band (5.2 GHz). The first stub (PW2) is extended from the other end of the first radiator and finely adjusts the signal received by the first radiator. The second stub (UW) is extended from the other end of the second radiator and finely adjusts the signal received by the second 178 Jung-Nam Lee et al. 5.2-GHz slot (c) (b) Fig. 1. Detailed view, (b) slot on ground plane, and (c) photograph of proposed MIMO antenna. radiator and a shorting plate electrically connecting the first radiator to the ground plane. In the antenna described in this paper, the first radiator is only connected to the coaxial feed line, and the second radiator may not be connected to the coaxial feed line but is connected to the ground plane. The signal flows along the ground plane. Therefore, even though the gain is slightly reduced, appropriate isolation between channels is provided, and it is possible to freely tune the received frequency signal. As shown in Fig. 1(b), to improve the isolation between the antennas mounted at the edges of the board, a plurality of slots may be formed in the ground plane [1]. Since it is possible to alter the path of the current flowing directly toward the neighboring antennas through the ground plane using individual slots, the isolation between antennas can be improved. III. Measured and Simulated Results Several cases for the available area are considered and ETRI Journal, Volume 35, Number 2, April 2013 2 1 Case 1 4 2 1 Case 2 3 4 Case 3 3 4 0 2 3 Return loss (dB) 1 –10 –20 S11 (measured) (a) S22 (measured) 0 S33 (measured) –30 S44 (measured) –10 0 –40 –10 S11 (case 1) –50 S21 (case 1) S31 (case 1) –60 S41 (case 1) 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0 Frequency (GHz) Magnitude (dB) (b) 2.8 3.2 3.6 4.0 4.4 4.8 Frequency (GHz) 5.2 5.6 6.0 –30 –40 –50 –60 –10 –70 2.0 –20 S21 (measured) S31 (measured) S41 (measured) –20 0 2.4 2.8 3.2 3.6 4.0 4.4 Frequency (GHz) 4.8 5.2 (b) –30 Fig. 3. Measured (a) return loss and (b) isolation (with slot). –40 S11 (case 2) S22 (case 2) S21 (case 2) S31 (case 2) S41 (case 2) –50 –60 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0 Frequency (GHz) (c) 0 –10 Magnitude (dB) 2.4 (a) –30 Isolation (dB) Magnitude (dB) 2.0 –20 –20 –30 –40 S11 (case 3) –50 S21 (case 3) –60 S41 (case 3) S31 (case 3) 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0 Frequency (GHz) (d) Fig. 2. Comparisons of simulated coupling and reflection coefficients of proposed antenna for three different configurations: (a) three cases, (b) case 1, (c) case 2, and (d) case 3. ETRI Journal, Volume 35, Number 2, April 2013 simulated using Ansys HFSS, and the coupling and reflection coefficients are compared to select the best position for the elements. Figure 2 shows the coupling and reflection coefficients for the three considered cases. The antenna elements are better matched at 2.45 GHz and 5.2 GHz for case 1, which is therefore chosen. Figure 3(a) shows the measured return losses of the proposed antenna for case 1. The antenna is measured using an Anritsu vector network analyzer in an anechoic chamber. The return losses are greater than 10 dB from 2.35 GHz to 2.6 GHz and from 5.1 GHz to 5.3 GHz. Figure 3(b) shows the measured isolation of the four-element antenna of Fig. 1. As can be seen from the figure, the antenna isolations have about 35 dB (S21), 29 dB (S31), and 30 dB (S41) at 2.45 GHz, and 45 dB (S21), 38 dB (S31), and 35 dB (S41) at 5.2 GHz, respectively. Many techniques for improving the isolation have been introduced [10]-[18]. It is difficult to integrate other elements inside the mobile terminal because the electronic band gap bores several holes into the ground plane [10]-[12]. The isolation is much improved when a short-circuit strip with Jung-Nam Lee et al. 179 0 S21(without slot) S31(without slot) S41(without slot) Return loss (dB) Isolation (dB) –10 0 S21(with slot) S31(with slot) S41(with slot) –20 –30 –10 –20 –40 Pw2=4 mm –30 –50 2.40 2.45 Frequency (GHz) (a) 0 –10 Pw2=5 mm Pw2=6 mm 2.50 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0 Frequency (GHz) (a) S21(without slot) S21(with slot) S31(without slot) S41(without slot) S31(with slot) S41(with slot) 0 –20 –10 Return loss (dB) Isolation (dB) Pw2=2 mm Pw2=3 mm –30 –40 –20 Uw=3 mm Uw=4 mm –50 –60 5.10 –30 5.15 5.20 Frequency (GHz) (b) 5.25 Uw=2 mm Uw=5 mm Uw=6 mm 5.30 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 5.6 6.0 Frequency (GHz) (b) Fig. 4. Comparison of measured isolations of proposed antenna: (a) 2.45-GHz isolations (with and without slot) and (b) 5.2-GHz isolations (with and without slot). 0 Without slot With slot S21 S31 S41 S21 S31 S41 2.45 GHz 18 dB 15 dB 17 dB 35 dB 29 dB 31 dB 5.2 GHz 15 dB 19 dB 15 dB 45 dB 38 dB 35 dB Return loss (dB) –10 Table 1. Isolations with and without slot. UH=8 mm UH=7 mm –20 –30 –40 many forms is added between the antennas [13]-[16]. However, the mobile terminal is affected manually. When using the parasitic element [17], [18], the space to arrange another element in the mobile terminal is decreased. In this paper, we form a slot on the ground plane to improve the isolation method [1]. Figure 4 and Table 1 show a comparison of the measured isolations with and without the use of a slot. When a slot is not used, the isolation at operating frequencies of 2.45 GHz and 5.2 GHz is measured at less than –20 dB. On the other hand, when a slot is used, the isolation of the operating frequency is 180 Jung-Nam Lee et al. UH=6 mm 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 Frequency (GHz) (c) 6.5 7.0 Fig. 5. Effects of antenna geometry (PW2, UW, and UH) on return loss: (a) PW2, (b) UW, and (c) UH. measured at over –30 dB. We study the effects of the antenna geometry (PW2, UW, and UH) on the return loss. The effects of a 2.45-GHz tuning stub length (PW2) on the return loss are shown in Fig. 5(a). From the figure, the length of the tuning stub has a profound effect on the lower frequency band (2.45 GHz) but only a small effect on the ETRI Journal, Volume 35, Number 2, April 2013 upper frequency band (5.2 GHz). As the length of the tuning stub (PW2) increases, the operating frequency band (2.45 GHz) is moved to a lower frequency. The effects of the 5.2-GHz radiator length (UW) and height (UH) on the return loss are shown in Figs. 5(b) and 5(c), respectively. As shown in the figures, changes of the 5.2-GHz radiator length (UW) and height (UH) have a negligible effect on the lower frequency band (2.45 GHz). Only characteristics of the upper frequency band (5.2 GHz) are changed. As the length (UW) and height (UH) of the 5.2-GHz radiator increase, the operating frequency band (5.2 GHz) is shifted to the lower frequency. According to the above results, the lower (2.45 GHz) and upper frequency (5.2 GHz) bands can be controlled by changing PW2, UW, and UH. To analyze in detail the effect of the isolation slot on the ground plane, the excited E-field and H-field and the surface current in the ground plane at 2.45 GHz and 5.2 GHz for the proposed MIMO system are studied. The E and H fields and the surface current are simulated using CST Microwave Studio [19]. Figures 6(a) through 6(c) show the simulated E and H fields and surface current for the case of antenna 1 excited at 2.45 GHz. As shown in the figures, it can be seen that the E and H fields and surface current are mostly focused in the 2.45GHz slot and less so in the 5.2-GHz slot. The isolations of S21, S31, and S41 are greatly improved, which agrees with the results shown in Fig. 4(a). The simulated results for the case of antenna 1 excited at 5.2 GHz are shown in Figs. 6(d) through 6(f). As shown in the figures, the E and H fields and surface current are mostly focused in the 5.2-GHz slot and less so in the 2.45-GHz slot. The simulated results in Fig. 4(b) show the improvement in the measured isolations of S21, S31, and S41. Figs. 6(g) and 6(h) respectively show the surface current of 2.45 GHz mostly focused in the 2.45-GHz radiator and the surface current of 5.2 GHz mostly focused in the 5.2-GHz radiator. The radiators (2.45 GHz and 5.2 GHz) do not affect one another. The measured radiation patterns of the proposed MIMO antenna at resonant frequencies of 2.45 GHz and 5.2 GHz are plotted in Fig. 7. The radiation pattern results for three planes are shown, and the Eθ and Eφ components appear to be comparable in all three planes. The radiation patterns of the proposed MIMO antenna are typical of those of conventional PIFAs. For frequencies of over 2.45 GHz (2.35 GHz to 2.6 GHz) and 5.2 GHz (5.1 GHz to 5.3 GHz), the radiation patterns are similar to those shown in Figs. 7(a) through 7(f). Stable measured peak MIMO antenna gains of about 3.2 dBi at 2.45 GHz and 4.2 dBi at 5.2 GHz are obtained. Calculation of the antenna correlation can be done using two methods. The first method uses a far-field radiation pattern [20], and the second method uses the scattering parameters [21]. In ETRI Journal, Volume 35, Number 2, April 2013 (a) (b) (c) (d) (e) (f) (g) (f) Fig. 6. Simulated E-field, H-field, and surface current at 2.45 and 5.2 GHz: (a) 2.45-GHz E-field, (b) 2.45-GHz H-field, (c) 2.45-GHz surface current, (d) 5.2-GHz E-field, (e) 5.2GHz H-field, (f) 5.2-GHz surface current, (g) 2.45-GHz radiator surface current, and (h) 5.2-GHz radiator surface current. this paper, the envelope correlation is calculated using the scattering parameters, based on a formula proposed by Blanch [22]: ρe = * | S11* S12 + S 21 S 22 |2 . 2 (1 − (| S11 | + | S21 | ))(1 − (| S22 |2 + | S12 |2 )) 2 In the design of small antennas on a mobile phone or small electric device, the radiation efficiencies are low, owing to the compressed electrical size of the mobile terminals. Therefore, the energy conservation in the envelope correlation is expected to be inadequate. The expression using the scattering parameters has a disadvantage in the case of small mobile terminal antennas. However, the envelope correlation calculated using the scattering parameters provides sufficiently Jung-Nam Lee et al. 181 0 0 330 0 30 300 30 60 –20 0 330 0 300 60 –20 330 0 –10 30 300 –10 60 90 –20 Eθ Eφ 240 0 210 180 0 120 90 –40 270 –20 240 0 150 210 300 60 90 –20 Eθ Eφ 240 0 210 330 0 –10 210 (a) 0 30 300 –30 270 90 210 330 0 –10 180 2.45 GHz xy-plane 120 150 z 30 90 300 –10 60 120 150 120 210 180 –10 120 150 60 0 300 60 –10 90 –30 270 –10 –10 90 Eθ Eφ Eθ Eφ 10 330 10 30 –10 60 120 150 –50 270 –10 330 10 5.2 GHz xz-plane 120 210 330 180 150 z 30 180 150 y 0 z 30 300 300 60 10 60 –20 90 Eθ Eφ 120 240 210 10 330 180 60 120 210 180 –10 150 90 Eθ Eφ –20 120 240 10 210 10 330 180 150 0 30 300 60 90 –30 270 Eθ Eφ –20 –10 120 240 120 240 0 210 10 330 180 150 210 10 30 90 –20 –20 120 90 210 180 150 (e) 5.2 GHz xy-plane 180 150 210 330 180 z 30 300 –20 210 180 330 –20 0 150 180 150 0 30 300 60 –40 270 120 240 120 210 x –20 90 Eθ Eφ 90 240 0 60 60 Eθ Eφ y 0 30 300 0 150 0 –40 270 –20 120 –40 270 0 –20 Eθ Eφ 240 330 0 60 Eθ Eφ 240 30 –40 270 0 60 0 300 0 150 300 –30 240 60 –30 270 –10 90 0 –50 270 90 Eθ Eφ –30 x Eθ Eφ 0 –30 –50 270 300 –20 10 30 300 10 x –10 –30 –10 330 –50 270 –10 y 0 30 0 0 –30 240 10 –10 90 Eθ Eφ –30 0 0 210 –30 270 –10 330 –20 –10 0 –30 –30 150 180 (d) 0 300 5.2GHz yz-plane 120 240 –20 240 180 210 10 30 –20 210 120 240 0 (c) –10 –10 0 300 0 150 180 90 Eθ Eφ 0 10 30 –20 240 0 330 10 0 –30 270 90 Eθ Eφ –20 60 –20 0 –20 –30 270 210 –20 240 330 30 –30 270 90 Eθ Eφ 240 0 –10 Eθ Eφ 0 –20 –10 –30 270 180 150 0 –10 300 60 –20 x 330 0 180 –20 10 30 60 210 210 (b) 300 y 0 z 120 240 x 0 0 90 Eθ Eφ y 30 150 –20 –10 150 300 –10 0 –20 Eθ Eφ 240 0 –10 60 –20 –10 330 0 180 60 –20 0 0 –20 120 30 300 –30 270 –10 –20 0 150 –20 Eθ Eφ 240 2.45 GHz xz-plane 120 –30 270 90 –20 180 180 330 –10 –40 270 120 210 0 60 Eθ Eφ 240 –10 0 30 300 –20 –40 270 120 150 0 330 0 –20 180 90 –30 270 –20 Eθ Eφ 30 330 0 2.45 GHz yz-plane 0 –20 –20 –40 270 330 0 90 Eθ Eφ 120 240 210 180 150 (f) Fig. 7. Measured radiation patterns of proposed MIMO antenna: (a) yz plane (2.45 GHz), (b) xz plane (2.45 GHz), (c) xy plane (2.45 GHz), (d) yz plane (5.2 GHz), (e) xz plane (5.2 GHz), and (f) xy plane (5.2 GHz). accurate results. Figure 8 shows the simulated envelope correlation 182 Jung-Nam Lee et al. coefficients obtained. The scattering parameters are found using CST Microwave Studio. The calculated envelope ETRI Journal, Volume 35, Number 2, April 2013 1.0 ρe 21 Mobile case ρe 31 Correlation coefficient 0.8 ρe 41 2.45 GHz 5.2 GHz ρe 32 ρe 42 0.6 LCD ρe 43 Battery 0.4 Head phantom 0.2 (a) 0.0 2 3 4 Frequency (GHz) 5 6 Fig. 8. Simulated envelope correlation coefficient. Table 2. Envelope correlation coefficients. Operating frequency 2.45 GHz ρe21 ρe31 ρe41 ρe32 ρe42 ρe43 0.0020 0.0027 0.0110 0.0110 0.0027 0.0028 Operating frequency 5.2 GHz ρe21 ρe31 ρe41 ρe32 ρe42 ρe43 0.0030 0.0005 0.0058 0.0058 0.0005 0.0030 correlation coefficients are depicted in Table 2. As shown in Table 2, the envelope correlation coefficients of the proposed MIMO antenna are substantially less than 0.5 at both 2.45 GHz and 5.2 GHz. Therefore, the four antennas are highly decorrelated. IV. SAR Simulation of Hand and Head Phantoms To analyze the characteristic of the MIMO antenna when the antenna is positioned in the hand and head, the antenna is simulated by applying the hand and head phantoms. The proposed antenna includes a mobile case, LCD, and battery. The mobile case is made of rubber (εr = 3), and the LCD and battery are PEC. The hand and head phantoms are applied to the earlier case. To model the hand and head phantoms, we use the shell (εr = 5, loss tangent = 0.05, and material density = 1,000 kg/m3) and liquid (εr = 42, loss tangent = 0.99, and material density = 1,000 kg/m3) dielectric material. Figure 9 shows the specific absorption rate (SAR) simulation model and the simulated 1-g SAR values for the proposed antenna. The SAR simulated using the IEEE C95.3 averaging method and the input power for the SAR testing is 0.126 W (24 dBm) for the WLAN frequency bands (2.45 and 5.2 GHz) [23]. The SAR values for 1-g tissue in the cases of the hand and ETRI Journal, Volume 35, Number 2, April 2013 (b) Fig. 9. SAR simulation modeling: (a) head phantom and (b) hand phantom. head are studied using the SAR simulation model provided by CST Microwave Studio. The return loss and isolation for the two cases at each operating frequency is given in Fig. 10. Figure 10 shows the simulated return loss, isolation, and correlation coefficients resulting from using the hand and head phantoms. It is clear that there is no distinguishable difference in return loss between using the hand phantom and using only the antenna (without the hand phantom). The isolation level increases with the hand phantom. However, the envelope correlation coefficients are substantially less than 0.5. When applying the head phantom, the lower (2.45 GHz) and upper (5.2 GHz) frequency bands move to the low frequency, and the level of isolation increases. Nevertheless, the envelope correlation coefficients are substantially less than 0.5. The comparison of the simulated results for the three cases (that is, using antenna only, hand phantom, or head phantom) at each operating frequency is given in Tables 3 and 4. Figure 11 shows the simulated SAR field distribution of the hand and head phantoms excited at 2.45 GHz and 5.2 GHz. We simulate the effect of the SAR by the hand and head phantoms. The input power is 24 dBm at 2.45 GHz and 5.2 GHz, and the average SAR of about 1 g is obtained for the position of the hand and head phantoms. The average SARs for when the MIMO antennas are located in the hand and head phantoms are shown in Table 5. V. Conclusion A dual-band MIMO antenna with high isolation for WLAN applications (2.45 GHz and 5.2 GHz) was designed and presented in this paper. The proposed antenna is composed of two radiators, the first with one end connected to a feed line and receiving a signal within the first frequency band (2.45 GHz) and the second with one end connected to the ground plane and receiving a signal within the second frequency band (5.2 GHz). To improve the isolation of the proposed antenna, Jung-Nam Lee et al. 183 0 0 S11 (with hand phantom) S22 (with hand phantom) S33 (with hand phantom) S44 (with hand phantom) –20 –30 2.0 2.4 2.8 3.2 S11 (with head phantom) S22 (with head phantom) S33 (with head phantom) S44 (with head phantom) –30 –40 3.6 4.0 4.4 4.8 Frequency (GHz) (a) 5.2 5.6 6.0 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 Frequency (GHz) (b) 5.2 5.6 6.0 1.0 0 1.0 0.9 –10 0.9 –20 0.8 –20 S21(with hand phantom) S31(with hand phantom) S41(with hand phantom) ρ21(with hand phantom) ρ31(with hand phantom) ρ41(with hand phantom) –40 –50 2.45 GHz –60 0.7 0.6 0.5 0.4 –70 0.3 –80 0.2 –90 0.1 –100 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 Frequency (GHz) 5.2 5.6 0.8 5.2 GHz –30 Isolation (dB) 5.2 GHz Correlation coefficient 0 –10 –30 Isolation (dB) –20 –40 2.45 GHz –50 –60 0.7 0.6 0.5 0.4 –70 0.3 –80 0.2 –90 0.1 –100 2.0 0.0 6.0 S21(with head phantom) S31(with head phantom) S41(with head phantom) ρ21(with head phantom) ρ31(with head phantom) ρ41(with head phantom) 2.4 2.8 3.2 3.6 4.0 4.4 4.8 Frequency (GHz) (c) 5.2 Correlation coefficient –10 Return loss (dB) Return loss (dB) –10 5.6 0.0 6.0 (d) Fig. 10. Simulated isolation and correlation coefficients: (a) return loss with hand phantom, (b) return loss with head phantom, (c) isolation and correlation coefficient with hand phantom, and (d) isolation and correlation coefficient with head phantom. Table 3. Simulated isolation and correlation coefficient comparison. Frequency 2.45 GHz Isolation (dB) 5.2 GHz Phantom S21 S31 S41 Only antenna 35 29 31 Hand 15 15 20 Head 13 14 17 Only antenna 45 Hand 20 23 14 Head 16 15 12 38 35 Frequency 2.45 GHz Correlation coefficient 5.2 GHz Phantom S21 S31 S41 Only antenna 0.0020 0.0027 0.011 Hand 0.027 0.059 0.044 Head 0.093 0.127 0.063 Only antenna 0.003 0.0005 0.0058 Hand 0.029 0.025 0.022 Head 0.042 0.068 0.080 Table 4. Simulated return loss comparison. Frequency Phantom Return loss 2.45 GHz (dB) 184 Port 1 Port 2 Port 3 Port 4 Only antenna 30 15 30 30 Hand 11 11 10 13 Head 12 10 10 11 Jung-Nam Lee et al. Frequency Phantom Return loss (dB) 5.2 GHz Port 1 Port 2 Port 3 Port 4 Only antenna 28 17 32 21 Hand 13 13 14 13 Head 16 9 10 8 ETRI Journal, Volume 35, Number 2, April 2013 (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k) (l) (m) (n) (o) (p) Fig. 11. Simulated SAR field distribution of hand and head phantoms: (a) port 1 at 2.45 GHz, (b) port 2 at 2.45 GHz, (c) port 3 at 2.45 GHz, (d) port 4 at 2.45 GHz, (e) port 1 at 5.2 GHz, (f) port 2 at 5.2 GHz, (g) port 3 at 5.2 GHz, (h) port 4 at 5.2 GHz, (i) port 1 at 2.45 GHz, (j) port 2 at 2.45 GHz, (k) port 3 at 2.45 GHz, (l) port 4 at 2.45 GHz, (m) port 1 at 5.2 GHz, (n) port 2 at 5.2 GHz, (o) port 3 at 5.2 GHz, and (p) port 4 at 5.2 GHz. Table 5. Simulated SAR values. Frequency SAR Phantom Port 1 Port 2 2.45 GHz SAR (W/kg) Hand 2.6 2.5 1.6 2.2 Head 0.7 0.7 0.49 0.52 ETRI Journal, Volume 35, Number 2, April 2013 Port 3 Port 4 Frequency SAR Phantom Port 1 Port 2 Port 3 Port 4 5.2 GHz SAR (W/kg) Hand 2.1 2.2 1.9 1.8 Head 0.3 0.43 0.32 0.19 Jung-Nam Lee et al. 185 narrow rectangular slots are inserted into the ground plane. The measured –10 dB impedance bandwidths are 10.1% (2.35 GHz to 2.6 GHz) and 3.85 % (5.1 GHz to 5.3 GHz) at each frequency band. The proposed four-element MIMO antenna has an isolation of better than 35 dB at 2.45 GHz and 45 dB at 5.2 GHz between each element. The antenna gain is 3.2 dBi at 2.45 GHz and 4.2 dBi at 5.2 GHz. We have shown that the envelope correlation coefficients of the proposed MIMO antenna are substantially less than 0.5. To analyze the characteristic of the MIMO antenna when the antenna was positioned in the hand and on the head, the antenna was simulated by applying the hand and head phantoms. Propag., vol. 54, Jan. 2006, pp. 90-100. [13] A. Diallo et al., “Study and Reduction of the Mutual Coupling Between Two Mobile Phone PIFAs Operating in the DCS 1800 and UMTS Bands,” IEEE Trans. Antennas Propag., vol. 54, Nov. 2006, pp. 3063-3074. [14] A. Chebihi et al., “A Novel Isolation Technique for Closely Spaced PIFAs for UMTS Mobile Phones,” IEEE Antennas Wireless Propag. Lett., vol. 7, 2008, pp. 665-668. [15] K.L. Wong and W.Y. Chen, “Small-Size Printed Loop-Type Antenna Integrated with Two Stacked Coupled-Fed Shorted Strip Monopoles for Eight-Band LTE/GSM/UMTS Operation in the Mobile Phone,” Microw. Opt. Technol. Lett., vol. 52, July 2010, pp. 1471-1476. [16] Z. Jin, J.H. Lim, and T.Y. Yun, “Small-Size and High-Isolation References MIMO Antenna for WLAN,” ETRI J., vol. 34. no. 1, Feb. 2012, [1] M. Karaboikis et al., “Compact Dual-Printed Inverted-F Antenna pp. 114-117. Diversity Systems for Portable Wireless Devices,” IEEE [17] K.S. Min, D.J. Kim, and Y.M. Moon, “Improved MIMO Antenna Antennas Wireless Propag. Lett., vol. 3, no. 1, 2004, pp. 9-14. by Mutual Coupling Suppression between Elements,” European [2] Y. Gao et al., “Modified PIFA and Its Array for MIMO Terminals,” Conf. Wireless Technol., Paris, France, Oct. 2005, pp. 125-128. IEE Proc. Microw. Antennas Propag., vol. 152, Aug. 2005, pp. [18] Z. Li, Z. Du, and K. Gong, “A Dual-Slot Diversity Antenna with 255-259. Isolation Enhancement Using Parasitic Elements for Mobile [3] K. Chung and J.H. Yoon, “Integrated MIMO Antenna with High Handsets,” Asia Pacific Microw. Conf., Singapore, Dec. 2009, pp. Isolation Characteristic,” Electron. Lett., vol. 43, Feb. 2007, pp. 1821-1824. 199-201. [19] Computer Simulation Technology (CST), Microwave Studio [4] C.C. Chiau, X. Chen, and C.G. Parini, “A Compact Four-Element (MWS), 3D EM simulation software. Diversity-Antenna Array for PDA Terminals in a MIMO System,” [20] R.G. Vaughan and J.B. Andersen, “Antenna Diversity in Mobile Microw. Opt. Technol. Lett., vol. 44, Mar. 2005, pp. 408-412. Communications,” IEEE Trans. Veh. Technol., vol. 36, Nov. 1987, [5] K.L. Wong et al., “Three-Antenna MIMO System for WLAN pp. 149-172. Operation in a PDA Phone,” Microw. Opt. Technol. Lett., vol. 48, [21] R.G. Vaughan and J.B. Andersen, Channels, Propagation and July 2006, pp. 1238-1242. Antennas for Mobile Communications, Stevenage, England: IET, [6] M. Manteghi and Y. Rahmat-Samii, “A Novel Miniaturized Jan. 2003. http://www.theiet.org/resources/books/electro/19103.cfm Triband PIFA for MIMO Applications,” Microw. Opt. Technol. [22] S. Blanch, J. Romeu, and I. Corbella, “Exact Representation of Lett., vol. 49, Mar. 2007, pp. 724-731. Antenna System Diversity Performance from Input Parameter [7] J. Thaysen and K.B. Jakobsen, “Design Considerations for Low Description,” Electron. Lett., vol. 39, no. 9, 2003, pp. 705-707. Antenna Correlation and Mutual Coupling Reduction in Multi- [23] IEC, “Human Exposure to Radio Frequency Fields from Handantenna Terminals,” European Trans. Telecommun., vol. 18, 2007, Held and Body-Mounted Wireless Communication Devices – pp. 319-326. Human Models, Instrumentation, and Procedures – Part 1: [8] B. Lindmark and S. Garcia-Garcia, “Compact Antenna Array for Procedure to Determine the Specific Absorption Rate (SAR) for MIMO Applications at 1800 and 2450 MHz,” Microw. Opt. Hand-Held Devices Used in Close Proximity to the Ear Technol. Lett., vol. 48, Oct. 2006, pp. 2034-2037. (Frequency Range of 300 MHz to 3 GHz),” IEC 62209-1, Feb. [9] Ansys, High Frequency Structure Simulator (HFSS), ver. 10, 2005. 2005. [10] D. Sievenpiper et al., “High-Impedance Electromagnetic Surfaces with a Forbidden Frequency Band,” IEEE Trans. Microw Theory Techniques, vol. 47, Nov. 1999, pp. 2059-2074. [11] F. Yang and Y. Rahmat-Samii, “Microstrip Antennas Integrated with Electromagnetic Band-Gap (EBG) Structures: A Low Mutual Coupling Design for Array Applications,” IEEE Trans. Antennas Propag., vol. 51, Oct. 2003, pp. 2936-2946. [12] L. Li et al., “Locally Resonant Cavity Cell Model for Electromagnetic Band Gap Structures,” IEEE Trans. Antennas 186 Jung-Nam Lee et al. ETRI Journal, Volume 35, Number 2, April 2013 Jung-Nam Lee was born in the Rep. of Korea in 1977. He received his BS and MS from the Department of Information and Communication Engineering, Hanbat National University, Daejeon, Rep. of Korea, in 2004 and 2006, respectively. He received his PhD in radio wave engineering from Hanbat National University in 2010. He thereafter joined the Mobile RF Research Team of ETRI, where he is currently a senior member of the engineering staff. His research interests are small antenna, RFID antenna, UWB antenna, and small base station antenna design. Kwang-Chun Lee received his BS/MS in electronics engineering from Chung-Ang University, Seoul, Rep. of Korea, in 1986 and 1988, respectively. Since 1988, he has been with ETRI, Daejeon, Rep. of Korea. His main research interests include high efficiency power amplifiers and RF technologies for mobile communication system design. Nam-Hoon Park received his BS from Chonnam University, Gwangju, Rep. of Korea, in 1984, his MS from Chung-Ang University, Seoul, Rep. of Korea, in 1987, and his PhD from Chungnam National University, Daejeon, Rep. of Korea, in 1999, all in computer science. Since 1988, he has been with ETRI, Daejeon, Rep. of Korea, where he has participated in various communication systems development. He is now working as the leader of the Mobile Communications Technology Research Department. His research interests include the design and development of computer networks, broadband wireless signaling networks, next-generation mobile communication systems, SDR, and cognitive radio. Jong-Kweon Park was born in the Rep. of Korea in 1969. He received his BS in electronics engineering from Kyungpook National University, Daegu, Rep. of Korea, in 1994, and his MS and PhD in electrical engineering from the Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Rep. of Korea, in 1997 and 2001, respectively. From 2001 to 2002, he was a research engineer at ETRI, Daejeon, Rep. of Korea. In 2002, he joined the Department of Information and Communication Engineering, Hanbat National University, where he is currently a professor. His research interests are wave scattering analysis and the design of fractal antennas, ultra-wideband (UWB) antennas, DVB-H antennas, and T-DMB antennas. ETRI Journal, Volume 35, Number 2, April 2013 Jung-Nam Lee et al. 187