sensors Article

Throughput Measurement of a Dual-Band MIMO Rectangular Dielectric Resonator Antenna for LTE Applications Jamal Nasir 1,2 , Mohd. Haizal Jamaluddin 1, *, Aftab Ahmad Khan 2 , Muhammad Ramlee Kamarudin 1 , Chee Yen Leow 1 and Owais Owais 2 1 2

*

Wireless Communication Centre, Universiti Teknologi Malaysia, 81310 UTM Skudai, Johor, Malaysia; [email protected] (J.N.); [email protected] (M.R.K.); [email protected] (B.L.C.Y.) Department of Electrical Engineering, COMSATS Institute of Information Technology, 22060 Abbottabad, Pakistan; [email protected] (A.A.K.); [email protected] (O.O.) Correspondence: [email protected]; Tel.: +60-7-553-6107 (ext. 36107)

Academic Editors: Yuh-Shyan Chen and Leonhard M. Reindl Received: 5 October 2016; Accepted: 10 January 2017; Published: 13 January 2017

Abstract: An L-shaped dual-band multiple-input multiple-output (MIMO) rectangular dielectric resonator antenna (RDRA) for long term evolution (LTE) applications is proposed. The presented antenna can transmit and receive information independently using fundamental TE111 and higher order TE121 modes of the DRA. TE111 degenerate mode covers LTE band 2 (1.85–1.99 GHz), 3 (1.71–1.88 GHz), and 9 (1.7499–1.7849 GHz) at fr = 1.8 GHz whereas TE121 covers LTE band 7 (2.5–2.69 GHz) at fr = 2.6 GHz, respectively. An efficient design method has been used to reduce mutual coupling between ports by changing the effective permittivity values of DRA by introducing a cylindrical air-gap at an optimal position in the dielectric resonator. This air-gap along with matching strips at the corners of the dielectric resonator keeps the isolation at a value more than 17 dB at both the bands. The diversity performance has also been evaluated by calculating the envelope correlation coefficient, diversity gain, and mean effective gain of the proposed design. MIMO performance has been evaluated by measuring the throughput of the proposed MIMO antenna. Experimental results successfully validate the presented design methodology in this work. Keywords: dielectric resonator antenna; MIMO; LTE; mutual coupling

1. Introduction LTE developed by third generation partnership project (3GPP) is a fourth generation (4G) wireless standard that provides 10 times the speed of 3G network [1]. It is designed to provide IP-based data, voice, and multimedia streaming at a speed ranging from 100 Mbps to 1 Gbps. Because of this dramatic increase in capacity and speed of data transmission, multiple antenna techniques have gained overwhelming effects in the area of wireless communication throughout the last decade. However, designing of antennas for such a high-speed data transmission poses challenges in terms of isolation between antenna elements and the size of the antenna which may ultimately affect system performance. A considerable amount of work has already been done in the area of microstrip patch antennas (MPAs) but they suffer from low bandwidth [2] and metallic losses at high frequencies. Dielectric resonators when used as a radiator have low losses, high radiation efficiency, and high bandwidth as compared to the conventional MPAs [3]. Importantly, various modes can be excited with diverse radiation characteristics using a single dielectric resonator which makes it an appropriate alternative for multifunction applications [4]. Ishimiya et al. for the first time gave the concept of a MIMO DRA and achieved 10 dB diversity gain [5] which is comparable with MIMO dipole array used in IEEE 802.11n, but no explicit design Sensors 2017, 17, 148; doi:10.3390/s17010148

www.mdpi.com/journal/sensors

Sensors 2017, 17, 148

2 of 15

Ishimiya et al. for the first time gave the concept of a MIMO DRA and achieved 10 dB diversity Sensors 2017, 17, 148 2 of 15 gain [5] which is comparable with MIMO dipole array used in IEEE 802.11n, but no explicit design method is given in this paper. A dual-port MIMO design at 700 MHz for LTE Femtocell base station applications is proposed in [6] A that achieved channel capacity to 11.1 with excellent port method is given in this paper. dual-port MIMO design at 700up MHz for bps/Hz LTE Femtocell base station isolation. In [7], the authors proposed a singlechannel elementcapacity dual-port DRA at 2.6 GHz applications is proposed in [6] that achieved upMIMO to 11.1rectangular bps/Hz with excellent port excited byIntwo lines. Acceptable results have been obtained of isolation. [7],symmetric the authorsmicrostrip proposed afeed single element dual-port MIMO rectangular DRAin at terms 2.6 GHz S-parameters, pattern plots, correlation effective gain, and diversity gain. excited by tworadiation symmetric microstrip feed lines. coefficient, Acceptablemean results have been obtained in terms of In [8], the authors proposed rectangular DRA atcoefficient, 2.6 GHz, mean excited by coplanar waveguide a S-parameters, radiation pattern plots, correlation effective gain, and diversityand gain. coaxial probe and proposed obtained reasonable terms correlation coefficient and diversity gain. In [8], the authors rectangularresults DRA atin2.6 GHz,ofexcited by coplanar waveguide and a coaxial proposed in [5–8] are designed forcorrelation single-band MIMO operation. A lot of research probeAntennas and obtained reasonable results in terms of coefficient and diversity gain. studies have been presented on dual-polarized patch MIMO antennas but, their A metallic losses at Antennas proposed in [5–8] are designed microstrip for single-band operation. lot of research high frequencies substantial. A compact and multiband DRA forbut, mobile devices studies have beenbecome presented on dual-polarized microstrip patch antennas theirhandheld metallic losses at for DVB-H, Wi-Fi,become and WiMAX is presented in [9]and using two dielectric However, devices there is high frequencies substantial. A compact multiband DRA resonators. for mobile handheld large ground sizeand andWiMAX high mutual coupling especially the higherHowever, band. Another for DVB-H, Wi-Fi, is presented in [9]values using ,two dielectricatresonators. there is dual-band DRA characteristics in literature is in cylindrical covering DCS large ground sizewith and MIMO high mutual couplingreported values, especially at the higher band. form Another dual-band and bandscharacteristics [10]. This design has good isolation but with large values of ground plane and DRAWLAN with MIMO reported in literature is in cylindrical form covering DCS and WLAN DRA bandsheight. [10]. This design has good isolation but with large values of ground plane and DRA height. In DRA with dual-band characteristics is reported using a In this this paper, a MIMO rectangular DRA single radiator to operate at 1.8 and and 2.6 GHz GHz bands for for LTE LTE applications. applications. A rectangular shape is chosen because it has one more more degree degree of of freedom freedom than cylindrical and two degrees degrees of of freedom freedom than than hemispherical DRAs. Lower frequency frequency band bandof ofthe t heproposed proposeddesign designmay maybebetermed termedasaswideband wideband asas it it covers three LTE bands and 9 simultaneously. The simulation presented MIMO antenna covers three LTE bands 2, 2, 3, 3, and 9 simultaneously. The simulation of of thethe presented MIMO antenna is is performed with the help of Ansoft 16.0 and aagreement good agreement found the between the performed with the help of Ansoft HFSSHFSS v 16.0vand a good is found is between simulated simulated and results. measured results. and measured The remainder remainder of this this paper paper covers antenna geometry description in Section 2. Antenna Antenna design design analysis including including effect effect of of electric electricfield fieldon onmutual mutualcoupling couplingwill willbebepresented presented Section Section 4 inin Section 3. 3. Section 4 is is dedicated discuss results detail and Section 5 concludes paper. dedicated to to discuss results in in detail and Section 5 concludes thethe paper.

2. Geometry Geometry of of the the Presented Presented Antenna Antenna 2. The geometry proposed design is shown in Figure It consists an L-shaped geometryofofthethe proposed design is shown in 1. Figure 1. It ofconsists of anrectangular L-shaped dielectric resonator excited by excited two symmetrical slots of equal coupled coupled through rectangular dielectric(DR) resonator (DR) by two symmetrical slotsdimensions of equal dimensions microstrip feed lines. DR is made up of ceramic material (ε = 10) and has a loss tanδ = 0.002, is r through microstrip feed lines. DR is made up of ceramic material (εr = 10) and has a loss tanδ =and 0.002, 2 . FR4 substrate 2 placed on a ground plane of size 100 × 100 mm (ε = 4.6) of the size of the ground plane r and is placed on a ground plane of size 100 × 100 mm . FR4 substrate (εr = 4.6) of the size of the with height 1.6with mm is the base of the design. Figure 1a shows air-gapaofcylindrical 22 mm in ground plane height 1.6 layer mm is base layer of the design.a cylindrical Figure 1a shows 2 ) at the height with a hole of 6 mm radius drilled in the DR along with two metallic strips (6 × 20 mm air-gap of 22 mm in height with a hole of 6 mm radius drilled in the DR along with two metallic optimized position tothe improve isolation. A top view of the design A is top shown inof Figure 1b, in is which pl strips (6 × 20 mm 2) at optimized position to improve isolation. view the design shown and pw is1b, theinlength width the coupling slots, + s2 + s3 is the stub length in Figure whichand pl and pw of is both the len gth and width of whereas both the s1 coupling slots, whereas s1 + s2to+ improve impedance andimpedance m_w is thematching width of the feed linesof(placed at the bottom s3 is the stub lengthmatching to improve andmicrostrip 𝑚_𝑤 is the width the microstrip feed surface of theatsubstrate). Both the coupling slots have thethe dimensions × have pl etched on the ground lines (placed the bottom surface of the substrate). Both coupling pw slots the dimensions pw plane and are fed by two 50 Ω microstrip feed lines. Table 1 lists the optimized dimensions of the × pl etched on the ground plane and are fed by two 50 Ω microstrip feed lines. Table 1 lists proposed design after aofthorough parametric optimized dimensions the proposed designanalysis. after a thorough parametric analysis.

(a)

(b)

Figure se nte d ante nna Ge ome try(a) (a)3-D 3-Dview; vie w;(b) (b)Top Topview. vie w. Figure1.1.Pre Presented antenna Geometry

Sensors 2017, 17, 148

3 of 15

Table 1. Presented design optimized dimensions. Parameter

Value (mm)

Parameter

Value (mm)

Parameter

Value (mm)

a sl pw s3

20 20 3 4.5

b sw s1 mw

25 6 5.25 3

h pl s2 rad

22 23 3 6

3. Antenna Design and Analysis This section presents the details about the dimension calculation of the DR and the parametric study of the design along with the effects of the cylindrical air-gap and metal strips on the mutual coupling between the ports. 3.1. Initial DR Dimensions and Parametric Study Initial values of the DRA dimensions are calculated using dual-band formulas given in [10]. While calculating the dimensions of the proposed dual-band DRA, all the equations given in [10] for dual band DRAs were used. Only the equations for kx1 and kx2 for TEy 111 and TEy 211 and ky1 and ky2 for TEx 111 and TEx 121 modes were modified because the excited higher order modes in the presented DRA are TEy 111 and TEy 211 , TEx 111 and TEx 121 as compared to the higher order mode excited in [10] which is TEy 113 . The modified values are given in Equations (1) and (2). Kx1 =

π 2π π π , Kx2 = and Kz1 = , Kz2 = a a 2h 2h

(1)

Ky1 =

π 2π π π , Ky2 = and Kz1 = , Kz2 = b b 2h 2h

(2)

where kx, ky, and kz are the wave numbers in the x, y, and z direction, respectively. The design formulas given in [10] are applicable to a rectangular shaped DRA only. Therefore, based on these formulas, the dimensions of the dual-band rectangular DRA were obtained. The DRA was simulated and corresponding S-parameters are shown in Figure 2a. S-parameters from port 2 are not shown as they are similar to port 1. It is clear from this figure that there are three resonances at 1.68 GHz, 2.33 GHz, and 2.54 GHz. The first resonance relates to TE111 mode with poor matching while the second resonance is due to some unknown higher order mode at 2.33 GHz with low simulated gain (3.8 dBi). The third resonance is because of the TE121 mode. In order to improve matching and shift the resonance frequencies to the required bands of interest, the size of the DRA needs to be changed. As the modes excited are TEx and TEy , the resonance frequency of the excited modes is sensitive to the dimension of the DRA along x and y-axis. Also, to maintain symmetry between the ports along with change in size in the x and y-axis, the rectangular shape was modified to an L-shaped DRA. Doing so, the size of the DRA has been modified in the x and y-axis (dimension a and b) which in turn changes the wave numbers of kx and ky, thus shifting the lower resonance frequency towards the higher band while merging the two resonances at the upper band to give a wide band. The S-parameters of the L-shaped DRA are shown in Figure 2b. As evident from the figure, the resonance at the lowest frequency has shifted to 1.8 GHz while at the upper frequency band, the two resonances have merged to obtain a wideband and matching has been improved. However, isolation between the ports at the bands of interest is poor. To improve port isolation, cylindrical air-gap and metal strips have been incorporated in the L-shaped DRA and the results are shown in Figure 2c.

Sensors 2017, 17, 148 Sensors 2017, 17, 148

4 of 15 4 of 15

Sensors 2017, 17, 148

4 of 15

(a) Rectangular DRA

(a) Rectangular DRA

(b) L-Shaped DRA

(b) L-Shaped DRA

(c) L- Shaped DRA with air-gap and metal strips Figure S-Parame te rsand Plotsmetal . (c) L- Shaped DRA2.2. with air-gap strips Figure S-Parameters Plots.

A parametric study on cylindrical air-gap radius rad length of metallic strips sl is performed Figure 2. S-Parame te rsand Plots . to evaluate the performance of the design. Figure 3 shows the effect themetallic cylindrical air -gap A parametric study on cylindrical air-gap radius rad and lengthofof strips sl is radius performed rad on reflection coefficient of port 1 (S 11). As both ports 1 and 2 are symmetrical, only the results ofradius A parametric study on cylindrical air-gap radius rad and length of metallic strips sl performed to evaluate the performance of the design. Figure 3 shows the effect of the cylindrical is air-gap port 1 are shown in the parametric study.Figure As can3be observed from the figure that the radius of the to evaluate the performance of the design. shows the effect of the cylindrical air -gap radius rad on reflection coefficient of port 1 (S11 ). As both ports 1 and 2 are symmetrical, only the results cylindrical air-gap has aof little reflection coefficient. minor shift in frequency be of coefficient porteffect 1 (S 11on ). As both ports and 2Aare symmetrical, only thecan results of rad porton1 reflection are shown in the parametric study. As can be 1observed from the figure that the radius of observed in Figure 3 at both frequency bands. Similarly, the effect of the cylindrical air-gap radius onthe port 1 are shown in the parametric study. As can be observed from the figure that the radius of the cylindrical air-gap has a little effect on reflection coefficient. A minor shift in frequency can be transmission coefficient (S 12 oreffect S21 ) is on shown in Figure 4. It is clearly observable this figure that cylindrical air-gap has a little reflection coefficient. A minor shiftfrom in frequency can , be observed in Figure the 3 atradius both of frequency bands. Similarly, the effect of theconsiderably cylindrical air-gap radius on by increasing the air -gap, the isolation at 2.6 GHz improves and its value observed in Figure 3 at both frequency bands. Similarly, the effect of the cylindrical air-gap radius on transmission or rad S21 )=is6 shown 4. isolation It is clearly observable from this figure reaches coefficient almost to 35(S dB mm. At in 1.8Figure GHz, the decreases with the increase in rad that, 12 for transmission coefficient (S 12 or S21 ) is shown in Figure 4. It is clearly observable from this figure that , but at a very low rateof and a valuethe of 13 dB at radat = 62.6 mm. by increasing the radius thehas air-gap, isolation GHz improves considerably and its value

by increasing the radius of the air -gap, the isolation at 2.6 GHz improves considerably and its value

reaches almost to to 3535 dBdB forfor rad = 6= mm. AtAt 1.81.8 GHz, the isolation decreases the increase reaches almost rad 6 mm. GHz, the isolation decreaseswith with the increaseininrad radbut at abut very low rate and has a value of 13 dB at rad = 6 mm. at a very low rate and has a value of 13 dB at rad = 6 mm.

Figure 3. |S 11 | for diffe re nt value s of cylinde r radius.

Figure 3. |S 11 | for diffe re nt value s of cylinde r radius. Figure 3. |S11 | for different values of cylinder radius.

Sensors 2017, 17, 148

5 of 15

Sensors 2017, 17, 148

5 of 15

Sensors 2017, 17, 148

5 of 15

Sensors 2017, 17, 148

5 of 15

Figure 4. |S 12 | for diffe re nt value s of cylinde r radius. Figure 4. |S12 | for different values of cylinder radius. Figuresheet 4. |S 12strip | for length diffe re ntslvalue s ofreflection cylinde r radius. The effect of the metallic on the and transmission coefficients is Figure 4. |S 12 | for diffe re nt value s of cylinde r radius. The effect of the metallic sheet strip length sl on the reflection and transmission shown in Figures 5 and 6, respectively. From Figure 5, it is quite apparent that the lengthcoefficients of the strip is The effect of the metallic sheet strip length sl on the reflection and transmission coefficients is shown in Figureseffect 5 and respectively. From Figure 5, itthe is quite apparent that the length ofeffect the strip has negligible on6,the reflection coefficient at both frequency bands. A considerable of The of 5the metallic sheet strip length sl on and transmission coefficients is shown ineffect Figures and 6, respectively. From Figure 5, the it is reflection quite apparent that the length of the strip can be observed of coefficient the ports asatshown in Figure 6. Thebands. isolation 1.8 GHz for effect sl = hasslnegligible effect on on the theisolation reflection both the frequency A at considerable shown in Figures 5 and 6, respectively. From Figure 5, itthe is frequency quite apparent that length of the strip has negligible effect on the reflection coefficient at both bands. A the considerable effect of 10can andbe15observed mm is almost 12isolation and 13 dB while the isolation improves to6.19 dBisolation when theatstrip lengthfor of sl on the of the ports as shown in Figure The 1.8 GHz has negligible effectononthe theisolation reflection both the bands. A considerable sl can be observed of coefficient the ports asatshown in frequency Figure 6. The isolation at 1.8 GHzeffect for slof = tomm 20 mm. At the122.6 GHz band, the the length of the improves strip shiftstoisolation depththe towards the sl =increases 10can andbe15 15 almost and 13of dB while isolation 19 when strip length sl observed on the12 isolation the portsthe as isolation shown inimproves Figure 6. The atthe 1.8 strip GHz for sl = 10 and mm isis almost and 13 dB while to 19isolation dBdB when length lower frequency giving a value of isolation of 38 dB atof 2.6the GHz for shifts a value of sl = 20depth mm. towards the increases mm. AtAt the 2.6 GHz the length strip isolation 10 andto1520 is almost 12 and 13 band, dB while isolation improves to 19 dB when the towards strip length increases tomm 20 mm. the 2.6 GHz band, thethe length of the strip shifts isolation depth the lower frequency a value dB 2.6 GHz forshifts value 20 mm. increases to 20giving mm. At 2.6ofGHz band,of the strip isolation depth towards the lower frequency giving athe value ofisolation isolation of38 38length dB at at of 2.6the GHz for aavalue ofofslsl= = 20 mm. lower frequency giving a value of isolation of 38 dB at 2.6 GHz for a value of sl = 20 mm.

Figure 5. |S 11 | for diffe re nt value of she e t he ight. Figure 5. |S 11 | for diffe re nt value of she e t he ight. Figure 5. |S 11 | for diffe re nt value of she e t he ight.

Figure 5. |S11 | for different value of sheet height.

Figure 6. |S 12 | for diffe re nt value of she e t he ight. Figure 6. |S 12 | for diffe re nt value of she e t he ight. Figure 6. |S 12 | for diffe re nt value of she e t he ight. Figure 6. |S12 | for different value of sheet height.

Sensors 2017, 17, 148 Sensors 2017, 17, 148

6 of 15 6 of 15

3.2. Electric Field Effects in DR

3.2. Electric Field Effects in DR of electric field in the DR, it clearly demonstrates the reason of mutual If we analyze the effects coupling by effects using the cylindrical and itthe metaldemonstrates strips. The the electric fields If wereduction analyze the of electric field air in -gap the DR, clearly reason of originating from both the ports interact with each other and are a cause of mutual coupling in the mutual coupling reduction by using the cylindrical air-gap and the metal strips. The electric fields DR. This interaction E-fields reduced cylindrical air-gap and in a the pairDR. of originating from both between the portsthe interact withiseach otherby andusing are a acause of mutual coupling metal strips. At the field intersection point, a cylindrical air-gap is introduced which changes the This interaction between the E-fields is reduced by using a cylindrical air-gap and a pair of metal strips. corresponding phases ofpoint, the E-field at the boundaries of the air-gap. When the field enters from a At the field intersection a cylindrical air-gap is introduced which changes the corresponding higher primitively material cylindrical air-gap, it deviates away from the anormal phases to of lower the E-field at the boundaries of the air-gap. When the field enters from higher direction to lower [11], thus reducing the interaction between the E-field from the two ports. This effect seen at primitively material cylindrical air-gap, it deviates away from the normal direction [11],can thusbereducing the where field eachthe port is tangentially oriented in the a considerable the boundaries interaction between the through E-field from two ports. This effect canresulting be seen at boundaries decrease in mutual coupling. In order to further reduce the coupling effects, two metallic strips are where field through each port is tangentially oriented resulting in a considerable decrease in mutual introduced at the corner of the DR shown in Figure 1a. It is well known that the tangential coupling. In order to further reduce the coupling effects, two metallic strips are introduced at the components of electric field vanish strikesthat on the a metallic surface, only the transverse corner of the DR shown in Figure 1a. Itwhen is wellitknown tangential components of electric field component can exist. This results in further deviation of the E-filed originating from each portresults from vanish when it strikes on a metallic surface, only the transverse component can exist. This each otherdeviation resulting in further reduction in mutual coupling. Air-gaps [12] and metallic are in further of athe E-filed originating from each port from each other resulting in strips a further being usedintomutual increase bandwidth and impedance matching, . However, in this design reduction coupling. Air-gaps [12] and metallic stripsrespectively are being used to increase bandwidth this combination has been used as a coupling reduction technique. The magnitude E-field plot and impedance matching, respectively. However, in this design this combination has been used asisa shown in reduction Figure 7. technique. The magnitude E-field plot is shown in Figure 7. coupling

(a) Without air-gap and metal strips

(b) With air-gap and metal strips

Figure 7. E-field Magnitude plot.

Figure 7. E-field Magnitude plot.

4. Antenna Performance Measurements 4. Antenna Performance Measurements This section includes the antenna and MIMO performance parameters. First, the antenna This section includes the antenna and MIMO performance parameters. First, the antenna performance parameters such as S-parameters, total efficiency, and radiation patterns are described; performance parameters such as S-parameters, total efficiency, and radiation patterns are described; followed by diversity and MIMO performance matrices like ECC, MEG, DG , and channel capacity followed by diversity and MIMO performance matrices like ECC, MEG, DG, and channel capacity (throughput) calculation. (throughput) calculation. 4.1. 4.1. Antenna Antenna Parameters Parameters Analysis Analysis For and 2.6 GHz, two orthogonal modes For the the antenna antenna to to be be resonating resonating at at 1.8 1.8 and 2.6 GHz, two orthogonal modes have have to to be be excited excited x x simultaneously. Port 1 excites dominant TE δ11 and high er order TEx δ21 modes whereas port 2 excites x simultaneously. Port 1 excites dominant TE δ11 and higher order TE δ21 modes whereas port 2 excites y TE and TEy2δ1 y modes, respectively. Because of the symmetry of both ports, the E-field distribution TEy1δ1 1δ1 and TE 2δ1 modes, respectively. Because of the symmetry of both ports, the E-field distribution for both modes is shown shown for both modes is is shown shown in in Figure Figure 88 for for port port 11 only. only. The The prototype prototype of of the the proposed proposed antenna antenna is in in Figure Figure 9. 9. The The ground ground plane plane is is etched etched on on FR4 FR4 substrate substrate and and the the DR DR is is made made up up of of Eccostock Eccostock HK-10 HK-10 material with tanδ = 0.002. Microstrip feed lines are placed at the opposite side of material with tanδ = 0.002. Microstrip feed lines are placed at the opposite side of the the substrate substrate to to feed DR through two slots of equal dimensions. Figure 10 demonstrates the comparison of both the feed DR through two slots of equal dimensions. Figure 10 demonstrates the comparison of both the simulated S-parametersfor for reference reference level level − −10 simulated and and measured measured S-parameters. S-parameters. S-parameters 10dB dBare aremeasured. measured.

Sensors 2017, 17, 148

7 of 15

Sensors 2017, 17, 148 Sensors Sensors2017, 2017,17, 17,148 148

7 of 15 7 7ofof1515

(a)

(b)

Figure 8. E-fie ld Distribution for (a) TEx δ1 (b) TEx δ21 .

(a) (a)

(b) (b)

x δ1 x δ21 xx δ1(b) x. . . x δ21 Figure E-fie Distribution TE Figure E-fieldldDistribution Distributionfor for(a) (a)TE TE TE Figure 8.8.8.E-field for (a) TE (b) TE δ1 (b) δ21

(a)

(b)

Figure 9. Prototype of the propose d MIMO ante nna (a) 3-D vie w (b) Bottom vie w.

Measured fractional bandwidth at 1.8 GHz is 18% (1.71–2.05 GHz) and at 2.6 GHz is 8% (2.5–2.7 GHz) through ports 1 and 2, respectively. Figure 10 indicates a good agreement between measured (a) (b) (a) (b) and simulated results. The antenna is well matched at both resonances , covering four LTE bands as Figure9.9.Prototype Prototypeofofthe thepropose proposeddMIMO MIMOante antenna nna(a) (a)3-D 3-Dvie vieww(b) (b)Bottom Bottomvie view.w. mentionedFigure earlier. Figure 9. Prototype of the proposed MIMO antenna (a) 3-D view (b) Bottom view. Measured Measuredfractional fractionalbandwidth bandwidthatat1.8 1.8GHz GHzisis18% 18%(1.71–2.05 (1.71–2.05GHz) GHz)and andatat2.6 2.6GHz GHzisis8% 8%(2.5–2.7 (2.5–2.7 GHz) GHz)through throughports ports11and and2,2,respectively. respectively.Figure Figure10 10indicates indicatesaagood goodagreement agreementbetween betweenmeasured measured and andsimulated simulatedresults. results.The Theantenna antennaisiswell wellmatched matchedatatboth bothresonances resonances, covering , coveringfour fourLTE LTEbands bandsasas mentioned earlier. mentioned earlier.

Figure 10. Measured and simulated S-parameters. Figure 10. Me asure d and simulate d S-parame te rs.

Measured fractional bandwidth at 1.8 GHz is 18% (1.71–2.05 GHz) and at 2.6 GHz is 8% (2.5–2.7 GHz) through ports 1 and 2, respectively. Figure 10 indicates a good agreement between measured and simulated results. The antenna is well matched at both resonances, covering four LTE Figure 10. Me asure d and simulate d S-parame te rs. bands as mentioned earlier. Figure 10. Me asure d and simulate d S-parame te rs. Simulated total radiation efficiencies through ports 1 and 2 are the same i.e., 94.42% at 1.8 GHz and 96% at 2.6 GHz, respectively. These values are calculated using Equations (3) and (4) [13].

2017, 17, 148 SensorsSensors 2017, 17, 148

8 of 158 of 15

SensorsSimulated 2017, 17, 148 total radiation efficiencies through ports 1 and 2 are the same i.e., 94.42% at 1.88 GHz of 15 2 Equations (3) and (4) [13]. 2 and 96% at 2.6 GHz, respectively. These values are calculated using µ1,tot = µ1,rad (1 − |s11 | − |s21 | )

(3)

µ2,tot = µ2,rad (1 − |s22 | − |s12 | )

(4)

Simulated total radiation efficiencies through ports 12and 2 are the same i.e., 94.42% at 1.8 GHz (3) 𝜇 1,𝑡𝑜𝑡 = 𝜇 1,𝑟𝑎𝑑 (1 − |𝑠11 | − | 𝑠21 |2 ) 2 and 96% at 2.6 GHz, respectively. These values are calculated using2 Equations (3) and (4) [13]. ( | 1122|2|2−−| 𝑠|𝑠2112|2|2)) = 𝜇𝜇1,𝑟𝑎𝑑 𝜇𝜇1,𝑡𝑜𝑡 2,𝑡𝑜𝑡= 2,𝑟𝑎𝑑(11−−|𝑠𝑠

(4) (3)

Figure 11 shows the simulated and measured gain and efficiency plots of the proposed MIMO 11 shows and measured gain the and efficiency plots of the proposed MIMO DRA. TheFigure measured gain the andsimulated efficiency only seen that (1 |2 − bands |𝑠 |2 ) of interest. It can be easily(4) 𝜇 2,𝑡𝑜𝑡is=shown 𝜇 − | 𝑠at 22 DRA. Th e measured gain and efficiency is 2,𝑟𝑎𝑑 shown only at the 12 bands of interest. It can be easily seen the measured gain at 1.8 GHz for ports 1 and 2 is 5.5 dBi and 5.2 dBi, respectively. While at 2.6 GHz, 11 shows theatsimulated andports measured anddBi efficiency of the proposed MIMO thatFigure the measured gain 1.8 GHz for 1 and gain 2 is 5.5 and 5.2 plots dBi, respectively. While at 2.6 the measured gain for ports 1 and 2 is 5.4 dBi and 5.3 dBi, respectively. Similarly, the measured DRA. measured gain gainfor and efficiency bands of interest. It can be seen GHz,Th thee measured ports 1 and 2isisshown 5.4 dBionly andat 5.3the dBi, respectively. Similarly, theeasily measured efficiency atmeasured 1.8 GHz for ports and 2 2isports 90%1and 89% while at2.6 2.6 GHz for ports 1 and 2 is that the gainfor atports 1.81GHz for 2 is 5.5 dBi at and 5.2 dBi, at93% 2.6 and efficiency at 1.8 GHz 1 and is 90% and 89% while GHz forrespectively. ports 1 and While 2 is 93% and 91%,GHz, respectively. measured gain for ports 1 and 2 is 5.4 dBi and 5.3 dBi, respectively. Similarly, the measured 91%,the respectively. efficiency at 1.8 GHz for ports 1 and 2 is 90% and 89% while at 2.6 GHz for ports 1 and 2 is 93% and 91%, respectively.

(a)

(b)

Figure 11. Simulate d and me asure d (a) Gain; (b) Efficie ncy.

(a) 11. Simulated and measured (a) Gain; (b) Efficiency. (b) Figure Figure 12a–d shows E-plane views dand 12e–h the H-plane Figure 11. Simulate andFigure me asure d (a)shows Gain; (b) Efficie ncy. views of the measured and simulated radiation patterns at 1.8 and 2.6 GHz through port s 1 and 2, respectively. Radiation Figure 12a–d shows E-plane views and Figure 12e–h shows the H-plane views of the measured Figure 12a–d shows E-plane views and Figure 12e–h shows the H-plane views of the measured patterns shown in Figure 12a,batpoint in broadside for both ports 1 and 2 at 1.8 Radiation GHz. and simulated radiation patterns 1.8 and 2.6 GHz directions through ports 1 and 2, respectively. and simulated patterns at 1.8 and 2.6 12 GHz s and 1 and 2, respectively. Radiation Similarly, radiation patterns inin Figure c,dthrough point in port +45° 315° for ports 1 and 2 at1.8 2.6GHz. patterns shown inradiation Figure 12a,bshown point broadside directions for both ports 1 and 2 at patterns shown in Figure 12a,b point in broadside for both 1 and 2 at 1.8 GHz. GHz, respectively. The cross polarization values in directions all the plots are wellports below the co polarization ◦ ◦ Similarly, radiation patterns shown in Figure 12 c,d point in +45 and 315 for ports 1 and 2 at 2.6 GHz, Similarly, radiation in Figure 12 c,d point +45° and 315° for ports 1 anddiversity 2 at 2.6 values. The same ispatterns true forshown the H-plane patterns. These in results clearly indicate pattern respectively. The crossThe polarization valuesvalues in all the plots are well below the the co polarization values. GHz, respectively. cross polarization in all the plots are well below co polarization operation of the presented MIMO antenna. The 3 - D radiation pattern plots of the proposed The same isThe true for the H-plane patterns. These results clearly indicate pattern diversity operation values. same is true forare the H-planein patterns. These results clearlyare indicate dual-band MIMO antenna presented Figure 13. The 3D patterns shown pattern for bothdiversity the ports of the presented MIMO antenna. The 3-D radiation pattern plots of the proposed dual-band MIMO operation the2.6presented MIMO antenna. The 3 - D radiation pattern plots of the proposed at 1.8 GHzofand GHz, respectively. antenna are presented in Figure 13. The 3-D patterns are shown for both the ports at 1.8 GHz dual-band MIMO antenna are presented in Figure 13. The 3- D patterns are shown for both the ports and at 1.8 respectively. GHz and 2.6 GHz, respectively. 2.6 GHz,

(a)

(b)

(a)

(b)

Figure 12. Cont.

Sensors 2017, 17, 148

9 of 15

Sensors 2017, 17, 148

9 of 15

(c)

(d)

(e)

(f)

(g)

(h)

Figure 12. Measure d and simulate d radiation patterns. (a) E-Plane Port 1 at 1.8 GHz; (b) E-Plane Port

Figure 12. Measured and simulated radiation patterns. (a) E-Plane Port 1 at 1.8 GHz; (b) E-Plane Port 2 2 at 1.8 GHz; (c) E-Plane Port 1 at 2.6 GHz; (d) E-Plane Port 2 at 2.6 GHz; (e) H-Plane Port 1 at 1.8 at 1.8GHz; GHz;(f)(c) E-Plane Port 1 at 2.6 GHz; (d) E-Plane Port 2 at 2.6 GHz; (e) H-Plane Port 1 at 1.8 GHz; H-Plane Port 2 at 1.8 GHz; (g) H-Plane Port 1 at 2.6 GHz; (h) H-Plane Port 2 at 2.6 GHz. (f) H-Plane Port 2 at 1.8 GHz; (g) H-Plane Port 1 at 2.6 GHz; (h) H-Plane Port 2 at 2.6 GHz.

Sensors 2017, 17, 148

10 of 15

Sensors 2017, 17, 148

10 of 15

Sensors 2017, 17, 148

10 of 15

Figure 13. 3-D Simulate d Radiation Patterns. (a) Port 1 at 1.8 GHz; (b) Port 2 at 1.8 GHz; (c) Port 1 at

Figure 13. 3-D Simulated Radiation Patterns. (a) Port 1 at 1.8 GHz; (b) Port 2 at 1.8 GHz; (c) Port 1 at 2.6 GHz; (d) Port 2 at 2.6 GHz. 2.6 GHz; (d) Port 2 at 2.6 GHz. Figure 13. 3-D Simulate d Radiation Patterns. (a) Port 1 at 1.8 GHz; (b) Port 2 at 1.8 GHz; (c) Port 1 at 4.2.2.6 Diversity GHz; (d)Performance Port 2 at 2.6Analysis GHz.

4.2. Diversity Performance Analysis

Envelope correlation coefficient (ECC) is the first diversity and MIMO parameter to be

4.2.Envelope Diversity Itcorrelation Performance Analysisbetween coefficient (ECC) is theatfirst and In MIMO parameter be analyzed. analyzed. gives a correlation signals the diversity receiving end. this work, ECC is to calculated byEnvelope bothcorrelation S-parameter using Equation (5) D radiation using (6) [14]. It gives ausing correlation between signals at(ECC) the receiving end. In thispattern work, isEquation calculated by coefficient is and the 3first diversity and ECC MIMO parameter to using be Simulated and measured ECC using S-parameters Equation (5) are shown in Figure 14. both S-parameter using Equationbetween (5) and signals 3D radiation using (6) [14]. and analyzed. It gives a correlation at the pattern receiving end. Equation In this work, ECC Simulated is calculated ∗ are shown ∗ 2 Figure 14. | measured S-parameters in 𝑆(5) 𝑆 + 𝑆 𝑆 by usingECC bothusing S-parameter using Equation Equation |(5) and 3 D radiation pattern using Equation (6) [14]. 11 12 21 22 𝜌 =

𝑒 |2 are |2 )) in Figure 14. Simulated and measured ECC using S-parameters (1 − (| )(1 − (|𝑆22 𝑆11 |2 + |𝑆21 | 2 )Equation + |𝑆shown 12 (5)

S∗ ∗ S12 + S∗∗ S22 22 |𝑆1111 𝑆12 + 𝑆21 | 21 𝑆22 ρe = 2 →∗ ( 2 ) →∗ ( 2 ) |2 + 𝜌𝑒 =(1 − (|S |ʃʃ|24𝜋2+ | 𝐹 𝜃, 𝜑 . 𝐹 𝜃, 𝜑 1 2 S ))( 1 − (| S | | 2 ))(1 − (|𝑆22𝑑Ω 2 + ||𝑆S12||2 ))) 11 21 (| | | | | ( ) 1 − 𝑆 + 𝑆 𝜌 = 𝑒

11

21 2

22

2

12

∗ ∗ ʃʃ 4𝜋 |𝐹1→ (𝜃, 𝜑) | 𝑑Ωʃʃ4𝜋 |𝐹2→ (𝜃, 𝜑) | 𝑑Ω

(5)

(6)

(5)

(5)

2 2

∗ →∗ ∗ ( θ, ϕ → ∗ ( θ, ϕ)) . F→ F→ )dΩ SS|ʃʃ4π | 𝜃, 𝜑 . 𝐹22 (𝜃, 𝜑) 𝑑Ω 4𝜋 𝐹11 ρe =𝜌𝑒 = →∗ ∗ 22 → ∗ ∗ 2 2 (𝜃, 𝜑) | 𝑑Ω F|→ |𝐹 2F → ʃʃ4𝜋 𝐹 (𝜃, 𝜑) | 𝑑Ωʃʃ4𝜋4π SS4π 2 ( θ, ϕ ) dΩ 1 1 ( θ, ϕ ) dΩSS

Figure 14. Me asure d and simulate d ECC.

Figure 14. asure d and simulate d ECC. Figure 14. Me Measured simulated ECC.

(6) (6)

Sensors 2017, 17, 148

11 of 15

ECC using far field pattern Equation (6) is found to be 0.0218 and 0.0165 at 1.8 and 2.6 GHz, Sensors and 2017, 17, 11 of 15 respectively is 148 shown as red dots in Figure 14. This figure also shows that ECC values are well below 0.5 in the desired frequency bands, which ensures good diversity performance [15]. ECC using far field pattern Equation (6) is found to be 0.0218 and 0.0165 at 1.8 and 2.6 GHz, Diversity gainand (DG) is another import parameter which good respectively is shown as red dots in Figure 14. This figure alsoassures shows that ECCdiversity values areand well MIMO performance. In this work, DG is calculated using Equation (7) [8]. Sensors 2017, 17, 148 11 of 15 below 0.5 in the desired frequency bands, which ensures good diversity performance [15]. Diversity gain (DG) is another import parameter which assures good diversity and MIMO ECC using pattern is = found performance. In far thisfield work, DG is Equation calculated(6) using Equation [8]. and 0.0165 at 1.8 and 2.6 GHz, DG 10eρto be(7)0.0218 (7) respectively and is shown as red dots in Figure 14. This figure also shows that ECC values are well 𝐷𝐺 ensures = 10𝑒𝜌 good diversity performance [15]. (7) below 0.5 in the desired frequency bands, which where qparameter which assures good diversity and MIMO Diversity gain (DG) is another import where 2 eρ = using (1 −Equation ) [8]. (8) |0.99ρe |(7) performance. In this work, DG is calculated 𝑒𝜌 = √(1 − |0.99𝜌𝑒 |2 ) (8) 𝐷𝐺 = diversity 10𝑒𝜌 (7) MIMO Figure 15 shows the measured and simulated gain of the proposed dual-band Figure 15 shows the measured and simulated diversity gain of the proposed dual -band MIMO antenna.where The diversity gain in the both the frequency bands is almost 10 dB. antenna. The diversity gain in the both the frequency bands is almost 10 dB. 𝑒𝜌 = √(1 − |0.99𝜌𝑒 |2 )

(8)

Figure 15 shows the measured and simulated diversity gain of the proposed dual -band MIMO antenna. The diversity gain in the both the frequency bands is almost 10 dB.

Figure 15. Me asure d and simulate d dive rsity gain.

Figure 15. Measured and simulated diversity gain. Mean effective gain (MEG) is the last and very important parameter to be analyzed for diversity performance. It is defined as the ratio of the average power received at the antenna to the sum of the Mean effective gain (MEG) is the last and very important parameter to be analyzed for diversity Figureand 15. Me asure d andpolarized simulate dwaves dive rsity gain. by isotropic antenna [16]. average power of the vertically horizontally received performance. It is defined as the ratio of the average power received at the antenna to the sum of the Using S-parameters MEG for each port, it can be measured using Equation (9) [17]. For similar Mean gain (MEG) is horizontally the last very important parameter to be analyzed for diversity averagepower power ofeffective the and polarized received by isotropic antenna [16]. levels of vertically each branch, power ratioand k, which is equalwaves to the difference in the magnitude of performance. It is defined as the ratio of the average power received at the antenna to the sum of the Using S-parameters MEGusing for each port,(10) it can MEGs, is calculated Equation [18].be measured using Equation (9) [17]. For similar power power of the vertically polarized waves received by isotropic antenna [16]. levels ofaverage each branch, power ratioand k, horizontally which is equal to 𝑀 the difference in the magnitude of MEGs, Using S-parameters MEG for each port, it can be measured 2using Equation (9) [17]. For similar is calculated using Equation (10)𝑀𝐸𝐺 [18]. = 0.5 [1 − ∑ |𝑆 | ] (9)of 𝑖 = 0.5𝜂 𝑖,𝑟𝑎𝑑 power levels of each branch, power ratio k, which is equal𝑖𝑗to the difference in the magnitude 𝑗=1 " # MEGs, is calculated using Equation (10) [18]. M

𝑀− MEG𝑘i ==|𝑀𝐸𝐺 0.5η1i,rad = 0.5 |Sij |2 | 1 − 𝑀𝐸𝐺 2 < 3𝑑𝐵 ∑

(10) 𝑀𝐸𝐺𝑖 = 0.5𝜂𝑖,𝑟𝑎𝑑 = 0.5 [1 − ∑ |𝑆j𝑖𝑗=| 12 ] (9) Figures 16 and 17 shows the simulated and measured MEGs and both the figures clearly show 𝑗=1 that k is almost equal to 0 at both the desired frequency bands.

k = | MEG1 − MEG2 | < 3dB

𝑘 = |𝑀𝐸𝐺1 − 𝑀𝐸𝐺2 | < 3𝑑𝐵

(9) (10)

(10)

Figures 16 and 17 shows the simulated and measured MEGs and both the figures clearly show Figures 16 and 17 shows the simulated and measured MEGs and both the figures clearly show that k isthat almost equalequal to 0 at thethe desired bands. k is almost to 0both at both desiredfrequency frequency bands.

Figure 16. Simulate d Me an Effe ctive Gain (MEG).

Figure 16. Simulate d Me an Effe ctive Gain (MEG).

Figure 16. Simulated Mean Effective Gain (MEG).

Sensors 2017, 17, 148

12 of 15 12 of 15

Sensors 2017, 17, 148

Figure 17. Me asure d Me an Effe ctive Gain (MEG). Figure 17. Measured Mean Effective Gain (MEG).

4.3. MIMO Performance Analysis 4.3. MIMO Performance Analysis This section presents the MIMO performance analysis of the proposed dual-band MIMO DRA. This section presents the MIMO performance analysis of the proposed dual-band MIMO DRA. MIMO performance in terms of data throughput has been measured. For the throughput MIMO performance in terms of data throughput has been measured. For the throughput measurements measurements of the proposed design, the setup shown in Figure 18 has been used. The procedure of the proposed design, the setup shown in Figure 18 has been used. The procedure presented in [19] presented in [19] has been used in this paper for the throughput measurements. Throughput, which has been used in this paper for the throughput measurements. Throughput, which is measured in is measured in bits per second (bps) in a communication network, is the amount of data transferred bits per second (bps) in a communication network, is the amount of data transferred per unit time per unit time through a communication link from source to destination. The a ggregate or system through a communication link from source to destination. The aggregate or system throughput throughput is the overall data rates transferred to all the destinations in a communi cation network. is the overall data rates transferred to all the destinations in a communication network. The term The term throughput is often used for maximum throughput. The maximum throughput (bits/sec or throughput is often used for maximum throughput. The maximum throughput (bits/sec or bps) of a bps) of a communication link or a node is also termed as its capacity. Figure 18 shows the MIMO communication link or a node is also termed as its capacity. Figure 18 shows the MIMO throughput throughput measurement setup. The figure shows the proposed dual-band MIMO antenna along measurement setup. The figure shows the proposed dual-band MIMO antenna along with two with two monopoles (separated by 0.5λ spacing). The separation between the proposed DR A and monopoles (separated by 0.5λ spacing). The separation between the proposed DRA and monopole is monopole is 30 cm. Both the antennas are connected to Wideband Radio Communication Tester 30 cm. Both the antennas are connected to Wideband Radio Communication Tester R&S® CMW500 R&S® CMW500 (LTE node B (eNodeB)) along with a Huawai E398 USB modem dongle (which acts as (LTE node B (eNodeB)) along with a Huawai E398 USB modem dongle (which acts as the LTE user the LTE us er equipment (UE)). By using dedicated Physical Downlink Shared Channel sub frames equipment (UE)). By using dedicated Physical Downlink Shared Channel sub frames the eNodeB the eNodeB (CMW500), the emulator transmits downlink data to the UE. The throughput is (CMW500), the emulator transmits downlink data to the UE. The throughput is calculated using the calculated using the Block Error Rate (BLER). The positive and negative acknowledgments returned Block Error Rate (BLER). The positive and negative acknowledgments returned by the UE are used to by the UE are used to determine BLER. The throughput measurements of the given MIMO antenna determine BLER. The throughput measurements of the given MIMO antenna were performed for a were performed for a SISO and 2 × 2 MIMO systems. Three types of modulations including QPSK, 16 SISO and 2 × 2 MIMO systems. Three types of modulations including QPSK, 16 QAM (quadrature QAM (quadrature amplitude modulation), and 64 QAM have been used to measure the throughput amplitude modulation), and 64 QAM have been used to measure the throughput for the SISO and for the SISO and MIMO system of the presented antenna. Tables 2 and 3 depict the measured MIMO system of the presented antenna. Tables 2 and 3 depict the measured throughput for the throughput for the SISO and MIMO antenna systems for the three modulation schemes along with SISO and MIMO antenna systems for the three modulation schemes along with the specified channel the specified channel quality indicator (CQI) at 1.8 GHz and 2.6 GHz, respectively. The parameters quality indicator (CQI) at 1.8 GHz and 2.6 GHz, respectively. The parameters used in the measurement used in the measurement setup include: bandwidth of 20 MHz, 100 number of resource blocks, −20 setup include: bandwidth of 20 MHz, 100 number of resource blocks, −20 dBm of transmitted power, dBm of transmitted power, frequency bands of 1.8 GHz (band 3), and 2.6 GHz (band 7). frequency bands of 1.8 GHz (band 3), and 2.6 GHz (band 7). From Tables 2 and 3, it can be clearly seen that the measured throughputs at both the frequency From Tables 2 and 3, it can be clearly seen that the measured throughputs at both the frequency bands are close to the maximum throughput. At 1.8 GHz for the SISO case, the proposed MIMO bands are close to the maximum throughput. At 1.8 GHz for the SISO case, the proposed MIMO DRA DRA achieves 99.9% of the theoretical maximum throughput for all modulation schemes. For the achieves 99.9% of the theoretical maximum throughput for all modulation schemes. For the case of case of MIMO at the same frequency, the presented antenna achieves 99.9% of the theoretical MIMO at the same frequency, the presented antenna achieves 99.9% of the theoretical maximum maximum throughput for all modulation schemes. Similarly, at 2.6 GHz for all the modulation throughput for all modulation schemes. Similarly, at 2.6 GHz for all the modulation schemes, schemes, the presented MIMO DRA achieves 99.9% of the maximum theoretical throughput for both the presented MIMO DRA achieves 99.9% of the maximum theoretical throughput for both SISO SISO and MIMO scenario. and MIMO scenario.

Sensors 2017, 17, 148

13 of 15

Sensors 2017, 17, 148

13 of 15

Figure 18. Channe l capacity me asure me nt se tup. Figure 18. Channel capacity measurement setup. Table 2. Me asure d throughput with thre e modulation sche me s at 1.8 GHz. Table 2. Measured throughput with three modulation schemes at 1.8 GHz. Measured Measured Modulation Scheme Maximum Maximum Modulation Measured Measured Maximum Maximum Throughput Throughput (CQI) Throughput SISO Scheme (CQI) Throughput Throughput Throughput SISO Throughput ThroughputMIMO MIMO SISOSISO MIMO MIMO QPSK(Mbps) (6) (6) 14.10 28.21 14.112 28.23 QPSK(Mbps) 14.10 28.21 14.112 28.23 16QAM (Mbps) (9) (9) 30.33 56.65 30.352 56.68 16QAM (Mbps) 30.33 56.65 30.352 56.68 64QAM (Mbps) (12)(12) 50.190 93.152 50.197 93.161 64QAM (Mbps) 50.190 93.152 50.197 93.161 Table3.3.Me Measured throughput with with thre threee modulation 2.62.6 GHz. Table asure d throughput modulationschemes sche me satat GHz.

Measured Modulation Measured Modulation Scheme Throughput Scheme (CQI) Throughput SISO (CQI) SISO QPSK(Mbps) (6) 14.11 QPSK(Mbps) (6) 14.11 16QAM (Mbps) (9) 30.350 16QAM (Mbps) 30.350 64QAM (Mbps) (12)(9) 50.191 64QAM (Mbps) (12) 50.191

Measured Measured Maximum Maximum Maximum Maximum Throughput Throughput MIMO Throughput Throughput SISO Throughput MIMO SISO Throughput MIMO MIMO 28.22 14.112 28.23 28.22 14.112 28.23 56.66 30.352 56.68 56.66 30.352 56.68 93.158 50.197 93.161 93.158 50.197 93.161

All the results analyzed in this work clearly indicate that the proposed antenna is suitable for All the results analyzed in this work clearly indicate that the proposed antenna is suitable for LTE applications and it is able to deliver an excellent throughput performance. The 2 × 2 MIMO LTE applications and it is able to deliver an excellent throughput performance. The 2 × 2 MIMO performance also justifies that the proposed antenna has a good level of isolation between ports to performance also justifies that the proposed antenna has a good level of isolation between ports to support the MIMO spatial multiplexing operation. support the MIMO spatial multiplexing operation. 5. Conclusions 5. Conclusions A dual-band dual-port MIMO RDRA fed by two symmetric coupling apertures is proposed A dual-band dual-port MIMO RDRA fed by two symmetric coupling apertures is proposed and and investigated. Two orthogonal radiating modes are excited in the DRA simultaneously at investigated. Two orthogonal radiating modes are excited in the DRA simultaneously at two two overlapping frequency ranges around 1.8 and 2.6 GHz. The measured isolation is better than overlapping frequency ranges around 1.8 and 2.6 GHz. The measured isolation is better than 1 7 dB 17 dB over the operating frequency ranges. Given the same frequencies and same dielectric material, over the operating frequency ranges. Given the same frequencies and same diel ectric material, the proposed antenna is the first with a rectangular shape to be operated at given frequencies with such

Sensors 2017, 17, 148

14 of 15

the proposed antenna is the first with a rectangular shape to be operated at given frequencies with such a high isolation values at both the bands for MIMO applications. ECC calculated from S-parameters and 3-D radiation pattern is well below 0.5 which is the required acceptable level. Measured DG is almost 10 dB and MEG ratios are maintained close to zero at both bands. The throughput measurement at both the frequency bands indicates superior performance to an SISO system in terms of more bits transferred per second. The presented results prove that the proposed antenna can provide considerable performance for LTE MIMO applications. Acknowledgments: The authors would like to thank the Ministry of Higher Education (MOHE) under FRGS (vote 4F283 and 4F733) and under Research University Grant (votes 11H59, 03G33, 05H62, and 04H36) and the Higher Center of Excellence Grant (vote 4J220 and 4J211). Also thanks to Universiti Teknologi Malaysia under vote number 4F818 and 12H35, the Ministry of Science and Innovation (MOSTI), under vote 4S076 for supporting this research work. Author Contributions: J.N. and C.Y.L. designed the experimental setup and performed the experiments; M.H.J. and O.O. conceived the idea and helped in finalizing the manuscript; M.R.K. and A.A.K. contributed in analyzing and writing the results. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3. 4.

5. 6. 7. 8. 9.

10. 11. 12. 13. 14. 15.

16.

Dahlman, E.; Parkvall, S.; Skold, J. 4 g: Lte/Lte-Advanced for Mobile Broadband; Academic Press: Cambridge, MA, USA, 2013. Ahsan, M.R.; Islam, M.T.; Ullah, M.H.; Singh, M.J.; Ali, M.T. Metasurface reflector (MSR) loading for high performance small microstrip antenna design. PLoS ONE 2015, 10, e0127185. [CrossRef] [PubMed] Ashoor, A.Z.; Ramahi, O.M. Dielectric resonator antenna arrays for microwave energy harvesting and far-field wireless power transfer. Prog. Electromagn. Res. C 2015, 59, 89–99. [CrossRef] Khan, A.A.; Jamaluddin, M.H.; Nasir, J.; Khan, R.; Aqeel, S.; Saleem, J. Design of a Dual-Band MIMO Dielectric Resonator Antenna with Pattern Diversity for WiMAX and WLAN applications. Prog. Electromagn. Res. M 2016, 50, 65–73. [CrossRef] Ishimiya, K.; Ying, Z.; Takada, J.-I. A compact mimo dra for 802.11n application. In Proceedings of the 2008 IEEE Antennas and Propagation Society International Symposium, Lund, Sweden, 5–11 July 2008; pp. 1–4. Yan, J.-B.; Bernhard, J.T. Design of a mimo dielectric resonator antenna for lte femtocell base stations. IEEE Trans. Antennas Propag. 2012, 60, 438–444. [CrossRef] Nasir, J.; Jamaluddin, M.H.; Khalily, M.; Kamarudin, M.R.; Ullah, I. Design of an mimo dielectric resonator antenna for 4g applications. Wirel. Pers. Commun. 2016, 88, 525–536. [CrossRef] Roslan, S.; Kamarudin, M.R.; Khalily, M.; Jamaluddin, M. An mimo rectangular dielectric resonator antenna for 4G applications. IEEE Antennas Wirel. Propag. Lett. 2014, 13, 321–324. [CrossRef] Huitema, L.; Koubeissi, M.; Mouhamadou, M.; Arnaud, E.; Decroze, C.; Monediere, T. Compact and multiband dielectric resonator antenna with pattern diversity for multistandard mobile handheld devices. IEEE Trans. Antennas Propag. 2011, 59, 4201–4208. [CrossRef] Sun, Y.X.; Leung, K.W. Dual-band and wideband dual-polarized cylindrical dielectric resonator antennas. IEEE Antennas Wirel. Propag. Lett. 2013, 12, 384–387. [CrossRef] Fang, X.; Leung, K. Designs of single-, dual-, wide-band rectangular dielectric resonator antennas. IEEE Trans. Antennas Propag. 2011, 59, 2409–2414. [CrossRef] Balanis, C.A. Antenna Theory: Analysis and Design; John Wiley & Sons: Hoboken, NJ, USA, 2009. Kishk, A.A.; Glisson, A. Effect of air gap on cylindrical dielectric resonator antenna operating in TM 01 mode. Electron. Lett. 1994, 30, 97–98. [CrossRef] Singh, H.S.; Meruva, B.R.; Pandey, G.K.; Bharti, P.K.; Meshram, M.K. Low mutual coupling between mimo antennas by using two folded shorting strips. Prog. Electromagn. Res. B 2013, 53, 205–221. [CrossRef] Nasir, J.; Jamaluddin, M.H.; Khalily, M.; Kamarudin, M.R.; Ullah, I.; Selvaraju, R. A reduced size dual port mimo dra with high isolation for 4g applications. Int. J. RF Microw. Comput. Aided Eng. 2015, 25, 495–501. [CrossRef] Lee, W.C. Mobile Communications Engineering; McGraw-Hill Professional: New York, NY, USA, 1982; p. 453.

Sensors 2017, 17, 148

17.

18. 19.

15 of 15

Alayon Glazunov, A. Theoretical analysis of mean effective gain of mobile terminal antennas in ricean channels. In Proceedings of the 2002 IEEE 56th Vehicular Technology Conference Proceedings, Vancouver, BC, Canada, 24–28 September 2002; pp. 1796–1800. Ding, Y.; Du, Z.; Gong, K.; Feng, Z. A novel dual-band printed diversity antenna for mobile terminals. IEEE Trans. Antennas Propag. 2007, 55, 2088–2096. [CrossRef] Moradikordalivand, A.; Rahman, T.A.; Leow, C.Y.; Ebrahimi, S. Dual-polarized MIMO antenna system for WiFi and LTE wireless access point applications. Int. J. Commun. Syst. 2014, 30, 1–11. [CrossRef] © 2017 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).