Shared radiator MIMO antenna for broadband applications
2018; Institution of Engineering and Technology; Volume: 12; Issue: 7 Linguagem: Inglês
10.1049/iet-map.2017.0331
ISSN1751-8733
AutoresSitu Rani Patre, Surya Pal Singh,
Tópico(s)Wireless Body Area Networks
ResumoIET Microwaves, Antennas & PropagationVolume 12, Issue 7 p. 1153-1159 Research ArticleFree Access Shared radiator MIMO antenna for broadband applications Situ Rani Patre, Situ Rani Patre Department of Electronics Engineering, Indian Institute of Technology (Banaras Hindu University), Varanasi, IndiaSearch for more papers by this authorSurya P. Singh, Corresponding Author Surya P. Singh spsingh.ece@iitbhu.ac.in Department of Electronics Engineering, Indian Institute of Technology (Banaras Hindu University), Varanasi, IndiaSearch for more papers by this author Situ Rani Patre, Situ Rani Patre Department of Electronics Engineering, Indian Institute of Technology (Banaras Hindu University), Varanasi, IndiaSearch for more papers by this authorSurya P. Singh, Corresponding Author Surya P. Singh spsingh.ece@iitbhu.ac.in Department of Electronics Engineering, Indian Institute of Technology (Banaras Hindu University), Varanasi, IndiaSearch for more papers by this author First published: 13 March 2018 https://doi.org/10.1049/iet-map.2017.0331Citations: 8AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Abstract In this study, a new compact broadband two-port multiple-input–multiple-output (MIMO) antenna using shared radiator is proposed. The radiator making an angle of 135° from X-axis is symmetrically shared by two tapered microstrip feed lines in orthogonal polarisations. The shared radiator MIMO antenna is more compact as compared with conventional MIMO antenna which uses separate radiator for each port. The proposed antenna is designed on low-cost FR4 substrate (dielectric constant = 4.4, and loss tangent = 0.02) of size 39 × 39 mm2 and provides a bandwidth of 136.63% (2.4–12.75 GHz) with reflection coefficients, S11/S22 less than or equal to −10 dB and values of mutual coupling between antenna ports S21/S12 less than or equal to −15 dB. Through the utilisation of end-loaded meandered line stub attached to modified curved ground plane, low mutual coupling between antenna ports is achieved. The simulated results for input and radiation characteristics of the proposed MIMO antenna are compared with corresponding experimental results. 1 Introduction The emerging wireless communication systems require compact broadband antennas which can provide high data rate. The ultrawideband (UWB) technology utilising UWB antennas gained importance due to its wide variety of applications for high-speed, short-range indoor wireless communication since the release of unlicensed UWB spectrum in the frequency range 3.1–10.6 GHz [[1]]. The high-speed indoor wireless communication performed in a dense environment typically suffers from multipath fading which degrades the signal quality. This issue can be resolved through the application of UWB and multiple-input–multiple-output (MIMO) technology concepts. The UWB antenna combined with MIMO technology can handle high data rate signals with minimal multipath fading [[2]], since MIMO technology employs multiple antennas which transmit/receive multiple spatial signals with different fading characteristics at the same time. It is very much unlikely that all received signals are influenced by deep fading. Thus, the system reliability in terms of minimal multipath fading can be enhanced through appropriate selection of received signals. One of the major hurdles in the design of MIMO antenna is to maintain high isolation between antenna ports over its operating bandwidth. In past few years, the design and development of compact MIMO antenna using planar UWB antennas have grown rapidly. In the early phase, planar monopole antennas (PMAs) operating in orthogonal polarisations were proposed for UWB-MIMO applications. The orthogonal polarisation configuration is achieved by placing multiple PMAs perpendicular to each other [[3]–[9]]. This type of perpendicular arrangement isolates antenna ports with/without the use of a stub or slot but increases the overall antenna size. To reduce overall size, the PMAs can be placed side-by-side instead of placing them in perpendicular arrangement. However, the side-by-side arrangement of PMAs decreases isolation between antenna ports. Therefore, extra isolation technique is required. Keeping this aspect in view, two UWB-MIMO antennas using half circular disk monopole and half square monopole antenna elements arranged in the side-by-side configuration are designed [[10]]. A vertical stub attached to ground plane is placed in between the antenna elements for maintaining proper isolation between antenna ports. In literature, different isolation techniques such as use of strips beneath the patches [[11]], parasitic decoupling elements [[12]], T-shaped ground stub along with slot [[13]], electromagnetic bandgap structure [[14]], comb-line structure [[15]] applied in between the PMAs and placed side-by-side to each other are also reported. Meanwhile, the quasi-self-complementary antennas arranged side-by-side without any extra isolation technique is also reported to form broadband MIMO antenna [[16], [17]]. Recently, common radiator operating in dual polarisation has been utilised as a two-port UWB-MIMO antenna [[18]]. The requirements for maintaining high isolation between antenna ports increases for compact MIMO antenna system utilising shared radiating patch. One of the solutions to obtain adequate isolation between antenna ports is the use of slot on shared radiator. In [[18]], high isolation is obtained through the combination of T-shaped slot on patch, a pair of slit on ground plane and a stub attached to ground plane. In [[19]], a compact UWB-MIMO antenna is proposed using a shared radiator fed by two asymmetrical-coplanar-strip lines and isolation is achieved with the help of an I-shaped slot in radiator, a stub attached to ground plane, a slot on ground plane and a parasitic strip on back of antenna substrate. Further, a circular patch is shared by two coplanar waveguide feeds and isolation is obtained through a single slot in patch [[20]], combination of slot in patch and stub attached to ground plane [[21]]. In [[21]], two band-rejections are also achieved using elliptical split ring slots. Furthermore in [[22]], isolation is maintained only by a cross-shaped slot inserted in a shared radiator, but antenna performance is not identical for both ports due to asymmetrical radiator configuration seen from two ports. In addition, most of the shared radiators are reported for UWB-MIMO applications and use slots in radiator patch for obtaining adequate isolation between antenna ports. In this paper, a broadband MIMO antenna using a shared radiator is proposed in which good isolation between antenna ports is achieved without introducing any slot in the patch. The proposed antenna comprises of a modified leaf-shaped radiating element shared symmetrically at 45° by two tapered microstrip lines and a common curved ground plane. Adequate isolation between antenna ports over its operating frequency band is obtained through the utilisation of end-loaded meandered line attached to modified curved ground plane. The proposed antenna is 39 × 39 mm2 in size and provides simulated −10 dB reflection coefficient bandwidth of 10.35 GHz (136.63%) over the frequency range 2.4–12.75 GHz with reasonable S21 ≤−15 dB over its operating frequency band. The designed antenna was fabricated and measurement of its input and radiation characteristics was made over the whole operating frequency range. The experimental results are compared with the respective simulation results obtained using Computer Simulation Technology Microwave Studio (CST MWS) software. 2 Antenna design Fig. 1a shows the geometrical configuration of proposed two-port MIMO antenna along with its dimensional parameters. The proposed antenna consists of single leaf-shaped radiating patch, which is shared by two tapered microstrip lines on one side of substrate (top-side) and an end-loaded meandered line stub attached to supporting curved ground plane on other side of substrate (bottom-side). The whole geometry of antenna is symmetrical about a line drawn at an angle of 45° from X-axis which results in identical performance of the antenna from either port. Fig. 1Open in figure viewerPowerPoint Geometries of (a) Proposed shared radiator MIMO antenna, (b) Different design stages of radiator, (c) Different stages of antenna The design of antenna starts with dual-feed circular disk monopole radiator having perimeter of approximately one-half of the free space wavelength at lower cutoff frequency of 2.2 GHz. The small portion of the basic radiator between the ports 1 and 2 (near the ground plane) is removed which helps in directing higher level of surface current from either port towards the upper boundary of radiator, thereby reducing mutual coupling between ports. However, this truncation of the basic radiator reduces the overall perimeter of radiator also to some extent. It is well known that the lower cutoff/first resonant frequency of the antenna is inversely proportional to the perimeter of radiator. The reduction in perimeter of the radiator increases the lower cutoff frequency of the antenna which is overcome by modifying the shape of the radiator. The shape of radiator was modified by superimposing a triangular patch on the diametrically opposite side of the truncated circular radiator. The modified shape of truncated circular radiator results in the proposed leaf-shaped radiator. The design stages involved in the evolution of radiator are shown in Fig. 1b. The length of outer boundary (excluding the truncated side) of proposed leaf-shaped radiator shown by dashed line in Fig. 1b is ∼λ/2 (where λ is the free space wavelength) at lower cutoff frequency of 2.2 GHz. The tapered microstrip line is utilised for proper impedance matching between 50 Ω SMA connector and the antenna. Two ports of the proposed MIMO antenna use single radiator instead of two radiators, thereby reducing overall size of the antenna as compared to conventional MIMO antenna system having two identical antenna elements placed in orthogonal or side-by-side arrangement. Due to the shared radiator and compact size of the proposed antenna in absence of ground plane stub, the mutual coupling between antenna ports increases. The high mutual coupling between ports can be reduced by utilising the basic idea of increasing the distance between antenna ports. For that reason, a metallic stub is attached to ground plane and length of the stub is increased by the use of meander line and pentagonal patch at the end of stub so that isolation can be increased for lower operating frequencies, since the operating frequency and length of active element is inversely proportional to each other. For improving the isolation between the ports further, the effective length of ground plane is also increased by creating slot in the ground plane where stub is attached, and by making the outer boundaries of ground plane curved. The configurations resulting from different stages of antenna design are shown in Fig. 1c. The optimised values of dimensional parameters of the proposed antenna are listed in Table 1. The proposed antenna is designed on cost-effective FR4 substrate having thickness of 0.8 mm and relative permittivity of 4.4. Table 1. Design parameters of the proposed antenna and corresponding dimensions in mm Parameter Value, mm Parameter Value, mm W 39 R1 13.3 Lf 2 s1 22.2 Lt 9.4 s2 7 Wf 1 w1 1.4 Wt 0.3 w2 11 Lg 10.4 w3 7.78 S_l 8.5 ml 3.8 S_w 11 mw 4.3 g 1 mg 1 3 Simulation and experimental studies 3.1 Simulation study The design and simulation of proposed shared radiator MIMO antenna were carried out using CST MWS software. The final configuration and optimum dimensions of the proposed antenna were found by optimising each parameter individually. The optimum parameter dimensions of the antenna are listed in Table 1. The antenna with optimum dimensions was then used to obtain its simulated input, radiation and diversity performance characteristics which are discussed in different sub-sections of results and discussion section. 3.2 Experimental study The proposed MIMO antenna with optimum dimensions was fabricated using T-Tech QC5000 Quick Circuit Prototyping Machine. The photographs of the fabricated antenna are shown in Fig. 2. The fabricated antenna was tested experimentally. The S-parameter–frequency characteristics of the antenna were measured using Anritsu make VNA Master Vector Network Analyser (model: MS2038C). The two-dimensional (2D) radiation patterns of the proposed MIMO antenna with port 1 excited and port 2 match terminated with 50 Ω load were measured in an anechoic chamber at discrete frequencies. For measuring the radiation patterns of the proposed antenna, 1–18 GHz broadband horn antenna was used as transmitting (Tx) antenna and the proposed antenna acting as receiving (Rx) antenna was placed on the antenna positioner system in the far-field region inside the anechoic chamber. The receiving (proposed) antenna was rotated along its own axis in H- and E-planes. The input port of transmitting horn and the output port of receiving antenna (the proposed antenna) to measure the magnitudes of scattering parameters (S21′) at different angles for the given plane and frequency of interest were connected, respectively, to ports 1 and 2 of Agilent make ENA series Vector Network Analyser (model: E5071C). Fig. 2Open in figure viewerPowerPoint Prototype of fabricated antenna (a) Front view, (b) Back view The gain values of the antenna are found using Friis formula (1) or (2) utilising the same setup as used for measurement of radiation patterns of the proposed antenna (1) (2) where |S21′| is the ratio of the received power 'PR' to the transmitted power 'PT', 'GT' and 'GR' are the gains of Tx and Rx antennas, respectively, 'R' is the distance between them and 'λ' is operating wavelength. Losses in the coaxial cables used on the transmitting and receiving sides were also taken into account while using the Friis formula to calculate the gain 'GR' of the proposed antenna at the frequencies of interest. However, the values of realised gain (GR′) are calculated using the following equation: (3) where S11 is the reflection coefficient at antenna port 1 when the signal is fed at port 1. Further, the total efficiency measurement was performed using Wheeler cap method as explained in [[17]]. Equation (4) was used to determine the total efficiency ηT of the proposed antenna from the measured parameters (4) where S11FS and S11WC are the values of reflection coefficient of the proposed antenna when the antenna was placed in free space and inside the Wheeler cap, respectively. Moreover, measured values of envelope correlation coefficient (ECC) were obtained from measured S-parameters using (8) [[23]]. Simulated and corresponding measured results for S-parameter–frequency characteristics, radiation patterns, gain–frequency characteristic, total efficiency–frequency characteristic and variations of ECC of the proposed antenna versus frequency are compared and discussed in different sub-sections of results and discussion section. 4 Results and discussion 4.1 S-parameters–frequency characteristics The variations of simulated and measured S-parameters (S11, S21, S12 and S22) of proposed MIMO antenna with frequency are shown in Fig. 3. Figs. 3a and b depict the reflection coefficient–frequency characteristics and transmission coefficient–frequency characteristics, respectively, when exciting antenna ports (port 1/port 2) one at a time. It is clear from Fig. 3 that respective simulated and measured S-parameter–frequency characteristics of the antenna nearly match each other. It is also clear from Fig. 3a that the S11 and S22 are almost identical to each other except for frequencies of more than 7.5 GHz whereas S21 and S12 (shown in Fig. 3b) are identical over the operating frequency range of the antenna. Fig. 3Open in figure viewerPowerPoint S-parameters–frequency characteristics of the proposed antenna (a) Reflection coefficient, (b) Transmission coefficient The discrepancy between S11 and S22 values at different frequencies may arise because of the shape of meandered line which does not look symmetrical from both input ports of the antenna while S21 and S12 values are identical because length of meander line in terms of wavelength at any frequency of interest is same for both ports. However, it is observed from Fig. 3a that simulated reflection coefficient (S11 or S22) values ≤−10 dB for the proposed antenna are obtained for frequencies starting from 2.4 GHz while its measured reflection coefficient values remain ≤−10 dB for frequencies starting from 2.5 GHz. Similarly, it is observed from Fig. 3b that simulated and measured transmission coefficient (mutual coupling, S21/S12) values ≤−15 dB between antenna ports are obtained over frequency ranges 2.00–12.75 and 2.50–12.45 GHz, respectively. Therefore, it is found that the antenna provides simulated and measured bandwidths with respect to −10 dB reflection coefficients and −15 dB mutual coupling between ports of 136.63% (2.4–12.75 GHz) and 133.11% (2.5–12.45 GHz), respectively, which are nearly in agreement with each other. In addition, the steps involved for isolation enhancement over broad range of frequency spectrum without disturbing patch geometry of the antenna are explained using S-parameter–frequency characteristics shown in Fig. 4. The S-parameters (S11 and S21)–frequency characteristics when exciting port 1 are considered for different antenna configurations due to almost identical S-parameter–frequency characteristics when exciting either of the antenna ports as shown in Fig. 3. Different antenna configurations are designated as Ant 1, 2, 3, 4 and Proposed Ant as shown in Fig. 1c. Fig. 4Open in figure viewerPowerPoint S-parameter–frequency characteristics of different antenna configurations when exciting port 1 (a) S11, (b) S21 First configuration, 'Ant 1' pertains to shared radiator MIMO antenna with rectangular ground plane without any modification. It is observed from Fig. 4a that the 'Ant 1' provides S11 values ≤−10 dB over wide range of frequencies while S21 is ≥−15 dB for most of the frequencies except the frequencies around 4.1 GHz as shown in Fig. 4b. The mutual coupling (S21) between ports 1 and 2 of 'Ant 1' for most of the frequencies is relatively high because the path length (for current on surface of ground plane) between ports 1and 2 is less. Therefore, in the second configuration specified as 'Ant 2', an end-loaded stub which is symmetrical about a line drawn at an angle of 45° with respect to X-axis is attached to ground plane to increase the path length. It can be observed that the 'Ant 2' provides reduced S21 values over broader frequency spectrum as compared to 'Ant 1' while reflection coefficient values remain equal or below −10 dB over the frequency range 2.13–2.43 GHz and for frequencies more than 2.6 GHz. It is to be noted from Fig. 4b that S21 values for 'Ant 2' have gone down as compared to the values for 'Ant 1' but these values are not well below −15 dB. In the third configuration specified as 'Ant 3', a rectangular slot is created on the ground plane where stub is attached. The slot on ground plane is arranged in such a manner that it further increases the path length between antenna ports. It is observed from Fig. 4 that the 'Ant 3' provides much reduced values of S21 for lower and higher frequencies as compared to 'Ant 2' whereas S11 values are maintained at or below −10 dB over the frequency range 2.16–2.37 GHz and for frequencies more than 2.44 GHz. But still possibility exists to further reduce the S21 values for lower frequencies. In the fourth configuration mentioned as 'Ant 4', the length of stub is increased by creating a meander line on stub. It can be further observed from Fig. 4 that 'Ant 4' provides smaller S21 values towards lower frequencies while slight increase in lower cutoff frequency is obtained on the basis of S11–frequency characteristics of 'Ant 4' as compared with 'Ant 3'. It is found that the reduced S21 values towards lower frequency side are achieved due to the meander line stub which increases the path length (for current flow on ground plane) between antenna ports. Though S21 values of 'Ant 4' are reduced below −14 dB over wider operating frequency range but slight increase in lower cutoff frequency (at which S11 = −10 dB) of 'Ant 4' as compared with 'Ant 3' needs further investigation. In the final configuration, that is, the Proposed Ant, the outer boundaries of ground plane are curved. It is observed from Figs. 4a and b that Proposed Ant provides slightly reduced lower cutoff frequency as compared to 'Ant 4' whereas S21 values are further reduced for lower frequencies and maintained at or below −15 dB over whole of its operating frequency range. It is noted that the curved boundaries of ground plane help in smooth transition of current from feed line to patch as well as around the boundary of ground plane. Moreover, it is inferred through discussion on S21–frequency characteristics of different antenna configurations that the modifications done in the ground plane shape have a very important role to play in the isolation enhancement between the antenna ports. 4.2 Surface current distributions To better understand the influence of different antenna configurations (or the effect of different modifications done on ground plane) on isolation between antenna ports 1 and 2, the current distributions on the surfaces of different antenna configurations (considered here) are discussed in this section. Fig. 5 depicts the surface current distributions for different antenna configurations when exciting port 1 at the frequencies of 2.5 and 6.0 GHz where S21 values are relatively higher for 'Ant 1' (first antenna configuration). It is to be noted that S21 value for 'Ant 1' is lowest among different antenna configurations at the frequency of 4.1 GHz. Fig. 5Open in figure viewerPowerPoint Simulated surface current distributions of different antenna configurations It can be clearly noticed from Fig. 5 that higher level of current is found near port 2 when exciting port 1 of 'Ant 1' at frequencies of 2.5 and 6.0 GHz which indicates that mutual coupling between ports 1 and 2 of 'Ant 1' is high. It is also noticed that for 'Ant 1', higher level of current reaches port 2 from port 1 through ground plane. Therefore, to minimise the mutual coupling, a stub was attached to ground plane in 'Ant 2' as shown in Fig. 1c. The stub increases the path length between ports 1 and 2. Subsequently, it is observed that for 'Ant 2' level of surface current is reduced near port 2 at 6.0 GHz while it is increased at 2.5 GHz as compared to 'Ant 1'. It is investigated from this observation that mutual coupling reduces at higher frequencies while it increases or remains higher for lower frequencies in 'Ant 2' due to the stub. In order to minimise mutual coupling at lower frequencies too, a slot was created in the ground plane of 'Ant 2' which further increases the path length between the antenna ports and the resultant antenna configuration is designated as 'Ant 3'. It is noticeable for 'Ant 3' that the levels of current near port 2 are reduced at both frequencies of 2.5 and 6.0 GHz as compared to 'Ant 2'. This observation indicates that the combination of stub and slot reduces the mutual coupling for both higher as well as lower frequencies due to the increased path length between the antenna ports. The mutual coupling remains higher for frequencies <2.4 GHz as can be seen from Fig. 4b. For this reason, the meander line stub is created and the resultant antenna configuration is designated as 'Ant 4' (shown in Fig. 1c). It can be observed for 'Ant 4' that levels of current near port 2 (proportional to mutual coupling) are still maintained at lower level at the frequencies of 2.5 and 6.0 GHz as shown in Fig. 5. Mutual coupling is further reduced below −14 dB for frequencies <2.4 GHz (shown in Fig. 4a) using meander line stub, which provides the increased path length as compared to 'Ant 3'. To further reduce the mutual coupling below −15 dB for frequencies <2.4 GHz, the ground plane is curved from the corner and the resulting antenna configuration is designated as 'Proposed Ant'. It is observed from the surface current distribution on 'Proposed Ant' that higher levels of current are smoothly distributed over the boundary of ground plane near port 1 whereas current distribution has lower levels near port 2 for frequencies of 2.5 and 6.0 GHz as compared to 'Ant 4'. Further, it is already observed from Fig. 4b that the S21 (mutual coupling) value of proposed Ant is reduced below −15 dB for frequencies lower than 2.5 GHz. It is inferred from the aforesaid observations that the reduced mutual coupling for lower frequencies is due to the smooth transition of current around the longer boundary of curved ground plane in Proposed Ant which supports lower frequencies (higher wavelength) as compared to straight ground plane in 'Ant 4'. 4.3 Radiation patterns Fig. 6a shows 3D radiation patterns in terms of realised gain of proposed MIMO antenna at 6 GHz when exciting port 1/2 while match-terminating port 2/1 with 50 Ω load. It is observed from Fig. 6a that the radiation patterns are slightly directive towards upper boundary of patch near excited port. When port 1 of the antenna is excited, maximum radiation occurs around −Y-axis (around patch boundary near port 1) whereas when port 2 is excited, maximum radiation takes place around −X-axis (around patch boundary near port 2). It is inferred from this observation that the radiation patterns for the excited ports 1 and 2 are mirror images of each other about a line drawn at an angle of 45° with respect to X-axis. Fig. 6Open in figure viewerPowerPoint Radiation patterns of the proposed MIMO antenna when exciting port 1 while match-terminating port 2 (a) 3D patterns, (b) 2D patterns Owing to the symmetrical patterns, 2D radiation patterns are discussed with only one port excited and other port match-terminated. Fig. 6b shows the simulated and measured normalised power patterns in H- and E-planes of proposed antenna with port 1 excited and port 2 match-terminated with 50 Ω load at discrete frequencies of 3, 6, 9 and 12 GHz. It is observed from Fig. 6b that the simulated and measured 2D radiation patterns deviate from each other and this deviation increases with frequency. The deviation in simulated and respective measured radiation patterns is due to fabrication and measurement errors, especially the errors caused in manual placement of the test antenna on the positioner system while making pattern measurements and in manual soldering of the coaxial connector to the antenna input. The slight tilt of test antenna plane with respect to the axial direction during manual placement can cause experimental radiation patterns to deviate from simulated patterns and this deviation would increase with frequency. It is also noticed that the simulated H-plane patterns are quasi-omnidirectional while simulated E-plane patterns are slightly directional at 270° (around the outer boundary of patch). Further, it is noticed that as frequency increases, patterns get distorted with multiple lobes. Generation of higher order modes is the probable cause of pattern distortion at higher frequencies. 4.4 Gain–/efficiency–frequency characteristics The variations of simulated and measured values of realised gain of the proposed antenna with frequency when exciting port 1 and match-terminating port 2 with 50 Ω load are depicted in Fig. 7a. It is clear from Fig. 7a that variations of simulated and corresponding experimental gain values of the antenna are nearly in agreement with each other over its operating frequency range. However, the simulated gain values vary over the range 2.33 dBi (at 2.5 GHz) to 5.22 dBi (at 13 GHz) while the experimental gain values vary in the range 2.0 dBi (at 2.5 GHz) to 4.92 dBi (at 12.5 GHz) over the frequency range 2–13 GHz. Fig. 7Open in figure viewerPowerPoint Realised gain–, total efficiency– and ECC–frequency characteristics of the proposed MIMO antenna when exciting port 1 and match-terminating port 2 (a) Realised gain, (b) Total efficiency, (c) Envelope correlation coefficient Further, the variations of simulated and measured values of total efficiency of proposed antenna with frequency are also shown in Fig. 7b. The trend of variation in simulated total efficiency values of the antenna with frequency is similar to that observed in the experimental values with frequency. It is observed from Fig. 7b that simulated and experimental values of total efficiency are more than 75 and 50%, respectively, over the operating frequency range of the proposed antenna. The total efficiency was measured using Wheeler cap method [[17]]. The discrepancy between simulated and experimental total efficiency values is due to the manual fabrication of Wheeler cap and experimental errors. It is to be noted that antenna must be properly grounded with the cap and placed exactly at the centre inside the completely closed cap to obtain accurate results. It is clearly noticed that the gain values increase while total efficiency values decrease with increase in operating frequency. The increase in gain with frequency is due to increase in the antenna dimensions in terms of wavelength while the decrease in total efficiency with increase in frequency is due to increase in substrate losses with frequency. Furthermore, the deviation between simulated and experimental total efficiency values increases with frequency due to bigger electrical size of cap at higher frequencies. 4.5 ECC–frequency characteristics The diversity performance of the proposed MIMO antenna was studied using ECC. ECC indicates correlation between the radiation patterns of antenna elements in a MIMO antenna system, and the value of ECC should be as low as possible. For a lossy antenna system, the ECC values can be computed using its relationship with far-field parameters [[24]], as described in (5)–(7) (5) where ρc is complex correlation coefficient related to far-field pattern (6) where (7) where 'XPR' represents cross-polarisation ratio of incident wave, is the vector radiation pattern associated with the antenna i = 1, 2, '*' symbolises complex conjugate, and and are the θ and ϕ components of power angular density functions of the incoming wave. It is assumed in the proposed study that the antenna operates in uniform multipath scattering environment with XPR equal to 1 and Pθ = Pϕ = 1/4π (However, in a practical scenario, ECC would depend on actual power distribution of incident wave – uniform, Gaussian, Laplacian distributions or their combinations.). The variation of ECC of the proposed antenna (computed through simulation using far-field parameters) versus frequency is shown in Fig. 7c. It is observed that the ECC values obtained using far-field parameters are <0.02 over the whole operating frequency spectrum. In addition, the ECC values are also computed through simulation using its relationship with S-parameters [[23]] (applicable for lossless MIMO antenna system and uniform scattering environment) as described in the following equation: (8) where S11 and S22 are the reflection coefficients of antenna ports 1 and 2, respectively, when signals are fed at ports 1 and 2, and S12 and S21 are, respectively, the values of coupling from antenna ports 2 to 1 and from ports 1 to 2. The variations of simulated and measured values of ECC related to S-parameters of the proposed MIMO antenna with frequency are also depicted in Fig. 7c. It is found from Fig. 7c that the simulated and measured values of ECC related to S-parameters are <0.005 and 0.006, respectively, and both variations are nearly in agreement with each other over whole operating frequency range of the proposed antenna. The deviation in the simulation results for ECC using far-field parameters and those using scattering parameters and measured results over the operating frequency range of the proposed antenna may be due to different assumptions made while arriving at the computed results using far-field parameters and S-parameters as well as experimental errors. Further, it is clearly noticed that the ECC values simulated/measured using far-field parameters and S-parameters over whole operating frequency range are well below the acceptable limit of 0.1. It indicates that the proposed MIMO antenna has good diversity performance. 5 Conclusion A new leaf-shaped compact two-port MIMO antenna having shared radiator for broadband applications has been proposed in this paper. The simulated and experimental studies of the antenna including variations of reflection coefficients and mutual coupling between antenna ports versus frequency, surface current distributions, radiation patterns as well as variations of realised gain, total efficiency and ECC values of the antenna versus frequency have been reported. The proposed antenna has been found to provide broad bandwidth and low mutual coupling between the ports over the frequency range 2.4–12.7 GHz, which covers UWB along with Bluetooth frequency spectrum. Low mutual coupling has been achieved with the help of end-loaded meander line stub attached to ground plane without creating any slot in shared radiator. The proposed antenna exhibits good realised gain as well as total efficiency along with good diversity performance (ECC <0.02) over the operating frequency spectrum of interest. Owing to these inherent properties, the proposed low-profile compact MIMO antenna may find usability in broadband diversity systems and compact wireless communication systems requiring UWB along with Bluetooth spectrum. 6 References [1] FCC 02–48: ' First report and order in the matter of revision of part 15 of the commission's rules regarding ultra-wideband transmission systems', 2002 [2]Kaiser, T., Zheng, F., Dimitrov, E.: 'An overview of ultra-wideband-systems with MIMO', Proc. IEEE, 2009, 97, pp. 285– 312 [3]Liu, L., Cheung, S.W., Yuk, T.I.: 'Compact MIMO antenna for portable devices in UWB applications', IEEE Trans. Antennas Propag., 2013, 61, (8), pp. 4257– 4264 [4]Gao, P., He, S., Wei, X., et al.: 'Compact printed UWB diversity slot antenna with 5.5 GHz band-notched characteristics', IEEE Antennas Wirel. Propag. Lett., 2014, 13, pp. 376– 379 [5]Ren, J., Hu, W., Yin, Y., et al.: 'Compact printed MIMO antenna for UWB applications', IEEE Antennas Wirel. Propag. Lett., 2014, 13, pp. 1517– 1520 [6]Kiem, N.K., Phuong, H.N.B., Chien, D.N.: 'Design of compact 4 × 4 UWB-MIMO antenna with WLAN band rejection', Int. J. Antennas Propag., 2014, 2014, pp. 1– 11, doi:10.1155/2014/539094 [7]Huang, H., Liu, Y., Zhang, S.-S., et al.: 'Compact polarization diversity ultrawideband MIMO antenna with triple band-notched characteristics', Microw. Opt. Technol. Lett., 2015, 57, (4), pp. 946– 953 [8]Tripathi, S., Mohan, A., Yadav, S.: 'Compact octagonal fractal UWB MIMO antenna with WLAN band-rejection', Microw. Opt. Technol. Lett., 2015, 57, (8), pp. 1919– 1925 [9]Tao, J., Feng, Q.Y.: 'Compact isolation enhanced UWB MIMO antenna with band-notch character', J. Electromagn. Waves Applicat., 2016, 30, pp. 2206– 2214, doi: 10.1080/09205071.2016.1217173 [10]Zhao, H., Zhang, F., Wang, C., et al.: 'A universal methodology for designing a UWB diversity antenna', J. Electromagn. Waves Applicat., 2014, 28, (10), pp. 1221– 1235 [11]Khan, M.S., Capobianco, A.D., Shafique, M.F., et al.: 'Isolation enhancement of a wideband MIMO antenna using floating parasitic element', Microw. Opt. Technol. Lett., 2015, 57, (7), pp. 1677– 1682 [12]Tang, T.-C., Lin, K.-H.: 'An ultrawideband MIMO antenna with dual band-notched function', IEEE Antennas Wirel. Propag. Lett., 2014, 13, pp. 1076– 1079 [13]Liu, L., Cheung, S.W., Yuk, T.I.: 'Compact MIMO antenna for portable UWB applications with band-notched characteristics', IEEE Trans. Antennas Propag., 2015, 63, (5), pp. 1917– 1924 [14]Li, Q., Feresidis, A.P., Mavridou, M., et al.: 'Miniaturized double-layer EBG structures for broadband mutual coupling reduction between UWB monopoles', IEEE Trans. Antennas Propag., 2015, 63, (3), pp. 1168– 1171 [15]Malekpour, N., Honarvar, M.A.: 'Design of high-isolation compact MIMO antenna for UWB application', Progress In Electromagnet. Res. C., 2016, 62, pp. 119– 129 [16]Liu, X.-L., Wang, Z.-D., Yin, Y.-Z., et al.: 'A compact ultrawideband MIMO antenna using QSCA for high isolation', IEEE Antennas Wirel. Propag. Lett., 2014, 13, pp. 1497– 1500 [17]Patre, S.R., Singh, S.P.: 'Broadband MIMO antenna using castor leaf-shaped quasi-self-complementary antenna with dual band-rejection characteristics', IET Microw. Antennas Propag., 2016, 10, (15), pp. 1673– 1681 [18]Mao, C.-X., Chu, Q.-X.: 'Compact coradiator UWB-MIMO antenna with dual polarization', IEEE Trans. Antennas Propag., 2014, 62, (9), pp. 4474– 4480 [19]Zhang, J.-Y., Zhang, F., Tian, W.-P., et al.: 'ACS-fed UWB-MIMO antenna with shared radiator', Electron. Lett., 2015, 51, (17), pp. 1301– 1302 [20]Srivastava, G., Kanuijia, B., Paulus, R.: 'UWB MIMO antenna with common radiator', Int. J. Microw. Wirel. Technol., 2016, 9, pp. 573– 580 [21]Srivastava, G., Kanuijia, B.K.: 'Compact dual band-notched UWB MIMO antenna with shared radiator', Microw. Opt. Technol. Lett., 2015, 57, (12), pp. 2886– 2891 [22]Khan, M.S., Capobianco, A.D., Iftikhar, A., et al.: 'A compact dual-polarized ultrawideband multiple-input-multiple-output antenna', Microw. Opt. Technol. Lett., 2016, 58, (1), pp. 163– 166 [23]Blanch, S., Romeu, J., Corbella, I.: 'Exact representation of antenna system diversity performance from input parameter description', Electron. Lett., 2003, 39, (9), pp. 705– 707 [24]Taga, T.: 'Analysis for mean effective gain of mobile antennas in land mobile radio environments', IEEE Trans. Veh. Technol., 1990, 39, (2), pp. 117– 131 Citing Literature Volume12, Issue7June 2018Pages 1153-1159 FiguresReferencesRelatedInformation
Referência(s)