Performance Evaluation of Next Generation Free-Space Optical

IEICE TRANS. ELECTRON., VOL.E90–C, NO.2 FEBRUARY 2007381

PAPER Special Section on Evolution of Microwave and Millimeter-Wave Photonics Technology

Performance Evaluation of Next Generation Free-Space OpticalCommunication System

Kamugisha KAZAURA†a), Member, Kazunori OMAE†, Student Member, Toshiji SUZUKI†,Mitsuji MATSUMOTO†, Edward MUTAFUNGWA††, Members, Tadaaki MURAKAMI†††, Nonmember,

Koichi TAKAHASHI†††, Hideki MATSUMOTO†††, Kazuhiko WAKAMORI†††,and Yoshinori ARIMOTO††††, Members

SUMMARY Free-space optical communication systems can providehigh-speed, improved capacity, cost effective and easy to deploy wirelessnetworks. Experimental investigation on the next generation free-space op-tical (FSO) communication system utilizing seamless connection of free-space and optical fiber links is presented. A compact antenna which utilizesa miniature fine positioning mirror (FPM) for high-speed beam control andsteering is described. The effect of atmospheric turbulence on the beamangle-of-arrival (AOA) fluctuations is shown. The FPM is able to miti-gate the power fluctuations at the fiber coupling port caused by this beamangle-of-arrival fluctuations. Experimental results of the FSO system ca-pable of offering stable performance in terms of measured bit-error-rate(BER) showing error free transmission at 2.5 Gbps over extended periodof time and improved fiber received power are presented. Also presentedare performance results showing stable operation when increasing the FSOcommunication system data rate from 2.5 Gbps to 10 Gbps as well as WDMexperiments.key words: free-space optical (FSO) communication, fine positioning mir-ror (FPM), scintillation, angle-of-arrival (AOA) fluctuations, atmosphericturbulence

1. Introduction

The increase demand of wireless links which are easier,faster and less expensive to deploy has renewed interest inthe use of free-space optics in digital transmission of signalin the atmosphere [1]–[6].

Recently, free-space optical (FSO) communication sys-tems have become low cost, simple and easy to install,and are therefore increasingly deployed to offer high-speed, broader bandwidth communication links. Opticalwireless communication systems (just like microwave andmillimetre-wave wireless communication systems) can eas-ily provide high-speed communications without the diffi-culty and cost of deploying high-capacity optical fiber ca-bles [7], [8].

Manuscript received June 5, 2006.Manuscript revised September 28, 2006.†The authors are with the Global Information and Telecommu-

nication Institute (GITI), Waseda University, Honjo-shi, 367-0035Japan.††The author is with the Helsinki University of Technology, Fin-

land.†††The authors are with the Advanced Info Communication Pro-

motion Community, Tokyo, 164-8512 Japan.††††The author is with the National Institute of Information

and Communication Technology (NICT), Koganei-shi, 184-8795Japan.

a) E-mail: [email protected]: 10.1093/ietele/e90–c.2.381

Currently, widely deployed FSO systems use 0.8 µmwavelength band and communication systems which areable to provide up to 1.5 Gbps are in practical use. How-ever, the use of 0.8 µm wavelength optical devices makesthe FSO system incompatible with most of the current highcapacity optical fiber systems. Therefore in order to over-come such technical barriers, devices and components de-veloped for long-haul optical fiber communication are ef-fectively utilized to achieve high-speed, improved capacityFSO communications system [9], [10]. In optical fiber com-munication using 1.55 µm band, wavelength division multi-plexing (WDM) technology is possible with eye safe limitsthus making it a suitable operating wavelength for FSO datalinks [11].

Because optical fiber communication devices and com-ponents are generally designed for operation with singlemode fiber (SMF), when compared to traditional opticalwireless system devices there is little or no difference in theirdesign. In reference [9] by applying multimode fiber a sim-plified optical system technique for coupling a free-spaceoptical beam into a fiber is presented. Unfortunately, mul-timode fibers are not widely deployed in optical fiber com-munication networks thus the availability of such devices islimited.

In this paper we present an experimental investigationof a next generation FSO utilizing seamless connection of afree-space optical beam to an SMF. The viability of the com-munication system using an optical antenna with a miniaturefine positioning mirror (FPM) for high-speed beam trackingand AOA fluctuation compensation is evaluated. This sys-tem was deployed in Tokyo for communication experimentsover a 1 km distance. Results of the experiment show thatby using an antenna with a high-speed beam tracking andcontrol unit, the system is able to reduce the intensity fluctu-ations of the received optical beam coupled to the SMF. Fur-thermore, we verified that a stable connection was achievedand an effective seamless free-space and fiber FSO systemwas realized.

Section 2 gives an outline of seamless connection offree-space and fiber system and Sect. 3 describes the devel-oped optical antenna used in our experimental performanceevaluation of next generation FSO communication system.In Sect. 4 the experimental results as well analysis are dis-cussed and in Sect. 5 the paper is concluded.

Copyright c© 2007 The Institute of Electronics, Information and Communication Engineers

382IEICE TRANS. ELECTRON., VOL.E90–C, NO.2 FEBRUARY 2007

2. Seamless Connection of Free-Space and Fiber Sys-tem

In traditional FSO systems a fiber transceiver converts theelectrical signal to optical signal. The electrical signal isamplified by a laser driver which provides enough current todrive the laser diode. Modulated light from the laser diodeis directed through space to the receiver, which focuses thebeam onto an Si APD. The APD converts the optical signalto an electrical signal. After noise filtering and reshapingthe electrical signal is converted at the fiber transceiver backto an optical signal. This process is depicted in Fig. 1(a).FSO communications systems which operate in this fashioncan transmit data rates up to 1.5 Gbps. They can not operateabove this data rate because of the power and bandwidthlimitation of the optical devices [12].

To overcome the above-mentioned limitations, technol-ogy originally designed for long haul fiber optics transmis-sion operating in 1.31 µm or 1.55 µm as transmitting wave-lengths is utilized. Unfortunately, the 1.31 µm wavelength isa poor choice for FSO transmission because of the high at-mospheric absorption near the 1.31 µm from water vapour.There is relatively less absorption for FSO transmission atthe 1.5 µm wavelength. This wavelength is appropriate forseamless connection of free-space and fiber system.

In seamless connection of free-space and fiber systemsan optical beam is emitted directly from a fiber (SMF) ter-mination to free-space using an optical antenna. At the re-ceiver, the transmitted optical beam is focused directly toa fiber and then sent down the fiber for detection. Thisis shown in Fig. 1(b). However, depending on the deploy-ment environment, the optical signal’s transmitted powermay not be sufficient for free-space transmission. Thereforethe signal power is boosted by an erbium doped fiber ampli-fier (EDFA) and the resulting high-powered optical signalcan be transferred from the SMF termination to free-spaceas shown in Fig. 1(b). In this method the need to convertthe optical signal from electrical to optical formats or vice

Fig. 1 FSO system using (a) O/E and E/O conversion and (b) seamlessconnection of FSO beam and single mode fiber.

versa for transmitting or receiving through space is elimi-nated. Furthermore, this system is protocol transparent, theneed for reconfiguration of the transmitting antenna is elimi-nated even when the nature of the transmitted signal changesdue to varying bit-rate, signal format (analogue or digital) orwavelength channel [13]. Since fiber and free-space opticaltransmission links carry the same optical signal, the schemecan utilize mature technologies and optical components de-veloped for high bit-rate fiber transmission.

Essentially, by omitting the propagation path turbu-lence factor in the atmosphere, a free-space transmissionchannel can be considered to be approximately equivalentto that of an SMF. Previously, FSO systems utilizing a free-space optical beam coupled to an SMF for communica-tion between orbiting satellites have been proposed in [14]–[16]. By seamlessly interfacing fiber and free-space chan-nels a hybrid optical transmission scheme that enables afiber transmitted optical signal to be emitted directly intofree-space is achieved as described previously. In the re-verse direction the free-space transmitted optical signal isfocused onto the fiber by using beam size converters, suchas lenses described in reference [13].

However, the optical beam transmitted through the at-mosphere requires a large aperture lens because of the hugebeam diameter at the receiver. It is difficult to focus the op-tical signal into an SMF which has a core diameter of about10 µm. Furthermore, the beam experiences atmospheric tur-bulence as it propagates through the atmosphere, as wellas vibrations of the device at the installation site and beamdistortion occurrence. The consequence of these effects isthe fluctuation of the beam angle-of-arrival (AOA) which inturn leads to significant variation in the power of the lightfocused into the SMF. It is therefore difficult to maintainan error free communications link due to the received sig-nal power occasionally dropping below the receiver sensi-tivity. As a result of these implementation difficulties, untilrecently, FSO systems have been restricted to research pub-lication and theoretical verification experiments [13], [16],[17].

In practical FSO communication systems various tech-niques have been developed or are applied to mitigate atmo-spheric effects such as scintillation or beam wander. Thesetechniques include adaptive optics (AO), use of large re-ceive apertures, diversity techniques and fast tracking anten-nas [18]. Adaptive optics techniques, originally developedfor atmospheric compensation in astronomical sites, restoresthe distorted wave-front to its original state before it wasdestroyed by atmospheric turbulence. Although AO haveshown limited success, they require bulky and computation-intensive systems to achieve wave-front sensing and correc-tion. Alternatively, the use of large receive apertures foratmospheric turbulence mitigation requires the antenna tele-scope to be equally large. The use of diversity techniquesincreases the likelihood that the detected signal will be readcorrectly by propagating the optical wave-front in at leasttwo distinct ways. Diversity can occur in the form of spatialdiversity (requiring multiple transmitters and/or receivers),

KAZAURA et al.: PERFORMANCE EVALUATION383

Fig. 2 Experimental FSO communication system setup.

temporal diversity (requiring signal to be transmitted twice,separated by a time delay) and wavelength diversity (requir-ing the transmission of data on at least two distinct wave-lengths). Even though diversity techniques are promising,they do require significant electronic overhead in the retim-ing and synchronization process [19].

It is also possible to improve the reliability of FSOsystem by utilizing coding schemes used in RF and wiredcommunication systems. Although coding provides an addi-tional layer of information security, studies have shown thateven the best Forward Error Correction (FEC) codes cannotnegate the effect of atmospheric turbulence alone.

To compensate for atmospheric turbulence induced sig-nal fading, the antenna used in our experiment utilizes aminiature fine positioning mirror to control and steer the re-ceived beam to the SMF connection port. The major incen-tive of incorporation of an FPM is the active prevention oflong term data loss by compensating or at least mitigating at-mospheric turbulence induced wave-front phase distortion.Compared to the techniques outlined above, the merit of theFPM employed in our antenna is that it manages to improvethe FSO system performance with less complexity and min-imum electronics overhead while maintaining the compactsize of the antenna.

The basic configuration of our experimental setup forseamless free-space and optical fiber transmission investiga-tion is shown in Fig. 2. Because the light is fiber coupled atboth ends, the EDFA receive electronics and other measure-ment and data collection devices can be conveniently placedinside the building. The fibers are run to the respectiverooftops and then coupled directly to the transceiver as de-picted in Fig. 2. The experimental hardware setup includingdata collection and other measurement devices are placedin the experiment room as shown in Fig. 3. The collecteddata include weather data (visibility, temperature, precipita-tion and fog), bit error rate (BER) and the optical receivedpower. Also placed in the experimental room is a PC forremote antenna monitoring and control.

The challenge in seamless connection of free-space andSMF systems is not only to design an effective beam track-

Fig. 3 Experimental hardware setup.

ing and antenna alignment technique, but also an efficientmechanism for focusing the light into the fiber at the re-ceiver. An active tracking is required to control and steerthe received light to the SMF. This is described in more de-tail in the next section.

3. Optical Antenna Description

The narrow transmission beam of a free-space optical signalmakes alignment of FSO communication terminals difficultcompared to the wider beam RF systems. FSO systems arefaced with the challenge of designing effective pointing andtracking mechanism that must keep the receiver aligned withthe transmitter. In this experiment a compact optical antennaspecifically designed to address this challenge is used. Theinternal structure of the optical antenna is shown in Fig. 4.The antenna uses a Cassegrain type telescope configurationdesigned using 3 free form surface (FFS) optics consistingof primary mirror, secondary mirror and a collimating mir-ror [20]. A 1.55 µm wavelength beam is used for data recep-tion and transmission offering full-duplex (simultaneous bi-directional) data at gigabit-per-second rates. A 0.98 µm bea-con is used for antenna alignment and tracking purpose. Thebeacon light is emitted from four output windows placed onthe four corners in front of the antenna. The antenna utilizesan automatic active tracking mechanism with a CCD cam-era for rough tracking or initial alignment and a quadrantdetector (QD) for accurate tracking. Coupling of the trans-mitted optical beam directly to the SMF is accomplished bya miniature fine positioning mirror. The FPM, which has avital function in fine tracking process, is placed at the fiberradiation pupil position vicinity. The information of the ar-rival beam fluctuations is provided by the QD in order toalways lead the horizontal optical axis to the fiber connec-tion port thus achieving a feedback control setup. The FPMantenna tracking speed is selected to be able to mitigate theeffects of random atmospheric turbulences on the receivedbeam and steer most of the received optical signal to theSMF.

By combining two tracking methods as outlined abovea more effective beam tracking and control is achieved. The


Fig. 4 Optical antenna internal structure.

CCD beam position detector and digital signal processing(DSP) performs actual beam position and analysis and feedthe information back to the system for counter measures.The QD used for fine tracking consists of four separate de-tector elements arranged in a matrix. Part of the 0.98 µmbeacon beam is directed to the QD by a beam splitter. Thefour elements of the QD collect the beam separately, andwhen the signal output for all the detectors is the same thenthe spot is located exactly at the middle of the detector ar-ray. If the beam spot moves, the amount of beam collectedby each different detector will be different, resulting in vari-ation of the level of the output signal. By analyzing andcomparing these four output signals the direction of the spotmovement on the detector array is determined and a correc-tive control signal is sent to drive the FPM mirror actuatorto control and steer the received 1.55 µm data signal to theSMF.

4. Experimental Results and Analysis

In order to establish the optimum antenna’s FPM trackingspeed, we first determine, approximately, the magnitude ofthe atmospheric turbulence in our deployment environment.This is done by examining the relation and correlation be-tween intensity fluctuation caused by the influence of atmo-spheric turbulence and the 1.55 µm beam intensity variationsas a result of AOA fluctuations. By doing so we can quan-tify, to some extent, the magnitude of atmospheric turbu-lence experienced by the propagating optical beam.

Figure 5 top shows the measured beam intensity fluc-tuation in terms of detected voltage level as a result of scin-tillation effect and Fig. 5 below shows the 1.55 µm commu-nication beam intensity variation because of AOA fluctua-tions caused by atmospheric turbulence. It is rather difficult,if not impossible, to measure the beam wave-front phasechanges during strong intensity scintillations in the receiveraperture [21]. Thus the AOA fluctuations are measured in

Fig. 5 Beam intensity fluctuation (top) caused by scintillation effects and(below) 1550 nm communication beam intensity variations as a result ofAOA fluctuation.

Fig. 6 Power spectra showing the relationship between intensity fluctu-ation caused by scintillation and 1550 nm communication beam variationscaused by AOA fluctuations.

terms of the detected electrical signal level (in Volts) whenthe antenna tracking was set to OFF. The measured data isrecorded after every 5 minutes. The 5 minutes periods aresplit into blocks of 3 seconds wherein the sampling rate is10 kHz.

By producing the power spectra of the time series datashown in Fig. 5 we can obtain the frequency characteristicsof the data and therefore are able to establish the similarityin the characteristics of the beam intensity fluctuation as aresult of scintillation and the beam AOA fluctuations. Fig-ure 6 depicts the frequency characteristics of the beam in-tensity fluctuations and AOA fluctuations. From the figureit can be observed that the AOA fluctuations of the 1.55 µmdata beam for frequency above 100 Hz are closely correlatedto scintillation variations observed on the 0.8 µm scintilla-tion measurement beam. It should be noted that the inten-sity variation as a result of scintillation is measured by anantenna installed on the same site thus having almost thesame propagation path as the antenna under test. Therefore,it can be correctly assumed that the atmospheric turbulence


Fig. 7 Fiber received power—(top) antenna tracking set to OFF and (bot-tom) antenna tracking set to 1 kHz.

will influence the AOA fluctuation of the 1.55 µm commu-nication beam in a similar way to the scintillation effect ob-served on the 0.8 µm wavelength beam.

From the above the FPM tracking speed which willenable the mirror to steer and focus to a great extent thereceived optical beam to the SMF can be closely approx-imated. From the experimental results and observation ofbeam AOA fluctuations frequency characteristics (obtainedby producing the power spectra as shown in Fig. 6) theselected FPM antenna tracking speed value should be setabove 500 Hz. In our case we set the tracking speed to 1 kHzwhich is currently considered to be the fastest tracking sys-tem for any available FSO communication system.

The improvement in the fiber received power is de-picted in Fig. 7. The figure shows the result for accuratetracking by controlling the beam AOA fluctuations. Figure 7top shows the fiber received power when the antenna FPMtracking is set to OFF while Fig. 7 bottom shows fiber re-ceived power when the antenna FPM tracking speed is set to1 kHz. It is observed that the high-speed tracking capabilityof the FPM manages to control and steer most of the re-ceived light to the SMF. The intensity fluctuations observedwhen the antenna FPM tracking is set to OFF are remarkablysuppressed and improvement in the fiber received power isrealized as depicted in Fig. 7 bottom.

To evaluate the communication quality of the systema 2.5 Gbps bit error rate tester (BERT) pattern generator isused to directly modulate (on-off keying) a single frequencydistributed feedback (DFB) laser at 1.55 µm wavelength,with a 2−23 − 1 pseudo random bit sequence (PRBS) pat-tern length. The data encoded optical signal, is amplified bya 100 mW EDFA. At the receiver the received optical beamis focused by the FPM to the SMF.

The optical circulator is used to isolate the transmittedand the received signals. At the reception side, the receivedsignal is equally split into two arms by a passive 3 dB cou-pler. Half of the received signal enters the O/E converter andthe data and clock signals extracted for BER measurementsand the other half of the received signal is used to monitor

Table 1 Selected specifications of equipment used in the experiment.

Parameter Specification Unit Value

Data rate Gbit/s 2.5E/O (directly mod. DFB laser) - OOK/NRZ

Test data pattern - 223− 1 PRBSBoost EDFA output mW 100Receiver optical filter 0.5 dB BW nm ± 11Antenna aperture mm 40

Rec. sensitivity (BER = 10−12) dBm −30E/O - Electrical/Optical, OOK/NRZ - On Off Keying withnon return to zero, PRBS - pseudo random bit sequence

Fig. 8 Bit error rate and fiber received power characteristics.

the received optical power (this setup is depicted in Fig. 2).The 3 dB output is coupled directly to an optical power me-ter with a 100 msec averaging time and the optical powermeter data is averaged and logged every 30 secs. Some ofthe primary specifications of the FSO communication de-vices used in the experiment are listed in Table 1.

The regenerated data and clock signals derived fromthe 2.5 Gbps optical signal, were applied to the 2.5 Gbpserror detector for link performance monitoring. The BERcharacteristics when the antenna FPM tracking speed is setto 1 kHz is shown in Fig. 8. For a 24 hour period, error freetransmission is achieved which confirms the stability andgood performance of the system. For this system the mini-mum back-to-back error-free received power is −30 dBm.

Even though the antenna tracking speed is enough tocompensate for atmospheric turbulence effects, occasionalburst errors occur as shown in Fig. 8. These burst errors areattributed to perhaps the non-linearity in the tracking sys-tem or the tracking dynamic range might be insufficient insituations of strong atmospheric turbulence. Another likelycause of the burst errors is phase scintillation due to atmo-spheric turbulence of the incoming signal beam resulting inthe decrease of the received power. These occasional bursterrors hardly have any influence on the overall performanceof the FSO system and the system can be used for reliablestable communication.

Consecutively by using a digital communication an-alyzer we evaluated the data handling capability of theFSO communication system using the eye pattern technique.The essential key wave shape parameters including period,rise/fall time, clock to data jitter, overshoot, ringing, noiseand signal to noise ratio are observed to be within accept-


Fig. 9 Eye pattern with Mask showing a standard STM16/OC48 test for2.5 Gbps transmission.

Fig. 10 10 Gbps transmission test eye pattern.

Fig. 11 WDM received signal spectrum.

able tolerance. This is depicted in Fig. 9 for a 2.5 Gbps sin-gle channel transmission and in Fig. 10 which depicts eyepattern for a single channel 1.5 µm data link operating at10 Gbps.

Alternatively, by using a relatively straightforwardmethod of increasing the bandwidth by employing WDMtechnology, four 2.5 Gbps individual channels with an out-put power of 100 mW per wavelength can be realised. Thefour 2.5 Gbps channels were combined for a total wirelessthroughput of 10 Gbps. By employing this technique stablecommunication was accomplished without any fluctuationor interferences between wavelengths as shown in Fig. 11.

The total bandwidth of this FSO communication system canbe increased considerably as more channels can be activatedin the more complex WDM or dense wavelength divisionmultiplexing (DWDM) schemes.

5. Conclusion

A free-space optical communication system using speciallydesigned compact antenna for easy, cost effective means ofconstructing a robust and reliable high-speed link for nextgeneration FSO system was investigated. The FSO commu-nication system offered seamless connection of free spaceand fiber system. The transceiver incorporates a FPM forhigh-speed beam tracking and control function, therefore,having the capability to mitigate the effects of atmosphericturbulence on the transmitted optical beam. The antennaFPM tracking speed is 1 kHz which is currently consideredthe fastest tracking speed for a FSO system. The FSO com-munication system performance was verified and error freetransmission over an extended period of time was demon-strated. The system performance expressed in terms of BERperformance was also evaluated and showed to be consis-tently above acceptable levels. Stable performance after in-creasing the system bandwidth using WDM technology wasalso attained.

Acknowledgments

This work is supported by a grant from the National Instituteof Information and Communications Technology (NICT) ofJapan.

References

[1] P.L. Eardley and D.R. Wisely, “1 Gbit/s optical free space link op-erating over 40 m — System and applications,” IEE Proc. Optoelec-tron., vol.143, pp.330–333, Dec. 1996.

[2] O. Bouchet, H. Sizun, C. Boisrobert, F. de Fornel, and P.N. Faven-nec, Free-Space Optics Propagation and Communication, ISTE,2006.

[3] K. Wakamori, T. Hayashi, H. Yamashita, Y. Kimura, and M. Hosoda,“155-Mbps ATM backbone for an interbuilding intranet using anoptical wireless system,” 10th International Symposium on Personal,Indoor and Mobile Radio Communications, vol.1, pp.85–90, Sept.1999.

[4] K. Wakamori, H. Yamashita, Y. Kimura, Y. Shimodaira, and F. Sato,“Regional broadband network by optical wireless system,” 2001Asia-Pacific Radio Science Conference, pp.D4-1-02, 2001.

[5] Infrared Communication Systems Association. http://www.icsa.org.jp

[6] Free-Space Optical Alliance. http://www.wcai.com/fsoalliance[7] T.H. Carbonneau and D.R. Wisely, “Opportunities and challenges

for optical wireless: The competitive advantage of free spacetelecommunications links in today’s crowded marketplace,” Wire-less Technologies and Systems: Millimeter-Wave and Optical, ed.P. Christopher, L. Langston, and G.S. Mecherle, vol.3232, pp.119–128, SPIE, Jan. 1998.

[8] H. Willebrand and B. Ghuman, Free Space Optics: Enabling OpticalConnectivity in Today’s Networks, Sams Publishing, 2002.

[9] G. Nykolak, P.F. Szajowski, D. Romain, G.E. Tourgee, H.M. Presby,and J.J. Auborn, “Update on 4 × 2.5 Gb/s, 4.4 km free-space optical


communications link: Availability and scintillation performance,”Optical Wireless Communications II, ed. E.J. Korevaar, vol.3850,pp.11–19, SPIE, Sept. 1999.

[10] F. Marioni, Z. Sodnik, and F.E. Zocchi, “2.5-Gb/s free-space opticslink over 1.1 km with direct fiber coupling to commercial devices,”in Free-Space Laser Communications IV, vol.5550, pp.60–69, Oct.2004.

[11] H. Willebrand and B. Ghuman, “Fiber optics without fiber,” IEEESpectr., vol.38, pp.40–45, Aug. 2001.

[12] I.I. Kim, M. Mitchel, and E. Korevaar, “Measurement of scintillationfor free-space laser communication at 785 nm and 1550 nm,” OpticalWireless Communications II, ed. E.J. Korevaar, vol.3850, pp.49–62,SPIE, Sept. 1999.

[13] Y. Aburakawa, T. Otsu, and Y. Yamao, “Fiber and free-space hy-brid optical networking for new generation mobile radio access net-works,” 5th International Symposium on Wireless Personal Multi-media Communications, pp.586–590, Oct. 2002.

[14] Y. Arimoto, Y. Hayano, and W. Klaus, “High-speed optical feeder-link system using adaptive optics,” Free-Space Laser Communica-tion Technologies IX, ed. G.S. Mecherle, vol.2990, pp.142–151,April 1997.

[15] R.W. Kaliski, S.M. Genco, D.A. Thompson, B. Breshears, T.O’Connor, K.B. Miller, E.W. Taylor, A.D. Sanchez, J.E. Winter, andR.M. Ewart, “Laser communication intersatellite links realized withcommercial off-the-shelf technology,” Free-Space Laser Communi-cation Technologies XI, ed. G.S. Mecherle, vol.3615, pp.170–178,SPIE, April 1999.

[16] D.Y. Song, Y.S. Hurh, J.W. Cho, J.H. Lim, D.W. Lee, J.S. Lee, andY. Chung, “4×10 Gb/s terrestrial optical free space transmission over1.2 km using an EDFA preamplifier with 100 GHz channel spacing,”Opt. Express, vol.7, pp.280–284, Dec. 2000.

[17] Y. Aburakawa and T. Otsu, “Dense wavelength division multiplexedoptical wireless link towards terabit transmission,” InternationalTopical Meeting on Microwave Photonics, pp.135–138, Oct. 2003.

[18] S. Bloom, E. Korevaar, J. Schuster, and H. Willebrand, “Understand-ing the performance of free-space optics,” J. Optical Networking,vol.2, pp.178–200, June 2003.

[19] L.M. Wasiczko, Techniques to mitigate the effects of atmosphericturbulence on free space optical communication links, PhD Thesis,Univ. of Maryland, College Park, 2004.

[20] K. Takahashi and Y. Arimoto, “Development of optical antennasutilizing free form surface optics for the high speed laser commu-nication systems,” Free-Space Laser Communication TechnologiesXVIII, ed. G.S. Mecherle, vol.6105, pp.40–49, SPIE, Jan. 2006.

[21] T. Weyrauch and M.A. Vorontsov, “Free-space laser communi-cations with adaptive optics: Atmospheric compensation exper-iments,” J. Optical and Fiber Communications Reports, vol.1,pp.355–379, Dec. 2004.

Kamugisha Kazaura received B.Eng. de-gree in Electronics and Communications En-gineering from the University of Bath in UKin 1995. He worked for Tanzania Telecom-munications Company Ltd. (TTCL) between1996 and 1999. He later joined Waseda Uni-versity’s Graduate School of Global Informa-tion and Telecommunication Studies where hereceived his M.Sc. in Information and Telecom-munication Engineering in 2002. Currently heis working on his PhD degree. His research in-

terest includes fixed and mobile high-speed wireless communications net-works, broadband wireless networks and free-space optical communicationsystems.

Kazunori Omae received the B.Sc. degreein Liberal Arts Industry and Technology fromthe University of the Air in 2004 and the M.Sc.degree in Information and TelecommunicationEngineering from Waseda University, GraduateSchool of Global Information and Telecommu-nication Studies in 2006. He is currently em-ployed with Cable and Wireless IDC Inc andhis research interests include free-space opticalcommunication systems, millimeter wave wire-less communication systems.

Toshiji Suzuki received the M.Sc. in Elec-tronic and Electrical Engineering from SophiaUniversity Graduate School of Science and En-gineering in 1972. From 2002 he joined theGlobal Information and Telecommunication In-stitute of Waseda University as a visiting re-searcher after a success career with Canon Elec-tronics. His research interests include free-spaceoptical communication systems, infrared andvisible light communication.

Mitsuji Matsumoto Since joining NTT labsin 1970, Dr. Matsumoto has been engaged in re-search and CCITT standardization activities inthe field of protocol architecture and terminaldesign for facsimile, telematics and multimediaservices and systems. He joined the Global In-formation and Telecommunication Institute ofWaseda University, Tokyo, Japan as professor in1996. Currently he is vice director of GITI. In2000–2004 study period he became Vice Chair-man of ITU-T SG16 (Multimedia) and since

2004 he is the vice president of Infrared Data Association.

Edward Mutafungwa received the B.Eng.degree in electronic systems engineering and theM.Sc. degree in telecommunications and infor-mation systems from the University of Essex,Colchester, U.K., in 1996 and 1997, respec-tively, and the Dr.Sc.Tech. degree in communi-cations engineering from the Helsinki Univer-sity of Technology (HUT), Espoo, Finland, in2004. Since 1997, he has been lecturing and re-searching in various projects at the Communica-tions Laboratory of HUT and has consulted for

various companies. His research interests lie within the general fields ofoptical networking, network design, broadband wireless communications,intelligent transport systems, and intelligent computing.

Tadaaki Murakami received the M.Sc.,from the University of Electro-Communicationin Tokyo in 1999. He is currently working forKoito Industry Ltd. which he joined in 1999.From 2005 he joined the Global Information andTelecommunication Institute of Waseda Univer-sity, Tokyo, Japan as a visiting researcher. Heis engaged in research and development of free-space optical communications equipment andapplied technology.


Koichi Takahashi received the M.Sc. fromMuroran Institute of Technology, Departmentof Engineering Applied Solid-State physics in1984. He currently works for Olympus Corpo-ration and in 2004 he joined the Global Informa-tion and Telecommunication Institute of WasedaUniversity as a visiting researcher. His researchinterest include development of microlithogra-phy lens with focus on optical system as well asresearch and development of a free curved sur-face optical system, optical communication re-

lated to free space optical communication.

Hideki Matsumoto received the M.Sc. fromNiigata University, Graduate School of Scienceand Engineering in 1982. In 2004 he joinedthe Global Information and TelecommunicationInstitute of Waseda University as a visiting re-searcher. He is engaged in research of opticalwireless LAN system, spread spectrum commu-nication technology, as well as research and de-velopment of optical transmission device.

Kazuhiko Wakamori received the M.Sc.from Shizuoka University, Graduate School ofInformation Engineering. From 1979 he joinedHamamatsu Photonics K.K. He has been a vis-iting researcher at the Global Information andTelecommunication Institute of Waseda Univer-sity from 2004. His research interest include de-velopment of optical communication for a high-speed optical device and research and develop-ment of free-space optical communication.

Yoshinori Arimoto received the M.Sc. inPhysics from Osaka University, Osaka, Japanin 1979. From 1979 he joined the Radio Re-search Laboratory (Communication ResearchLaboratory), Ministry of Posts and Telecom-munication. During 1983–1985 he was withthe Telecommunications Satellite Corporationof Japan (TSCJ). From 1990 to 1993 he workedat the Advanced Telecommunications ResearchInstitute (ATR) Optical and Radio Communica-tions Research Laboratory. In 1999 to 2001 he

was with the National Space Development Agency of Japan. His researchinterest include communication related to space infrastructure, such as free-space laser communication.

Documents

Performance Evaluation of Next Generation Free-Space Optical