NRZ data and the clock are synchronized with each other byadjusting the ODL, NRZ-to-RZ format conversion will be realized.

Figure 12 shows the experimental results of NRZ-to-RZ formatconversion. Figure 12(a) shows the waveform of the original NRZdata, the recovered clock, TOAD’s switching window, and theconverted RZ data. As can be seen from this figure, there is noevident pattern effect in the converted RZ data and its extinctionratio is about 10 dB. Figure 12(b) shows the optical spectrum ofthe original NRZ data, the recovered clock, and the converted RZdata, respectively. In this figure, the original NRZ data has no splitspectrum line, whereas in the converted RZ data, evident spectrumlines are generated, spaced 10-GHz apart (the spectrum interval is0.08 nm). The experimental results clearly demonstrate the appli-cability of the scheme.

5. CONCLUSIONS

This paper has proposed an NRZ-to-PRZ converter consisting ofan SOA and an AWG. Through numerical simulation, the effectsof the parameters (including the power, wavelength, bit rate of theNRZ data, and the bandwidth of AWG and SOA current) on clockcomponent extraction are investigated, which is a useful guide forthe performance optimization. In the experiments, the clock-to-modulation ratio of the NRZ data at 10 Gb/s was enhanced morethan 15 dB using the proposed converter, which is beneficial forfurther clock recovery. All-optical clock recovery from NRZ dataat 10 Gb/s was successfully demonstrated with this converter anda mode-locked SOA fiber laser. Furthermore, using the recoveredclock as the control light of TOAD, NRZ-to-RZ format conversionat 10 Gb/s with extinction ratio of 10 dB was further realized,emphasizing the practical applicability of the proposed clock-recovery scheme for NRZ data. This scheme is a promisingmethod for high-speed clock recovery from NRZ data in futurecommunication networks.

ACKNOWLEDGMENT

This work is supported by National Natural Science Foundation ofChina (NSFC) program.

REFERENCES

1. Y.H. Li and G. Li, 80-Gb/s all-optical clock recovery using two-section gain-coupled DFB laser, Electron Lett 38 (2002), 892–893.

2. K. Vlachos, G. Theophilopoulos, A. Hatziefremidis, and H. Avramo-poulos, 30-Gb/s all-optical clock recovery circuit, IEEE Photon Tech-nol Lett 12 (2000), 705–707.

3. H.J. Lee, H.G. Kim, J.Y. Choi, and H.K. Lee, All-optical clockrecovery from NRZ data with simple NRZ-to-PRZ converter based onself-phase modulation of semiconductor optical amplifier, ElectronLett 35 (1999), 989–990.

4. W. Mao, M. Al-Mumin, X. Wang, and G. Li, All-optical enhancementof clock and clock-to-data suppression ratio of NRZ data, IEEE PhotonTechnol Lett 13 (2001), 239–241.

5. H.K. Lee, J.T. Ahn, M.Y. Jeon, K.H. Kim, D.S. Lim, and C.H. Lee,All-optical clock recovery from NRZ data of 10 Gb/s, IEEE PhotonTechnol Lett 11 (1999), 730–732.

6. W. Hong, D. Huang, F. Cai, and Y. Wang, Simultaneous clockcomponent extraction and wavelength conversion of NRZ signal usingan SOA loop mirror, IEEE Photon Technol Lett 16 (2004), 1116–1118.

7. G.P. Agrawal and N.A. Olsson, Self-phase modulation and spectralbroadening of optical pulses in semiconductor laser amplifiers, IEEE JQuant Electron 25 (1989), 2297–2306.

8. H. Wei, H. Dexiu, S. Junqiang, and L. Deming, Numerical simulationof recovery enhancement by a CW pump light in semiconductoroptical amplifiers, Opt Commun 214 (2002), 335–341.

9. A.E. Willner and W. Shieh, Optical spectral and power parameters for

all-optical wavelength shifting: single stage, fanout and cascadability,J Lightwave Technol 13 (1995), 771–781.

10. W. Mao, X. Wang, Al-Mumin, et al. All-optical clock extraction ofNRZ data using semiconductor optical amplifiers, CLEO2000, 2000,p. 130.

11. L. Yin, G. Liu, J. Wu, and J. Liu, Reduction of pattern effect forSOA-based clock recovery using CW assist light, Optical Eng (toappear).

12. M. Eiselt, W. Pieper, and H.G. Weber, SLALOM: Semiconductorlaser amplifier in a loop mirror, J Lightwave Technol 13 (1995),2099–2112.

© 2006 Wiley Periodicals, Inc.

A COMPACT, BROADBAND ANTENNAFOR PLANETARY SURFACE-TO-SURFACE WIRELESSCOMMUNICATIONS

Philip Barr, Afroz Zaman, and Felix MirandaAntenna, Microwave, and Optical Systems BranchNational Aeronautics and Space AdministrationGlenn Research CenterCleveland, OH 44135

Received 3 October 2005

ABSTRACT: The compact microstrip monopole antenna (CMMA) is anovel antenna design that combines a microstrip patch antenna with a3D structure to attain a highly directive, broadband, compact antenna.A tri-lobed patch (TLP) is designed to minimize the patch’s area whilereducing the antenna’s operating frequency. A grounding wall (GW)connects the patch to the ground plane and a vertical-enclosure wall(VEW) extends up, away from portions of the patch’s perimeter. This

Figure 1 Design technique used for laying out the TLP in the CMMA:(a) a depiction of the equilateral hexagon used to determine the location ofthe six uniform circles with 2.3-mm radius; (b) the darker region depictsthe area used to define the TLP from part a; (c) the TLP design used in theCMMA (this patch was fabricated on 1.57-mm Duroid� 5880 with arelative dielectric constant of �r � 2.2). [Color figure can be viewed in theonline issue, which is available at www.interscience.wiley.com.]

DOI 10.1002/mop MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 48, No. 3, March 2006 521

VEW supplies the antenna with a higher directivity in the radial direc-tion and also reduces the operating frequency. The CMMA was de-signed to operate at 2.23 GHz, but experimental results have shown thisantenna resonates at 2.05 GHz which is on the order of approximately�0/11.6 with respect to the antenna’s largest dimension, with a directiv-ity and bandwidth of 6.0 dBi and 130 MHz (6.3%), respectively. Thisminiature, radially emitting antenna makes the CMMA attractive forplanetary-based surface-to-surface communications. © 2006 Wiley Peri-odicals, Inc. Microwave Opt Technol Lett 48: 521–524, 2006;Published online in Wiley InterScience (www.interscience.wiley.com).DOI 10.1002/mop.21397

Key words: compact; broadband antennas; planetary surface-to-surfacecommunications; tri-lobed patch

Figure 2 A 3D diagram of the CMMA. The GW on the right connectingthe TLP to the ground plane is 1.57-mm high and W � 1.20 mm wide. Acoaxial probe is fed from the bottom through the substrate d � 0.75 mmfrom the GW. The VEW on the left extends H � 11.0 mm up (away fromthe TLP’s perimeter) while wrapping 31.2 mm of the perimeter oppositethe GW. [Color figure can be viewed in the online issue, which is availableat www.interscience.wiley.com.]

Figure 3 The CMMA prototype fabricated at NASA GRC. The VEW isshown on the right with the hole for the RF probe at the base of the CMMAon the left. [Color figure can be viewed in the online issue, which isavailable at www.interscience.wiley.com.]

Figure 4 Measured and simulated return losses for the CMMA. Theoperating frequency was measured at 2.05 GHz with a 130-MHz band-width ( ). The antenna was designed to operate at 2.23 GHz with abandwidth of 50 MHz (—). Also shown are the simulated data for thereturn loss of the CMMA taking into account the effects of the epoxy (�r �4) used to attach the VEW to the TLP (- - - - -)

Figure 5 Plots comparing the directivity of the measured and simulatedradial radiation patterns of the CMMA at 2.05 GHz: (a) E-plane radiationpattern; (b) H-plane radiation pattern

522 MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 48, No. 3, March 2006 DOI 10.1002/mop

1. INTRODUCTION

NASA’s vision for space exploration calls for extending the hu-man presence throughout the solar system through both human androbotic missions [1]. This vision will be realized with a return tothe moon by 2020 in preparation for future human missions toMars and beyond. As NASA begins exploring these destinations aflexible, reliable, and cost-effective wireless communications net-work will be needed. This wireless communications network willbe required to support various types of communications, includingvoice, data, imaging, and video, to and from an assortment ofmobile and nonmobile nodes. To achieve these goals, a concertedresearch effort into miniature antenna technologies is ongoing atthe NASA Glenn Research Center (GRC). The Compact Micros-trip Monopole Antenna (CMMA) presented in this paper is one ofthe outcomes of this ongoing effort.

The CMMA radiates like a miniature monopole antenna with aheight of less than �0/11.6, where �0 is the wavelength of theelectromagnetic signal in free space. This small size was achievedby attaching a Grounding Wall (GW) along a small section of the

perimeter of a Tri-Lobed Patch (TLP) [2]. Efforts by other groups(for example, the Dime Antenna [3, 4]) have shown that theoperating frequency of these grounded patches was determined bythe length of the ungrounded perimeter, plus the length of the GW.The TLP achieves a larger perimeter while maintaining a smallerarea in order to reduce the operating frequency while minimizingsize. Although this approach significantly reduces the antenna’soperating frequency, the directivity is greatly diminished. To ad-dress this problem, a Vertical Enclosure Wall (VEW) was attachedto portions of the TLP’s perimeter to increase the antenna’s direc-tivity and further reduce the antenna’s operating frequency (in ourcase, by as much as 500 MHz).

2. COMPACT MICROSTRIP MONOPOLE ANTENNA (CMMA)DESIGN AND FABRICATION

The CMMA was designed and simulated using Zealand’s IE3Delectromagnetic simulator to radiate at 2.23 GHz, a frequencywithin the band being considered for Lunar and Martian surface-to-surface communications [5, 6]. The TLP was designed on a

Figure 6 Average current distribution through the CMMA: (a) top view of the average current distribution (magnitude) through the TLP; (b) averagecurrent distribution (magnitude) through the VEW as seen along the line of sight of the plane containing the probe feed and parallel to the TLP; (c) averagecurrent distribution through the VEW looking at it from the side; (d) top view of the vector current distribution through the TLP. [Color figure can be viewedin the online issue, which is available at www.interscience.wiley.com.]

DOI 10.1002/mop MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 48, No. 3, March 2006 523

1.57-mm-thick substrate with a relative dielectric constant of �r �2.2 over a ground plane. A series of six connected circles locatedat the corners of an equilateral hexagon defines the layout of thispatch. The diameter of each circle is 4.60 mm, which is also equalto one side of the equilateral hexagon. Figure 1 illustrates thedesign technique used for the TLP. In Figure 2 the locations of theGW and the VEW are depicted. An approximately 1.20-mm-wideGW was attached from the perimeter of the patch to the groundplane, spanning the 1.57-mm thickness of the substrate. A VEW of11.0 mm in height and 31.2 mm in length was attached to theperimeter of the TLP and distributed evenly around the two lobes,which are directly opposite the lobe containing the GW.

The prototype shown in Figure 3 was fabricated at NASA GRCon 1.57-mm Duroid� 5880 with a relative dielectric constant of�r � 2.2. The copper surface of the Duroid� 5880 was etched topattern the TLP used in the CMMA. The GW and VEW werefabricated using copper tape. The adhesive from the copper tapewas removed using a ketone-based solvent. The 50-�m-thick wallswere shaped appropriately and attached to the patch using Able-bond, a conductive epoxy. A coaxial probe was inserted throughthe Duroid� 5880 to feed an RF signal from below the antennathrough the substrate to the TLP. This probe was attached 0.75 mmfrom the GW. The outer conductor of the coaxial probe wasremoved near the ground plane to prevent electrical shorting.

3. EXPERIMENTAL TESTING AND RESULTS

The CMMA was measured using an HP 8510C network analyzer,which showed an optimal S11 performance at 2.05 GHz. Themeasured difference in this operating frequency from simulationcould be a result of fabrication constraints. The relative dielectricconstant of the epoxy (which is nominally 4) may have been afactor contributing to the lowering of the operating frequency.Simulations that included this epoxy layer have shown a reductionin the operating frequency to what was observed in experiments.The 10-dB bandwidth of the fabricated antenna was approximately130 MHz (6.3%), which was larger than the simulated one of 50MHz. Figure 4 compares the measured return loss of the CMMAwith that of simulation. The broadening of the resonance is a resultof the nonoptimal matching and coupling losses between the RFprobe feed and the antenna.

Experimental measurements of the CMMA’s radiation patternswere performed in the cylindrical near-field antenna range atNASA GRC [7]. The measured radiation patterns for the E-planeand H-plane are shown in Figure 5. The radial radiation charac-teristic of the CMMA indicates that the dominant radiation mech-anism of the antenna is the surface current flowing on the VEW.This current is fed at the base of the wall and is distributed in amanner determined by the TLP. Note that this distribution is afunction of frequency and, hence, it is the TLP that controls thefrequency behavior of the antenna. The current flows vertically andmust terminate at the top edge of the wall, resulting in the radiatedfield. These currents can be viewed in the current distribution plotshown in Figure 6. The directivity of the CMMA prototype wascalculated from the radiation pattern to be approximately 6.0 dBi,which is comparable to the simulated directivity of 5.0 dBi and thequarter-wave monopole of 5.16 dBi [8].

4. CONCLUSION

With the CMMA’s unique combination of a VEW and a microstrippatch, this miniature antenna was found to retain its high directiv-ity and large bandwidth properties, as compared to typical minia-ture antennas which trade these properties to achieve their smallsizes. The measured results have shown that the CMMA is capable

of operating at less than �0/11.6 with no trade off in its directivity.The measured bandwidth and directivity for this antenna were 130MHz and 6.0 dBi, respectively. The CMMA has the potential formany applications where monopole antennas are typically used.Many mobile components in NASA’s future missions, such assensors, rovers, robots, and so forth, will require radially directedantennas, since the antenna’s orientation and position cannot bepredicted. The CMMA’s small size makes it more suitable forintegration than other larger monopole antennas. Future work forthis antenna will include an investigation into a linear array of theCMMA elements in order to produce a highly directive radiallyradiating antenna.

ACKNOWLEDGMENTS

The authors would like to acknowledge Dr. Kevin Lambert ofAnalex Corporation and Dr. Richard Lee of the NASA GlennResearch Center for their contributions to this project.

REFERENCES

1. National Aeronautics and Space Administration, The new age of ex-ploration. NASA’s direction for 2005 and beyond, 2005, http://www-.nasa.gov/pdf/107490main_FY06_Direction.pdf.

2. R. Waterhouse, Small microstrip patch antenna, IEE Electron Lett 31(1995), 604–605.

3. B.A. Cetiner, L. Jofre, F. Flaviis, N. Alexopoulos, and G.P. Li, Aminiature broadband antenna for portable communications terminals,IEEE Antennas Propagat Soc Int Symp Dig 4 (2001), 108–111.

4. B.A. Cetiner, L. Jofre, and F. Flaviis, Reconfigurable miniature mul-tielement antenna for wireless networking, IEEE Radio Wireless Conf,2001, pp. 203–206.

5. D.L. Brandel, Spectrum recommendations for lunar relay and lunarproximity links, Presentation to NASA Architecture Sub-Group forSpectrum, 2004.

6. Space Frequency Coordination Group, Frequency assignment guide-lines for communications in the Mars region, 2005, http://www.sfcgon-line.org/handbook/rec/rec22–1r1.doc.

7. K. Lambert, G. Anzic, R. Zakrajsek, and A. Zaman, Overview ofantenna measurement facilities at the NASA Glenn Research Center,NASA/TM—2002-211883, Proc Antenna Measurement Tech Assoc24th Ann Mtg and Symp, 2002, pp. 549–554.

8. W.L. Stutzman and G.A. Thiele, Antenna theory and design, 2nd ed.,Wiley, New York, 1998, p. 68.

© 2006 Wiley Periodicals, Inc.

CHARACTERIZATION OF A PLANARSELF-COMPLEMENTARY SQUARE-SPIRAL ANTENNA IN THE THz REGION

E. R. Brown,1 A. W. M. Lee,1 B. S. Navi,2 and J. E. Bjarnason1

1 University of CaliforniaSanta Barbara, CA 931062 Archer School for GirlsLos Angeles, CA 90049

Received 22 September 2005

ABSTRACT: This paper describes a compact, self-complementarysquare-spiral antenna on a GaAs substrate with a broadside high-direc-tivity (�29 dB) frequency-independent pattern when coupled through asilicon hyperhemisphere. The driving-point resistance undulates between�100 and 300� from 200 GHz to 1 THz—much higher than the 72�value from Booker’s modified formula, but quite beneficial for couplingto high-impedance broadband devices. © 2006 Wiley Periodicals, Inc.

524 MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 48, No. 3, March 2006 DOI 10.1002/mop