Sie sind auf Seite 1von 8

3184

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 54, NO. 11, NOVEMBER 2006

Correlation-Based Pattern Stability Analysis and a Figure of Merit for UWB Antennas
Tharaka Dissanayake, Student Member, IEEE, and Karu P. Esselle, Senior Member, IEEE
AbstractCombining the concepts of pulse delity, time-domain correlation patterns, and frequency-domain pattern stability, we propose the concept of frequency-domain correlation patterns and a gure of merit, called pattern stability factor (PSF), to characterize pattern stability of wideband antennas. The frequency-domain correlation pattern represents the relationship between radiation in a reference direction to radiation in all other directions of interest, over a specied bandwidth. With practical examples, we demonstrate that this tool is very useful when selecting a single transfer function (and a reference direction) for the antenna when designing or optimizing an ultrawideband (UWB) system. Careful selection of the reference direction is crucial for antennas with relatively less stable patterns, but even antennas with relatively stable patterns can perform poorly overall if an inappropriate reference direction is chosen. The PSF represents the overall pattern stability of an antenna for a specic bandwidth and a range of directions. It is useful as a quantitative measure: a) to compare the suitability of different antennas for a given wideband or multiband application; b) to assess pattern stability improvement methods; and c) to quantify the effect of packaging, etc., on stability. We also propose the concept of PSF bandwidth and show how it can be used to identify stable frequency bands of a given antenna within its impedance bandwidth, and how it compares with the impedance bandwidth of some UWB antennas. Index TermsAntenna bandwidth, broadband, characterization, correlation, delity, multiband, optimization, pattern stability factor (PSF), radiation pattern, stability, transfer function, ultrawideband (UWB), wideband.

I. INTRODUCTION ARAMETERS such as pulse delity and correlation energy patterns are presented in [1] for time-domain characterization of antennas. Pulse delity is a measure of pulse deformation caused by the antenna, whereas a time-domain correlation pattern represents relationship between radiated pulses in different directions. Furthermore, recently in [3], concepts of correlation energy pattern and correlation energy efciency have been introduced. In this paper, we use those denitions to create new parameters to characterize and quantify performance of ultrawideband (UWB) antennas and demonstrate their use through practical applications. In UWB system design and optimization schemes, the antenna is represented by a single transfer function [4][7]. However, modern UWB antennas are complicated, and their transfer functions depend on the direction. Even the transfer function of an ideal monopole can become direction dependent, when the

antenna is integrated to the case of an electronic system. Therefore, it is necessary to select a single transfer function (or equivalently, a single reference direction) out of an innite number of choices available, for system design and optimization. This paper describes a novel approach to select the best transfer function for a given antenna in a given application. It is based on pulse correlation concepts but conducted in the frequency domain. Its applications to antennas with relatively stable and unstable patterns are demonstrated through examples. We investigate the close relationship between antenna pattern stability and frequency-domain correlation patterns and highlight the advantages of the latter. Pattern comparison between two different frequencies is possible as presented in [2]. However, in the context of UWB antennas, stability within an entire band should be considered. Hence, we propose a new gure of merit to represent overall pattern stability performance of a UWB antenna. Based on this gure of merit: a) pattern stability performance of different antennas can be compared quantitatively; b) effects of ground-plane modications, packaging, integration of antennas with cases, element interactions [12], etc., can be examined; c) pattern stability improvement methods can be assessed [13]; d) a pattern stability bandwidth can be dened. Section II of this paper summarizes background theory. New frequency-domain correlation patterns and the gure of merit are dened in Section III. The applications of these concepts to several UWB antenna congurations are given in Section IV to demonstrate their practical value. Sections V and VI present a brief discussion and conclusions, respectively. II. THEORY A. Basis Energy patterns for pulse radiating antennas are dened in the time domain in [3] as (1) and

Manuscript received September 15, 2005; revised March 28, 2006. This work was supported by Macquarie University under an International Scholarship. The authors are with Macquarie University, Sydney NSW 2109, Australia (e-mail: tharaka@ics.mq.edu.au; esselle@ics.mq.edu.au; karu@ieee.org). Digital Object Identier 10.1109/TAP.2006.883965 0018-926X/$20.00 2006 IEEE

(2)

DISSANAYAKE AND ESSELLE: CORRELATION BASED PATTERN STABILITY ANALYSIS

3185

where is correlated energy, is total energy, is the inis the time-domain raditrinsic impedance of free space, , and is the template ated pulse at a far eld point , function that can be preselected. In general, the radiated pulse of a UWB antenna depends on the direction. In this paper, we select the radiated electric-eld waveform in a particular direction, hereafter called the reference . Then, to quantify the correlation direction, as the template of the radiated pulse in the reference direction to any other direction of interest , we dene , which is identical to the normalized correlation coefcient in [3] when the radiated eld in the reference direction is chosen as the template function

is independent between two directions. Note that as of the electric-eld magnitude, the antenna need not be an in all omnidirectional radiation pattern to maximize directions. For an antenna with low dispersion, the integral in the numerator of (4) is real, as phase values are cancelled out. Moreover, normalized radiation patterns can give a good feel of pattern stability and hence can be used to calculate (4), as a direct consequence of (P2). III. FREQUENCY-DOMAIN CORRELATION PATTERNS AND FIGURE OF MERIT A. Frequency-Domain Correlation Pattern First we dene the frequency-domain correlation pattern , which represents how the radiated waveform in the is correlated to all other directions of reference direction interest on average, as

(3)

Note that the template function is selected as equal to the electric eld in the reference direction. Therefore, the correlation . For other direcin that direction is maximized and tions, correlation energy depends on the relationship between the pulse in the given direction and the reference direction. We shall now show that for a pattern stable antenna, pulses in all the directions are highly correlated by paying attention to the propfactor in (3). erties of B. Properties of the Factor

(6) where is the range of operating directions of the antenna (i.e., eld of view) and the integration is with respect to the direction of interest . For an antenna intended for spherical coverage, is a sphere and the integration is double over a solid angle of 4 . If operation is limited to a plane (e.g., azimuth), it becomes a line integral. Likewise, the integration can be limited to any arc, solid angle, or polarization. Note that reaches in all directions, a the maximum value of unity if condition difcult to achieve in practice. However, we shall consider this ideal situation as a theoretical reference. In practical system design or optimization, we search for a direction in is maximized. This direction is considered unique, which as it is the best direction that can be selected for optimization. Optimization of a UWB antenna could be to achieve the highest data rates through amplitude maximization [5], or through reduced intersymbol interference, while complying with Federal Communications Commission (FCC) regulations [4], [6], [7]. Signal processing theorems have been extensively used in optimizing signals for UWB systems. In almost all of those investigations, the antenna is considered as a two-port device and represented by a single transfer function [10]. However, in the case of many UWB antennas, the radiating transfer function depends on the direction. The direction along which is maximum is the direction that is best correlated to all other directions on average. Hence, the transfer function in this direction is the best one to represent the antenna in UWB system design and optimization. As an example, Appendix A shows the importance of pattern stability in achieving pulse amplitude optimization [5] simultaneously in all directions. for the opGenerating the correlation pattern curve erating directional range is especially important when an antenna has unstable radiation patterns within the operating band. It clearly shows the best reference direction(s) for the operating bandwidth concerned. If an antenna has an extremely stable patwill be close to unity for all the tern throughout the band, reference directions under consideration.

The following observations on (3) are found useful. P1) The value of and is maximum if and only , where k is a constant in time. if P2) Its value is independent of electric-eld amplitude. Therefore, value remains unchanged under radiation pattern normalization. P3) It can be calculated in the frequency domain using Parseval theorem and Fourier integral theorem. in the frequency domain Therefore, we can dene in the time domain (3) as equivalent to

(4)

where stands for complex conjugate and BW is the bandwidth of interest. Now, from property (P1) and Fourier transform, the condition for maximization can be expressed as (5a) or in logarithmic scale (5b) where and are independent of frequency and represents the gain (absolute or normalized) in decibels. Obviously (5) is the condition for ideal pattern stability. stands for the relationship between the radiated elds

3186

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 54, NO. 11, NOVEMBER 2006

B. Antenna Figure of Merit An ideally stable pattern can be expressed as (5) in the frequency domain. In other words, an ideal antenna would have for all and , and for all reference directions . Based on these observations, we are in a position to dene a single gure of merit to quantify the overall pattern stability performance of an antenna. We dene the pattern stability factor (PSF) for a given frequency band as (7) where is again the range of operating directions of the antenna and the integration is now with respect to the reference direction . It is obvious that PSF is unity for an ideally pattern stable antenna. For good all-round performance in a UWB system, PSF may be desirable. Selection of is solely based on the application, to cover only the directions of interest, not necessarily the complete sphere surrounding the antenna. Even when the directions of interest form a solid angle (three-dimensional shape) that normally require double integrations, using interpolation schemes, the integration may be done approximately over two two-dimensional planes using single integrals to save computation time and resources [9]. If the directions of interest are limited to one plane, only single integrals are required for all calculations. C. PSF Bandwidth One may note that (8) In other words, any antenna is an almost perfectly stable antenna within a sufciently narrow bandwidth. As we increase the operating bandwidth, the PSF of a typical antenna tends to deteriorate. With this observation, it is sensible to dene an antenna bandwidth based on its pattern stability performance. For example, 0.95 PSF bandwidth indicates a band within which overall pattern stability is better than 0.95. Antenna engineers would like to align this PSF bandwidth with the impedance bandwidth, to achieve good overall performance for a given wideband application. D. PSF Calculation The calculation of PSF is done as follows. First, the range of directions of interest is discretized, and the frequency-domain radiated electric eld values (at a xed distance in the far eld) are calculated in these discrete directions over the bandwidth of interest. This can be achieved directly with frequency-domain software such as HFSS, or time-domain solvers such as Microwave Studio followed by fast Fourier transform. Then, these electric eld values are substituted in (4) to calculate for all pairs of discrete directions. Finally frequency-domain correlation patterns and PSF are calculated according to (6) and (7), respectively.1
1MATLAB codes useful for these computations will be available at www. elec.ics.mq.edu.au/celane.

The PSF can also be calculated using experimental data. When both the amplitude and phase of the measured radiation patterns are available, they can be substituted in (4) just like theoretical data. Note that absolute eld strength is not required; a scaled or normalized pattern is sufcient. For antennas with stable phase centers, the phase values in (4) cancel out and hence only amplitude (scalar) measurements are sufcient. Alternatively, transmission between two identical antennas can be measured to determine PSF. In this case, one antenna is rotated with respect to the other, and the transmission from the input of one antenna to the output of the other is measured using a vector network analyzer [10]. The measured is substituted for the electric eld values in (4). IV. APPLICATIONS In this section, we investigate two example applications of antenna correlation patterns and PSF. In the rst example, we quantitatively estimate the improvement of the pattern stability (and pattern stability bandwidth) of a planar monopole antenna when its symmetry is improved by extending it to a cross-planar monopole conguration [13]. In the second example, we quantify the degradation of pattern stability and overall bandwidth of an L-shaped wide slot antenna when its ground plane is modied with the addition of sidewalls [14]. In both cases, we discuss how useful the correlation patterns are when a reference direction is chosen for UWB system design and optimization. A. Planar Monopoles Planar monopoles are popular as UWB antennas, due to their extremely wide impedance bandwidth (see Fig. 4). The planar element of the antenna can be square, circular, elliptical, or even triangular [16]. It has been observed that, even though planar monopoles are considered similar to whip antennas, even on the orthogonal (H) plane, the radiation in different directions varies signicantly, mainly due to the geometrical asymmetry and niteness of the ground plane. The cross-planar monopole has been proposed to improve the pattern stability of the conventional planar monopole antenna. Fig. 1 shows a planar monopole antenna and an improved cross-planar monopole antenna. We have simulated both antennas using CST Microwave Studio time-domain software. Since major improvement in stability is observed in the XY plane (i.e., H plane), our investigation is limited to an component on the XY plane. In our simulations, probes are in the XY plane. placed in the far eld, 1 m away, parallel to The correlation patterns and PSF have been calculated on the azimuth plane. Table I summarizes the PSF of the two monopole antennas under investigation for the frequency range of 3.110.6 GHz on the azimuth plane. It is evident that stability has signicantly improved by introducing the cross-planar element to the monopole. To understand this improvement in the conventional way, the normalized radiation patterns of the two antennas are shown in Fig. 2 for the range of frequencies concerned. From Fig. 2(a), it can be seen that the planar antenna has highly unstable radiation pattern with frequency in the directions of 90 and 270 , whereas the variation in other directions is relatively low. In contrast, the patterns of the cross-planar monopole in

DISSANAYAKE AND ESSELLE: CORRELATION BASED PATTERN STABILITY ANALYSIS

3187

Fig. 1. Planar and cross-planar monopole antennas on nite ground planes. The two antennas have the same dimensions. The inner conductor of the probe feed has a radius of 1.12 mm and the outer conductor inner radius is 4 mm. Relative dielectric constant of the coaxial line material is 2.33.

TABLE I PATTERN STABILITY FACTOR OF FOUR ANTENNAS

Fig. 2. E radiation pattern on the XY plane (a) for planar monopole antenna and (b) for cross-planar monopole antenna.

Fig. 2(b) have only a 4 dB variation within the whole frequency range considered. The higher PSF value of the cross-planar antenna represents this improvement in an averaged and quantied manner. To identify the best reference direction for system design, optimization, or any such purpose, we calculated the correlation patterns of these two antennas over the azimuth plane in the frequency range of 3.110.6 GHz, and the results are shown in Fig. 3. (Due to symmetry, results for only one quadrant are shown.) The cross-planar antenna has a signicantly higher correlation pattern for all directions concerned. For comparison, [17], which (as further Fig. 3 also shows the average of discussed in Appendix A) is the peak electric-eld amplitude in direction when a system is optimized for peak amplitude in the reference direction , divided by the peak electric-eld amplitude in direction when a system is optimized for the peak amplitude in the same direction. The averaging is done with respect to the direction by integrating over the azimuth plane as

Fig. 3. Frequency-domain correlation patterns and P curves for planar and cross-planar monopole antennas. The x-axis represents the reference direction ().

(9)

We can recognize that a higher value of in a given reference direction means a better overall pulse amplitude optimization in

all the directions of interest, when the system is optimized for this reference direction. It can be seen from Fig. 3 that as a result of good pattern stability, an optimized UWB system based on a cross-planar antenna can reach nearly maximum pulse amplitude in all directions independent of the selection of the reference direction. However, in case of the planar monopole, the selection of reference direction is crucial to achieve the best curve in Fig. 3 indicates optimization results overall. The that the reference direction should be in the range of 4060 . A system optimized for any other reference direction will not perform well in all directions overall. In fact, we can reach the same conclusion simply observing the correlation pattern ( ) curve for the planar monopole an-

3188

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 54, NO. 11, NOVEMBER 2006

Fig. 4. Reection coefcient magnitude against frequency for each antenna given in the examples.

Fig. 6. (a) L-slot antenna on planar ground plane and (b) modied L-slot antenna with side walls. Ground plane is located at z = 0, stub and feed at z = 0:8, and sidewalls are from z = 0 to 40.

Fig. 5. Overall bandwidth as a function of the lower limit of the band, for the two planar monopole antennas.

tenna, because both and curves have the same shape. The correlation pattern indicates that the reference directions in the range of 4060 are the best to use as they are more correlated with all directions on average. Hence, using the correlation pattern curve, the best reference direction (and the transfer function) can be selected without cumbersome calculations of or (which also require antenna impedance over the whole frequency range of interest). We also calculated the overall bandwidth of the two monopole antennas. We dene the overall bandwidth with respect to both impedance matching and pattern stability. Within this bandwidth the return loss of the antenna is greater than 10 dB and the PSF is greater than 0.95. The return loss of the two monopole antennas is shown in Fig. 4. Fig 5 shows their overall bandwidth against the lower limit of the band, which was used as the independent variable for each calculation. The upper frequency limit of operation is taken as 10.6 GHz. For the cross-planar monopole, the PSF bandwidth aligns with the impedance bandwidth, and hence the overall bandwidth covers the full UWB band from 3.1 to 10.6 GHz. On the other hand, the largest overall bandwidth of the planar monopole is only from 6.25 to 10.6 GHz. Expected degree of stability can only be

achieved for less than 2.5 GHz bandwidth when the operating band starts below 6.1 GHz. With this type of analysis, it is easy to identify stable bands within the impedance bandwidth of a given antenna. It is not easy to predict sharp variations, for example, seen around 6.1 GHz in the planar monopole, using conventional pattern stability analysis, which is based on radiation patterns in Fig. 2. B. L-Slot Antenna A compact L-slot UWB antenna, which can be integrated with the metal casing of a communication device, has been proposed in [14]. This L-slot antenna on a planar ground plane (40 40 mm), and a modied integrated L-slot antenna with two sidewalls perpendicular to the ground plane (40 mm tall), are depicted in Fig. 6. When modied, impedance matching degrades due to the effects of the metal walls but the antenna can be retuned to compensate for those effects by adjusting a few parameters. Both antennas in Fig. 6 are well matched (with return loss dB) to operate in the band of 5.510 GHz, as shown in Fig. 4. However, in UWB applications, it is equally important to investigate the effects of sidewalls on radiation pattern stapattern on the XZ plane bility. In this example, we consider (which is the azimuth plane if the antenna is mounted vertically on a UWB system).

DISSANAYAKE AND ESSELLE: CORRELATION BASED PATTERN STABILITY ANALYSIS

3189

Fig. 9. Overall bandwidth as a function of lower limit of the band for the two L-slot antennas.

Fig. 7. E radiation pattern on the XZ plane: (a) for the original L-slot antenna and (b) for modied L-slot antenna with side walls.

worst directions are only about 20 away. This example highlights the importance of correlation patterns when selecting the reference direction (and hence reference transfer function) in system design and optimization. Interestingly, 200 is the worst reference direction for the original antenna, and for that and values are even less than those of the direction its modied antenna. However, its performance is much better if most of the other reference directions are used, as expected from its high PSF value. We note from this analysis that a) a stable antenna optimized for the wrong direction can perform more poorly than a less stable antenna optimized for the correct reference direction and b) the operating environment (case, metal walls, user, etc.) should be considered when selecting the reference direction and transfer function in system design and optimization. The return loss of the two antennas is plotted in Fig. 5 and their overall bandwidths are shown in Fig. 9. For the original antenna, the PSF and impedance bandwidths overlap from 6.1 to 10 GHz. After integration with sidewalls, the PSF bandwidth decreases signicantly, and usually the overall bandwidth falls to 11.5 GHz. V. DISCUSSION

Fig. 8. Frequency-domain correlation patterns and P curves for the original and modied L-slot antennas. The x-axis represents the reference direction ().

Table I lists the PSF for the two antennas in the intended operating frequency range of 5.510 GHz. Due to the modication, the PSF of the antenna with sidewalls has degraded. To explain this, the radiation patterns of the two antennas are shown in Fig. 7. One may note that for larger angles, the pattern of the modied antenna is highly unstable with frequency. The correlation patterns ( curves) of the two antennas are curves for comparison. shown in Fig. 8, together with the The correlation pattern of the less stable modied antenna varies signicantly with the reference direction. The best reference direction for the modied antenna is 200 but the two

For antennas with a xed phase center, the phase values in (4) cancel out and hence all calculations can be done with amplitudes only. Therefore, measured scalar radiation patterns (normalized or otherwise) can be used in correlation pattern and PSF calculations. Network analyzer measurement data obtained through a link between two identical antennas, while rotating one, can also be used [10], [11], [15]. Note that correlation patterns and PSF are insensitive to the absolute values of the patterns; only the shape matters. The range of directions used for PSF and correlation pattern calculations should be limited to the eld of view of the application. This range may be formed by several unconnected regions, for example, to exclude patterns nulls or other uninterested directions when required. For certain specic applications, it may be desirable to emphasize some frequencies at the

3190

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 54, NO. 11, NOVEMBER 2006

expense of others by including a predened frequency-depenin denition of dent weighting function

is the transfer function in the reference direcwhere , and are generator impedance, antenna input tion ; , impedance, and antenna input resistance, respectively. The constraint of 1 W input energy gives the Lagrange multiplier as

(A1b) (10) This may be the case when the power density spectrum of the radiated eld is capped by a frequency-dependent mask, as in UWB systems developed to operate in the United States under FCC regulations. The correlation patterns and PSF may be calculated accordingly. Although the scope of this paper is limited to transmitting antennas, the concepts and parameters can be extended to receiving antennas. VI. CONCLUSIONS The proposed PSF helps the antenna designer to quantitatively compare wideband radiation pattern stability performance of different antennas for a given application. The operating environment of the antenna should be included in its calculation as the PSF could be signicantly affected by the case, metal walls, user, etc. PSF bandwidth calculations allow one to locate pattern-stable bands within the impedance bandwidth of a given antenna. Subtle but signicant effects on stability, resulting from the frequency dependence of radiation patterns, can be easily detected from PSF bandwidth calculations. The overall antenna bandwidth indicates the frequency range within which an antenna is well matched and pattern-stable. When a UWB system is designed or optimized, the antenna is usually replaced with a single two-port transfer function. As the transfer function of modern wideband antennas is directiondependent in general, a reference direction should be chosen, and the transfer function in that direction is used for system design. The selection of the reference direction is vital in the case of relatively unstable antennas such as the modied L-slot antenna in Fig. 6(b) because its overall performance depends strongly on the choice of the reference direction. The frequency-domain correlation patterns proposed here are a convenient and efcient tool to identify the best reference directions for a given antenna in a given application. They are very useful even in the case of a relatively stable antenna, such as the original L-slot antenna in Fig. 6(a), to nd directions not to be used as a reference. A relatively stable antenna optimized for the wrong direction can perform more poorly than a less stable antenna optimized for the correct reference direction. In the case of printed UWB antennas, the best and worst reference directions cannot be identied by simply examining the antenna geometry or radiation patterns. APPENDIX A To maximize the pulse amplitude, an optimum generator function is given by [5] (A1a) Note that if another direction is selected for optimization, we may end up with a different source pulse and a Lagrange multiplier (A2a) (A2b) In a particular direction , we may consider two scenarios for and E-eld: one is when the antenna is excited by (A1a), . In the latter case, we the other is when excited by (2a), expect the maximum possible peak amplitude in the time domain for that direction. In this formulation, the maximum am. Therefore, we may compare the plitude is expected at time-domain maximum amplitudes for the two cases

(A3) where

Note that if the pattern is perfectly stable (A4) where K is a constant independent of frequency. Then from (A3), . In other words, for an ideally pattern stable antenna, it is possible to maximize the radiated pulse peak amplitude in all directions simultaneously. However, in a nonideal case, the reference direction to be chosen for optimization is important, as shown in Section III. REFERENCES
[1] D. Lamensdorf and L. Susman, Base band-pulse-antenna techniques, IEEE Antennas Propag. Mag., vol. 36, pp. 2030, Feb. 1994. [2] J. M. Rius, M. C. Santos, and J. Parron, Figure of merit for multiband antennas, IEEE Trans. Antennas Propag., vol. 51, pp. 31773179, Nov. 2003. [3] J. S. McLean, H. Foltz, and R. Sutton, Pattern descriptors for UWB antennas, IEEE Trans. Antennas Propag., vol. 53, pp. 553559, Jan. 2005. [4] Z. N. Chen, X. H. Wu, H. F. Li, N. Yang, and M. Y. W. Chia, Consideration for source pulses and antenna in UWB radio systems, IEEE Trans. Antennas Propag., vol. 52, pp. 17391748, Jul. 2004.

DISSANAYAKE AND ESSELLE: CORRELATION BASED PATTERN STABILITY ANALYSIS

3191

[5] D. M. Pozar, Waveform optimization for ultra wideband radio systems, IEEE Trans. Antennas Propag., vol. 51, pp. 23352345, Sep. 2003. [6] Y. Wu, A. F. Molisch, S.-Y. Kung, and J. Zhang, Impulse radio pulse shaping for ultra-wide bandwidth (UWB) sytems, in IEEE Int. Symp. Personal, Indoor Mobile Radio Commun., 2003. [7] G. Lu, P. Spasojevic, and L. Greenstein, Antenna and pulse designs for meeting UWB spectrum density requirements, in IEEE Conf. Ultrawideband Syst. Tech., Nov. 2003. [8] T.-G. Ma and S.-K. Jeng, Planar miniature tapered-slot-fed anuular slot antenna for ultrawide-band radios, IEEE Trans. Antennas Propag., vol. 53, pp. 11941202, Mar. 2005. [9] F. Gil, A. R. Carlo, J. M. Ferreira, C. Pardelinha, and L. M. Correia, A 3-D interpolation method for base-station-antenna radiation patterns, IEEE Antennas Propag. Mag., vol. 43, pp. 132137, Apr. 2001. [10] S. Zwierzchowski and P. Jazayeri, A system and network analysis approach to antenna design for UWB communication, in IEEE Int. Symp. Antennas Propag., Jun. 2003, vol. 1, pp. 826829. [11] B. Scheers, M. Acheroy, and V. Vorst, Time domain simulation and characterization of TEM horns using a normalized impulse response, Proc. Inst. Elect. Eng. Microw. Antennas Propag., vol. 147, no. 6, Dec. 2000. [12] G. Kotyrba and H. J. Chaloupka, On signal distortion in compact UWB arrays due to element interaction, in IEEE Int. Symp. Antennas Propag., Jul. 2005, vol. 1A, pp. 614617. [13] M. J. Amman, R. Sierra Cordoba, M. Uzelac, J. A. Evans, and A. T. Schwartzbatcher, On pattern stability of the crossed planar monopole, Microwave Opt. Technol. Lett., vol. 40, no. 4, Feb. 20, 2004. [14] T. Dissanayake, K. P. Esselle, and Y. Ge, Integrated compact ultrawideband L-shaped wide slot antenna, in Asia Pacic Microwave Conf., 2005, vol. 1, pp. 47. [15] C. E. Baum, E. G. Farr, and C. A. Frost, Transient gain of antennas related to the traditional continous wave (CW) denition of gain, Sensor and Simulation Notes, Jul. 2, 1997, Note 412. [16] N. P. Agrawall, G. Kumar, and K. P. Ray, Wideband planar monopole antennas, IEEE Trans. Antennas Propag., vol. 46, pp. 294295, Feb. 1998. [17] T. Dissanayake and K. P. Esselle, Waveform optimization of UWB radio systems over a wide range of directions, Electron. Lett., pp. 384385, Mar. 2006. Tharaka Dissanayake (S02) received the B.Sc.Eng. degree in electronics and telecommunications engineering (with rst-class honors) from the University of Moratuwa, Sri Lanka, in 2002. He is currently pursuing the Ph.D. degree in electronics at Macquarie University, Sydney, Australia. From 2002 to 2003, he was an Assistant Lecturer at the University of Moratuwa. He has been a Visiting Lecturer at Kotalawala Defence Academy and Arthur C. Clerk Centre for Modern Technology. He was a Junior Research Assistant in statistics for the Postgraduate Institute of Science, University of Peradeniya, Sri Lanka, in 1997. His current research interests are in ultrawideband antennas, antenna diversity, and microwave imaging techniques.

Karu P. Esselle (M92SM96) received the B.Sc. degree in electronic and telecommunication engineering (with rst-class honors) from the University of Moratuwa, Sri Lanka, in 1983 and the M.A.Sc. and Ph.D. degrees in electrical engineering from the University of Ottawa, Ottawa, ON, Canada, in 1987 and 1990, respectively. He is Director of Postgraduate Research in the Division of Information and Communication Sciences and Director of the Centre for Electromagnetic and Antenna Engineering, Macquarie University, Sydney, Australia. He has coauthored more than 150 papers, including ve invited book chapters. He is a Research Grant Assessor for several organizations, including the Australian Research Council (ARC), Hong Kong University Grants Commission, and Netherlands Organization for Scientic Research. He led the establishment of a national antenna testing facility funded by ARC and seven collaborating institutions, and he is currently directing it. His industry experience includes employment as a Faculty Hire Design Expert by Hewlett-Packard Laboratory, Palo Alto, CA, and several consultancies for local and international companies. His current research interests include metamaterials, photonic crystals, PBG/EBG structures, and frequency-selective surfaces; their microwave applications; antennas based on them; broadband antennas for wireless communication systems including WLANs and ultrawideband systems; smart antenna systems; antennas for radio astronomy, moment methods and FDTD methods for antennas and microwave circuits; and closed-form Greens functions. He was a Canadian Government Laboratory Visiting Postdoctoral Fellow from 1990 to 1992. In 1992, he joined Macquarie University, Sydney, where he is an Associate Professor in electronics. He was a Visiting Professor of the University of Victoria, Canada, in 19961997 and a Visiting Scientist of the CSIRO ICT Centre in 2002 and 2005. Dr. Esselle received a Canadian Commonwealth Scholarship from 1985 to 1990, an URSI Young Scientist Award in 1990, and the Senior Researcher Award from the Raj Mittra Travel Grant committee of the IEEE Antennas and Propagation Society (AP-S) in 2002. He has served on technical program committees or international committees of many international conferences, including the IEEE AP-S International Symposium, IEEE Vehicular Technology International Symposium. He was an Organizer and Publicity Chair of the 2000 Asia-Pacic Microwave Conference. He is the former Chair of the IEEE New South Wales MTT/AP Joint Chapter. At present he is Chair of the Educational Activities Committee of the IEEE NSW Section, and a member of the IEEE NSW Executive Committee. He pioneered the formation of the Macquarie University IEEE Student Branch and serves as the Branch Counselor at present. His recent awards include the 2004 Macquarie University Inaugural Invention Award for best innovation. His students and research associates have also received awards such as the IEEE MTT-S Graduate Fellowships, IEEE NSW Best Student Paper (Max Symon) award, and URSI Young Scientist Award.

Das könnte Ihnen auch gefallen