Performance analysis of optimized millimeter-wave fiber radio links

IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 54, NO. 2, FEBRUARY 2006 921

Performance Analysis of OptimizedMillimeter-Wave Fiber Radio Links

Teddy Kurniawan, Student Member, IEEE, Ampalavanapillai Nirmalathas, Senior Member, IEEE,Christina Lim, Member, IEEE, Dalma Novak, Senior Member, IEEE, and Rod Waterhouse, Senior Member, IEEE

Abstract—We present a comprehensive performance analysisof several optimized fiber radio distribution schemes for mil-limeter-wave radio services. The analysis includes the noise andnonlinear characteristics of the transmitter (Tx)–receiver (Rx)pair integrated with the analog optical link in the downlinktransmission of a given wavelength without optical amplification.Investigations are focused on four configurations of optimizedfiber radio links, which were derived by considering the bestperforming possible Tx–Rx configuration and specifications ofcommercially available devices to support multichannel subcar-rier multiplexed transmission. It was found that the nonlinearcharacteristics of the Mach–Zehnder modulator are the majorsource of performance degradation of the fiber radio links. Acomparison of RF-over-fiber and IF-over-fiber transport schemesalso shows that RF-over-fiber can yield 3-dB improvement inperformance compared to IF-over-fiber techniques.

Index Terms—Analog systems, noise, nonlinearities, opticalfiber, transceivers.

I. INTRODUCTION

B ROAD-BAND wireless access (BWA) has attracted con-siderable interest to deliver broad-band communications

to last mile users [1]–[3]. In addition, there has been rapiddevelopments in the deployment of optical metropolitan areanetworks (MANs) and optical access networks, driven by thelow loss of fiber and its high bandwidth capabilities [4]–[6]. Byexploiting high-bandwidth wavelength-division multiplexing(WDM) networks, an integrated fiber radio backbone network(known as fiber radio) could be developed to interconnectantenna base stations (BSs) with the central office. In thesesystems, the radio subcarriers will be placed on to a partic-ular wavelength channel to be delivered to the designated BSbefore being transmitted to users via the wireless link. TheBS will then perform at least three basic functions, which are:1) amplification; 2) multiplexing; and 3) demultiplexing, aswell as electronic-to-optical (E/O) and optical-to-electronic

Manuscript received March 29, 2005; revised September 14, 2005. Thiswork was supported by the Australian Research Council under the DiscoveryProject Scheme and by the Australian Research Council under DiscoveryProject DP0452223.

T. Kurniawan was with the Australian Photonics Cooperative ResearchCentre, Photonics Research Laboratory, Department of Electrical andElectronic Engineering, The University of Melbourne, Victoria, Vic. 3010,Australia. He is now with the Brain Science Institute, Swinburne University ofTechnology, Melbourne, Vic. 3122, Australia.

A. Nirmalathas, C. Lim, D. Novak, and R. Waterhouse are with theAustralian Photonics Cooperative Research Centre, Photonics ResearchLaboratory, Department of Electrical and Electronic Engineering,The University of Melbourne, Victoria, Vic. 3010, Australia (e-mail:[email protected]).

Digital Object Identifier 10.1109/TMTT.2005.863047

Fig. 1. Investigated architecture of the fiber radio system for downlinktransmission.

(O/E) conversions. Fig. 1 shows a higher level overview of thisfiber-radio system architecture.

There are several BWA systems currently under investigation.These include the local multipoint distribution system (LMDS),which operates within the 28–33-GHz band and can deliver datarates to users of up to 40 Mb/s occupying noise equivalent band-widths of 20 MHz when quadrature phase-shift keying (QPSK)modulation is used [7]–[9]. For an LMDS link having a 50-mcell radius, a line-of-sight (LOS) signal transmission will re-quire an RF signal power of 12 dBm and a carrier to interferenceand noise ratio (CINR) of 21 dB to achieve the target bit errorrate (BER) of 10 . This assumes a 4-dB loss in the BS and a3-dB NF at the user terminal [10]–[13].

In such analog optical link transmission, the system per-formance limitations will arise from the cumulative effectsof noise and distortions from device nonlinearities, as well asthe crosstalk arising from impairments in the optical network.Previous investigations of such effects have been carried out bya number of authors [14]–[18]. For example, Castleford et al.studied the effect of in-band and out-of-band optical crosstalkbetween WDM channels in a WDM fiber radio backbone net-work [14]. Mizuguti et al. investigated the effect of nonlinearcharacteristics of directly modulated laser diodes (LDs) onthe fiber radio link performance [15]. A detailed investigationof distortion effects caused by the use of a Mach–Zehndermodulator (MZM) in an externally modulated fiber radio linkwas reported by Way [16] and Cox [17]. Ackerman et al. laterestablished the maximum dynamic range (DR) and minimumlink insertion loss of an externally modulated analog opticallink [18]. Sabella [19] then extended the aforementionedinvestigations by undertaking a performance analysis of thefiber radio transmission system. This study considered the linkrequirements due to rain-induced effects on the transmissionof millimeter-wave signals, as well as different modulationformats.

0018-9480/$20.00 © 2006 IEEE

Authorized licensed use limited to: CHUNG ANG UNIVERSITY. Downloaded on July 7, 2009 at 05:40 from IEEE Xplore. Restrictions apply.

922 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 54, NO. 2, FEBRUARY 2006

TABLE IMATHEMATICAL TRANSFER FUNCTIONS USED TO DEFINE MATLAB-BASED MODELS

In this paper, we focus our studies on the effects of the noiseand intermodulation distortions in a fiber radio link to establish

the limitations on system performance. In addition, we use thedeveloped models to compare different configurations of fiber

Authorized licensed use limited to: CHUNG ANG UNIVERSITY. Downloaded on July 7, 2009 at 05:40 from IEEE Xplore. Restrictions apply.

KURNIAWAN et al.: PERFORMANCE ANALYSIS OF OPTIMIZED MILLIMETER-WAVE FIBER RADIO LINKS 923

radio subsystems comprising microwave/millimeter-wave sub-systems, as well as the analog fiber transmission link. This paperis organized as follows. In Section II, we describe the approachadopted to investigate fiber radio link performance through nu-merical simulations. This is then followed by Section III, whichdescribes in detail the fiber radio system configurations/archi-tectures considered and provides a comparison of their perfor-mance characteristics. In Section IV, we discuss the sensitivityof the key performance indicators to the variation of critical sub-system parameters. Finally, we present a summary of our inves-tigations in Section V.

II. INVESTIGATION METHODS

We have developed phenomenological numerical modelsof various subsystems and components of the fiber radio linkand the models were constructed using parameters that can bereadily measured or characterized. We first identified the elec-tronic and optical devices used in the complete transmissionlink of the fiber radio system. In the electronic/RF domain,these components comprise the amplifier, mixer, and bandpassfilter (BPF), while in the optical domain, they include the MZM,standard single-mode fiber (SMF), and the p-i-n photodiode(PD). The noise (through noise factor) and nonlinear propertiesof the RF amplifier, mixer, MZM, and the p-i-n PD weremathematically modeled and the models were verified throughexperiment. The simulation tools were able to be cascaded togive the performance of the entire link in MATLAB. We thenoptimized the Tx–Rx configurations considering the speci-fications of commercially available electronic devices, radiosystem design, and subsystem simulations in a multichannelenvironment. Afterwards, the obtained optimized transceiverconfigurations were incorporated into the analog optical link.In our investigation, a particular modulation scheme of op-tical single-sideband carrier (OSSB C) employing a dualelectrode mode MZM was chosen to eliminate the destructiveinterference occurred in the double sideband transmission dueto the chromatic dispersion effect [20].

A. Modeling

To model the performance of the fiber radio links, standardwell-established transfer functions/responses of the electronicamplifier [21], BPF [22], MZM [20], SMF [23], [24], and PD[25], [26] were used. We assume that the transfer functions de-scribe relationships between input and output signals and thesetransfer functions are time invariant. The transfer functions usedto define the linear and nonlinear models of different elementsare tabulated in Table I. As an example, the mixer model wasdeveloped independently to capture its nonlinear behavior in amultichannel environment when driven by the specified localoscillator (LO) power that is obtainable from the device specifi-cations. Mathematically, the mixer model defining the relation-ship between the input signal and the output signal isgiven in Table I. The desired output signals of a mixer comesfrom frequency components of the term . Apart from thedesired output, distortions will also appear at the mixer outputas a result of the nonlinear transfer function depicted in Table I.It needs to be highlighted that the power series equation that de-

Fig. 2. (a) Measured and (b) simulated two-tone test of a mixer.

scribes the mixer transfer function in fact comprises additionalhigher order terms. However, for time efficiency in running sim-ulation, and the fact that higher power series are negligible whenthe mixer is driven effectively at its LO power, the power seriesis limited up to the fourth order.

B. Verified Simulation Tools

An example of the verification of the simulation tools by ex-periment is given in Fig. 2. A two-tone experiment to confirmthe nonlinear transfer function of a mixer is performed with aZAM-42 mixer from MiniCircuits, Brooklyn, NY. It operatesin the 1.5–4.2-GHz band for the LO/RF port and dc–500 MHzat the IF port with a 7-dBm effective LO power. By applyinga two-tone input of equal power level at 100 and 105 MHz tothe IF port and by varying the power level, the conversion loss,RF–LO isolation, RF–IF isolation, third-order input interceptpoint (IIP3), and second-order input intercept point (IIP2) char-acteristics of 9.5, 18, and 33 dB and 9 and 16.5 dBm, respec-tively, were attained. When these values were fed to the devel-oped tools together with the carriers at a power of 10.5 dBm,the signal spectrum from the simulation shows strong correla-tion with the spectrum from the experiment, as shown in Fig. 2.For Fig. 2(a) and (b), the -axis represents the power in dBmand the -axis represents the frequency in megahertz.

III. PERFORMANCE ANALYSIS

We considered two signal transport schemes (RF-over-fiberand IF-over-fiber) and, for each transport scheme, we inves-

Authorized licensed use limited to: CHUNG ANG UNIVERSITY. Downloaded on July 7, 2009 at 05:40 from IEEE Xplore. Restrictions apply.

924 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 54, NO. 2, FEBRUARY 2006

Fig. 3. RF-1 configuration.

Fig. 4. RF-2 configuration.

Fig. 5. IF-1 configuration.

tigated many transmitter and receiver driver circuit configura-tions, selecting the best two for each transport technique. In de-ciding the optimal configurations of the transmitter driver cir-cuits, we chose one based entirely on achieving the transmis-sion performance target, while the other one was selected on thebasis of better performance at a lower cost, which offers perfor-mance at a lower cost. In contrast, the receiver circuit config-urations were based on the receiver performance and the po-tential for realizing simple and low power BSs. Figs. 3–6 showthe link configurations under investigation along with the cor-responding device or subsystem parameters. For the electronicdevices, we take into account the noise figure (NF), gain profile(G), or loss profile (L), 1-dB compression point , IIP3,and IIP2, as well as the RF–LO and RF–IF isolation of a mixer.

For the optical subsystems, the MZM switching voltageand loss (L), the SMF loss and its dispersion character-istic (D), the laser relative intensity noise (RIN), the p-i-n PDdark current , and thermal noise were considered, as wellas the nonlinear properties of the MZM and PD. The values ofthese parameters were obtained from the available commercialdevices matched with the link requirements to comply with thewireless scenario described earlier in Section I.

Table II summarizes the subsystems characteristics of theTx–Rx pairs for both RF- and IF-over-fiber schemes that wereused in the investigation. The input power and output power

of the transmitters and receivers, as well the inducedloss (L) and required gain (G) are numerically shown. From [8],we can expect that one BS is able to serve eight users in a cell.

Authorized licensed use limited to: CHUNG ANG UNIVERSITY. Downloaded on July 7, 2009 at 05:40 from IEEE Xplore. Restrictions apply.

KURNIAWAN et al.: PERFORMANCE ANALYSIS OF OPTIMIZED MILLIMETER-WAVE FIBER RADIO LINKS 925

Fig. 6. IF-2 configuration.

TABLE IICHARACTERISTICS OF TRANSMITTER AND RECEIVER PAIRS FOR

IF- AND RF-OVER-FIBER SCHEMES

TABLE IIICINR OBSERVATION AMONG THE TRANSMITTER, OPTICAL, AND RECEIVER

SUBSYSTEMS FROM THE FOUR CONFIGURATIONS

We assume that the signals entering the Tx have no ad-ditional noise other than that contributed by thermal noise( 174 dBm/Hz). Initially, we evaluated the three-tone exci-tation (three subcarriers of equal power) of the links, whichare described schematically in Figs. 3–6. The performancecharacteristics in terms of CINR evaluated for the three keysubsystem blocks (Tx, opt, and Rx) for RF- and IF-over-fibertransport schemes are summarized in Table III. It should benoted that the values presented in Table III correspond to themiddle of the three subcarriers used and represent the worstcase scenarios.

By comparing the results in Table III, it can be seen thatthe overall system performance is limited by the performanceof the optical subsystem block, primarily due to the nonlineartransfer function of the MZM used in the link. Furthermore, itcan be concluded that the RF-over-fiber scheme delivers betterthan 3-dB improvement in performance compared to that of theIF-over-fiber scheme. The reason for this is that, in the IF-over-fiber schemes, the signals received at the Rx are of low power,and undergo further signal-processing steps with mixers andamplifiers, which introduce additional nonlinearities.

TABLE IVSFDR CHARACTERISTICS

Fig. 7. Performance comparison of the investigated links for transmission ofthree subcarriers.

The spurious-free dynamic range (SFDR) as a function of thenumber of subcarriers (for , and ) for the four linksshown in Figs. 3–6 are given in Table IV. For a target SFDR of77 dB Hz [27], [28], it can be deduced that the system canonly support five or fewer subcarriers.

As the input power to the transmitter subsystem is varied, theoptical modulation index (OMI) per subcarrier within the MZMwill also vary. This, in turn, will cause a change in the achievableCINR for the various links. Fig. 7 shows the calculated CINRfor each carrier as a function of OMI in the links under consid-eration for a three-tone excitation . Figs. 8 and 9 showthe same relationship for the number of subcarriers being fiveand eight.

In Figs. 7–9, the dashed line corresponds to the target specifi-cation in terms of a 21-dB CINR requirement in regards with thewireless scenario explained in Section I. Comparing the SFDRvalues in Table IV and considering the results in Figs. 7–9,we can deduce that the fiber radio links are only feasible forchannel numbers of five of less. Furthermore, as can be clearly

Authorized licensed use limited to: CHUNG ANG UNIVERSITY. Downloaded on July 7, 2009 at 05:40 from IEEE Xplore. Restrictions apply.

926 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 54, NO. 2, FEBRUARY 2006

Fig. 8. Performance comparison of the investigated links for transmission offive subcarriers.

Fig. 9. Performance comparison of the investigated links for transmission.

seen from Figs. 7–9, it is evident that there exists an optimumvalue of OMI depending on the number of channels in the link.These optimum values arise from the interplay between the car-rier-to-noise ratio (CNR), which increases as the signal powerincreases, and the carrier-to-interference ratio (CIR), which de-creases as the signal power increases. In addition, the maximumCINR and optimum OMI both decrease as the number of sub-carriers in the link is increased.

IV. TOLERANCE OBSERVATION

Probably the most obvious means to improve the fiberradio system performance is to linearize the MZM. However,MZM linearization is typically complex to realize in practice[29]–[31]. Another possible approach to achieving an improve-ment in performance is to increase the IIP3 and IIP2 of theelectronic/RF devices. Tables V and VI show the effect onsystem performance (CINR), when the IIP3 and IIP2 valuesare increased by 3 dB at certain critical stages in the link,considering one stage at a time. The positions of each stagecan be seen in Figs. 3–6. All values in Tables V and VI are indecibels.

It can be observed from Tables V and VI that, for the RF-over-fiber scheme, considerable improvement in performance can beachieved by increasing the linear characteristics of the partic-ular device, which appears as the last stage in the receiver con-figuration. On the other hand, for the IF-over-fiber scheme, fur-ther improvement in the performance of the overall link can beachieved by increasing the linearity of the second stage withinthe receiver, namely, the mixer

TABLE VCINR OBSERVATION AT VARIOUS STAGES OF RF OVER FIBER

CONFIGURATIONS WHEN NONLINEAR PARAMETER OF

EACH STAGE IS ALTERED INDEPENDENTLY

TABLE VICINR OBSERVATION AT VARIOUS STAGES OF IF OVER FIBER CONFIGURATIONS

WHEN NONLINEAR PARAMETER OF EACH STAGE IS ALTERED INDEPENDENTLY

V. CONCLUSION

We have presented performance analyses of the completefiber radio transmission system for several millimeter-wavefiber radio distribution schemes. By taking into account noise,the nonlinear transfer function of all the devices in the link, aswell as possible signal transport schemes and transmitter/re-ceiver configurations with the specifications of commerciallyavailable devices, calculations of the link performance have beobtained.

For the first time, numerical results were presented in theanalysis of millimeter-wave fiber radio systems, which includethe microwave circuits, as well as the analog optical transmis-sion link. It was found that the MZM nonlinearity is the lim-iting factor in the overall link performance, both for RF- andIF-over-fiber schemes. In addition, the RF-over-fiber techniquesdeliver improved performance (up to 3 dB) compared to theIF-over-fiber signal transport scheme. Furthermore, for the par-ticular wireless application of LMDS described in this study,the investigated links are feasible only for delivering five sub-carriers in a fiber radio system based on frequency division mul-tiple access (FDMA).

The SFDR has also been presented and discussed, as wellas the existence of an optimum OMI, which depends on thenumber of subcarriers in various link configurations. We alsocarried out a tolerance analysis of the improvement in systemperformance that results when the IIP3 and IIP2 values of theelectronic devices employed in the link are increased. It wasfound that the receiver plays a significant role in improving thesystem performance compared to the transmitter.

REFERENCES

[1] A. Nirmalathas, C. Lim, D. Novak, and R. B. Waterhouse, “Progress inmillimeter-wave fiber-radio access networks,” Ann. Telecommun., vol.56, pp. 27–38, 2001.

[2] K. Biesecker, “The promise of broad-band wireless,” Inf. Technol. Pro-fessional, vol. 2, pp. 31–39, 2000.

[3] C. D. Gavrilovich, Jr., “Broadband communication on the highways oftomorrow,” IEEE Commun. Mag., vol. 39, pp. 146–154, 2001.

[4] L. G. Kazovsky, K. Shrikhande, I. M. White, M. Rogge, and D.Wonglumsom, “Optical metropolitan area networks,” presented at theOpt. Fiber Commun. Conf. and Exhibition, 2001.

Authorized licensed use limited to: CHUNG ANG UNIVERSITY. Downloaded on July 7, 2009 at 05:40 from IEEE Xplore. Restrictions apply.

KURNIAWAN et al.: PERFORMANCE ANALYSIS OF OPTIMIZED MILLIMETER-WAVE FIBER RADIO LINKS 927

[5] P. A. Humblet, “The direction of optical technology in the metro area,”presented at the Opt. Fiber Commun. Conf. and Exhibition, 2001.

[6] I. Tomkos, “Metropolitan area optical networks,” IEEE Circuits DevicesMag., vol. 19, pp. 24–30, 2003.

[7] A. Nordbotten, “LMDS systems and their application,” IEEE Commun.Mag., vol. 38, pp. 150–154, 2000.

[8] LMDS Cell Sizing and Availability, IEEE Standard 802.16sc-99/17,1999.

[9] K. H. Park, D. K. Shin, J. S. Lee, and M. H. Sunwoo, “Design of aQPSK/16 QAM LMDS downstream receiver ASIC chip,” in IEEESignal Process. Syst. Des. Implementation Workshop, Piscataway, NJ,2000, pp. 210–217.

[10] L. W. Couch II, Digital and Analog Communication System. Engle-wood Cliffs, NJ: Prentice-Hall, 2001.

[11] L. E. Larson, Ed., RF and Microwave Circuit Design for Wireless Com-munication. Norwood, MA: Artech House, 1996.

[12] H. Kawamura, N. Imai, E. Ogawa, and H. Inomata, “High-speed datatransmission using millimeter-wave fiber-optic links,” IEICE Trans.Commun., vol. E79-B, pp. 1784–91, 1996.

[13] R. L. Freeman, Radio System Design For Telecommunication System(1–100 GHz). New York: Wiley, 1987.

[14] D. Castleford, A. Nirmalathas, D. Novak, and R. S. Tucker, “Opticalcrosstalk in fiber-radio WDM networks,” IEEE Trans. Microw. TheoryTech., vol. 49, no. 10, pp. 2030–2035, Oct. 2001.

[15] H. Mizuguti, T. Okuno, S. Komaki, and N. Morinaga, “Performanceanalysis of optical fiber link for microcellular mobile communicationsystems,” IEICE Trans. Electron., vol. E76-C, pp. 271–278, 1993.

[16] W. I. Way, “Optical fiber-based microcellular systems: An overview,”IEICE Trans. Commun., vol. E76-B, pp. 1091–102, 1993.

[17] C. H. Cox III, “High-performance fiber-optic links for microwave appli-cations,” in IEEE MTT-S Int. Microw. Symp. Dig., 1993, pp. 719–722.

[18] E. Ackerman, S. Wanuga, D. Kasemset, A. S. Daryoush, and N. R.Samant, “Maximum dynamic range operation of a microwave externalmodulation fiber-optic link,” IEEE Trans. Microw. Theory Tech., vol. 41,no. 8, pp. 1299–1306, Aug. 1993.

[19] R. Sabella, “Performance analysis of wireless broadband systems em-ploying optical fiber links,” IEEE Trans. Commun., vol. 47, no. 5, pp.715–721, May 1999.

[20] G. H. Smith, D. Novak, and Z. Ahmed, “Overcoming chromatic-dis-persion effects in fiber-wireless systems incorporating external modula-tors,” IEEE Trans. Microw. Theory Tech., vol. 45, no. 8, pp. 1410–1415,Aug. 1997.

[21] B. Razavi, RF Microelectronics. Englewood Cliffs, NJ: Prentice-Hall,1998.

[22] L. C. Ludeman, Fundamental of Digital Signal Processing. New York:Harper & Row, 1986.

[23] Z. Ahmed, D. Novak, R. B. Waterhouse, and L. Hai-Fen, “37-GHz fiber-wireless system for distribution of broad-band signals,” IEEE Trans. Mi-crow. Theory Tech., vol. 45, no. 8, pp. 1431–1435, Aug. 1997.

[24] D. K. Mynbaev and L. L. Scheiner, Fiber-Optic Communications Tech-nology. Englewood Cliffs, NJ: Prentice-Hall, 2001.

[25] S. B. Alexander, Optical Communication Receiver Design. London,U.K.: IEE Press, 1997.

[26] K. J. Williams, R. D. Esman, and M. Dagenais, “Nonlinearities in p-i-nmicrowave photodetectors,” J. Lightw. Technol., vol. 14, no. 1, pp.84–96, Jan. 1996.

[27] J. C. Fan, C. L. Lu, and L. G. Kazovsky, “Dynamic range require-ments for microcellular personal communication systems using analogfiber-optic links,” IEEE Trans. Microw. Theory Tech., vol. 45, no. 8, pp.1390–1397, Aug. 1997.

[28] W. Jemison et al., Fiber Radio Using a Mode-Locked MicrochipOptical Transmitter—System Considerations and Performance Eval-uation. Philadelphia, PA: CMLE Publications, Drexel Univ., 2001.[Online]. Available: http://www.ece.drexel.edu/CMLE/papers.

[29] E. I. Ackerman, “Broad-band linearization of a Mach–Zehnder elec-trooptic modulator,” IEEE Trans. Microw. Theory Tech., vol. 47, no. 12,pp. 2271–2279, Dec. 1999.

[30] W. B. Bridges and J. H. Schaffner, “Distortion in linearized electroopticmodulators,” IEEE Trans. Microw. Theory Tech., vol. 43, no. 9, pp.2184–2197, Sep. 1995.

[31] U. V. Cummings and W. B. Bridges, “Bandwidth of linearized elec-trooptic modulators,” J. Lightw. Technol., vol. 16, no. 8, pp. 1482–1490,Aug. 1998.

Teddy Kurniawan (S’02) was born in Jakarta,Indonesia, in 1979. He received the B.E. degree(first-class honors) in electrical engineering fromAtma Jaya Catholic University, Jakarta, Indonesia,in 2001, the M.Eng.Sc. degree in electrical andelectronic engineering from The University of Mel-bourne, Victoria, Australia, in 2004, and is currentlyworking toward the Ph.D. degree at the SwinburneUniversity of Technology, Melbourne, Victoria,Australia.

His research interests are in the area of device non-linearities of microwave photonics, performance and architecture of fiber radiosystem, and dosimetry studies due to RF radiation.

Ampalavanapillai (Thas) Nirmalathas (S’96–M’97–SM’03) received the B.E. (with honors) inelectrical and electronic engineering and Ph.D.degree from The University of Melbourne, Victoria,Australia, in 1993 and 1997, respectively.

In 1997, he joined the Photonics Research Labo-ratory (PRL), The University of Melbourne, wherehe was a Research Fellow, Senior Research Fellow,and Senior Lecturer. He is currently an Associate Pro-fessor and Reader with the Department of Electricaland Electronic Engineering, The University of Mel-

bourne. He has also held other positions such as the Director of PRL and the Pro-gram Manager of the Telecommunications Technologies Research Program ofthe Australian Photonics Cooperative Research Centre (APCRC) (2001–2004).In 2004, he was a Guest Researcher of the Ultra-Fast Photonic Network Group,National Institute of Information and Communication Technology (NICT), Ko-ganei, Japan, and a Visiting Scientist with the Lightwave Department, Insti-tute for Infocom Research (I2R), Singapore. He is currently a contributed staffmember with the National ICT Australia (NICTA) and the Program Leader forthe Network Technologies Program of the NICTA. He has authored or coau-thored over 130 technical papers and has given a number of invited presenta-tions at leading international conferences. He holds two international patentsand one provisional patent. His current research interests include fiber-wirelessnetworks, optical access networks, optical signal monitoring, photonic packetswitching technologies, ultra-fast optical communications systems, and appli-cations of mode-locked semiconductor lasers.

Dr. Nirmalathas is one of the representatives on the Steering Committee ofthe IEEE Lasers and Electro-Optics Society (CLEO) Pacific Rim Conferenceand a member of the Steering Committee of the International Conference onOptical Internet (COIN). He has been a member of committees associated witha number of international conferences in his field of expertise.

Christina Lim (S’98–M’00) received the B.E. (withfirst-class honors) and Ph.D. degrees in electricaland electronic engineering from The University ofMelbourne, Victoria, Australia, in 1995 and 2000,respectively.

In 1999, she joined the Photonics Research Labo-ratory (PRL), a member of the Australian PhotonicsCooperative Research Centre (APCRC), Universityof Melbourne, where she is currently a Senior Re-search Fellow. Her research interests include fiber-wireless access technology, modeling of optical and

wireless communication systems, microwave photonics, application of mode-locked lasers, optical network architectures, and optical signal monitoring.

Dr. Lim was one of the recipients of the 1999 IEEE Lasers & Electro-OpticsSociety (IEEE LEOS) Graduate Student Fellowship.

Authorized licensed use limited to: CHUNG ANG UNIVERSITY. Downloaded on July 7, 2009 at 05:40 from IEEE Xplore. Restrictions apply.

928 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 54, NO. 2, FEBRUARY 2006

Dalma Novak (S’90–M’91–SM’02) received theB.E. (with first-class honors) and Ph.D. degreesin electrical engineering from the University ofQueensland, Brisbane, Qld., Australia, in 1987 and1992, respectively. Her doctoral thesis concerned thedynamic behavior of directly modulated semicon-ductor lasers.

From January 1992 to August 1992, she wasa Lecturer with the Department of Electrical andComputer Engineering, University of Queensland.In September 1992, she joined the Photonics

Research Laboratory (PRL), Australian Photonics Cooperative ResearchCentre (APCRC), Department of Electrical and Electronic Engineering, TheUniversity of Melbourne, Victoria, Australia, where her responsibilities overthe years have included Deputy Director and Research Director of the PRL,CRC Key Researcher, Director of the CRC Melbourne Division, and CRCEducation Director. From December 1999 to March 2001, she was a Directorof Australian Photonics Pty Ltd. From July 2000 to January 2001, she was aVisiting Researcher with the Department of Electrical Engineering, Universityof California at Los Angeles (UCLA), and with the Naval Research Laboratory,Washington, DC. In June 2001, she joined Dorsál Networks Inc., which laterbecame part of the Corvis Corporation, Columbia, MD, where she was aTechnical Section Lead until 2003. She is currently a Professorial Fellowwith The University of Melbourne. She has authored or coauthored over 160papers. Her research interests include fiber-radio communication systems andhigh-speed opto-electronic devices and circuits.

Rod Waterhouse (S’90–M’92–SM’02) received theB.E. (Hons.), M.Eng.Sc. (Research), and Ph.D. de-grees from the University of Queensland, Brisbane,Australia, in 1987, 1990, and 1994, respectively.

In 1994, he joined the School of Electrical andComputer Engineering, RMIT University, as aLecturer, and became a Senior Lecturer in 1997.From mid–2000 to the beginning of 2001, he was aVisiting Professor with the Department of Electricaland Computer Engineering, University of Californiaat Los Angeles (UCLA) for three months and then

a Visiting Researcher with the Photonics Technology Branch, Naval ResearchLaboratories, Washington, DC, for three months while on sabbatical. In June2001, he took a leave of absence from RMIT and joined Dorsal Networks,which later became part of the Corvis Corporation, Columbia, MD. Since2004, he has been a Senior Fellow with The University of Melbourne, Victoria,Australia. He has authored or coauthored over 150 papers. He holds threepatents. His research interests include optically distributed wireless systems,photonic devices, optical systems, and wireless communication technologies.

Authorized licensed use limited to: CHUNG ANG UNIVERSITY. Downloaded on July 7, 2009 at 05:40 from IEEE Xplore. Restrictions apply.