Download - Photonic True-Time Delay Beamforming Based on Polarization-Domain Interferometers

2492 JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 28, NO. 17, SEPTEMBER 1, 2010

Photonic True-Time Delay Beamforming Based onPolarization-Domain Interferometers

Miguel V. Drummond, Student Member, IEEE, Paulo P. Monteiro, Member, IEEE, andRogério N. Nogueira, Member, IEEE

Abstract—In this paper, we propose a novel photonic true-timedelay beamforming system for phased array antennas. The systemrelies on tunable delay lines which are based on Mach–Zehnderdelay interferometers (MZDIs) with tunable coupling ratio. As theMZDIs are implemented on the polarization domain, a single op-tical source and a single piece of polarization maintaining fiber arerequired. The proposed implementation is theoretically assessedand beam squinting is investigated by simulation. A proof-of-con-cept experiment that validates the operation principle of the pro-posed delay lines is presented.

Index Terms—Mach–Zehnder delay interferometer (MZDI),phased array antennas, photonic true-time delay beamforming.

I. INTRODUCTION

D EMANDING wireless applications such as electronicwarfare systems and broadband wireless networks

require advanced antenna systems which can provide highsensitivity, broad bandwidth operation, and wide and preciseangular control [1]. With this end, phased array antennas(PAAs) with photonic beamforming systems have been in-tensively investigated over the last years. Such systems sharethe advantages of microwave photonics, such as low losses,high time-bandwidth products, light weight and immunity toelectromagnetic interference. As broad bandwidth operation isrequired, photonic true-time delay (TTD) beamforming shouldbe used to avoid beam squinting.

Different photonic TTD beamforming techniques basedon different optical delay lines have been proposed [2]–[10].Optical fibers with different lengths can be used to providediscretely tunable delay lines [2]. The photonic beamformingsystem therefore consists on a switchable fiber optic network.This approach has scalability problems. For a PAA composedby several antenna elements, a fiber optic network with manyfibers is required, which results in a bulky system. Another wellknown approach is to take advantage of chromatic dispersion[4]–[6]. In [4], tunable laser sources are multiplexed and

Manuscript received January 18, 2010; revised June 20, 2010 and June 28,2010; accepted June 30, 2010. Date of publication July 15, 2010; date of currentversion August 18, 2010. This work was supported by the THRONE (PTDC/EEA-TEL/66840/2006) Fundação para a Ciência e Tecnologia (FCT) Project.The work of M. V. Drummond was supported by the FCT under the SFRH/BD/40250/2007 scholarship.

The authors are with the Instituto de Telecomuniçacões, Universidadede Aveiro, 3810-193 Aveiro, Portugal (e-mail: [email protected]; [email protected]; [email protected]).

Color versions of one or more of the figures in this paper are available onlineat http://ieeexplore.ieee.org.

Digital Object Identifier 10.1109/JLT.2010.2057408

modulated by the input RF signal, where is the number ofantenna elements of the PAA. The multiplexed signals are thenpropagated in a dispersive medium, thus acquiring a time delaydependent on their wavelength. Tunable delays among themodulated signals are achieved by tuning the wavelength of theoptical sources. This technique also has scalability problems,as the number of tunable lasers increases with the number ofantenna elements of the PAA. The dispersive medium can beimplemented with a chirped fiber Bragg grating (CFBG) [4]or a dispersive optical fiber [5]. A CFBG offers a compactimplementation. However, the group delay ripple must be lowto avoid signal distortion and beam squinting [7].

A novel technique based on ring resonators has recently in-troduced a new concept in [8]. Instead of having a constant timedelay across the entire spectral bandwidth of the optical signal,the time delay only needs to be constant at the modulated RF car-rier, as long as the phase of the optical carrier is adjusted. Suchconcept relaxes the bandwidth requirement of the tunable delaylines down to the spectral bandwidth of the data signal mod-ulated onto the RF carrier. This is especially important whenhigh-frequency RF carriers are employed. However, this con-cept is limited to optical single sideband signals with monochro-matic optical carrier.

In this paper, we propose a novel photonic TTD beamformingtechnique based on an improved concept: the amplitude andgroup delay responses only need to be correct at the opticaland RF carriers. The proposed scheme allows using double orsingle sideband signals with monochromatic or modulated op-tical carrier. The tunable delay lines are implemented with po-larization-domain Mach–Zehnder interferometers. Each inter-ferometer has a tunable coupling ratio, which allows tuning thetime delay between 0 and , where is the differential groupdelay (DGD) of the polarization-domain interferometer (PDI).The optical and RF carriers are set at different maxima of thePDI’s amplitude response, thus sharing the same frequency re-sponse. As a result, the bandwidth requirement of the tunabledelay lines is relaxed down to the spectral bandwidth of the datasignal modulated onto the RF carrier. The proposed beamformerstructure uses only one optical source that needs not to be tun-able and a single birefringent medium.

The remainder of this paper is organized as follows. A math-ematical description of the proposed photonic beamformingsystem is presented in Section II. Section III presents sim-ulation results which analyze the impact of beam squinting.A proof-of-concept experiment with the purpose of assessingthe operation principle of one PDI is described in Section IV.Section V states the main conclusions of this work.

0733-8724/$26.00 © 2010 IEEE

DRUMMOND et al.: PHOTONIC TTD BEAMFORMING BASED ON POLARIZATION-DOMAIN INTERFEROMETERS 2493

Fig. 1. Photonic TTD beamforming scheme considering a PAA with �� antenna elements.

II. OPERATION PRINCIPLE

The photonic TTD beamforming scheme is depicted in Fig. 1.A laser source provides a CW signal with a wavelength of .A Mach–Zehnder modulator is used to modulate the CW signalwith the input RF signal. The polarization controller PC 0 setsthe state-of-polarization (SOP) of the modulated signal at 45with the slow axis of the polarization-maintaining fiber (PMF),which has a differential group delay of . Thus, at the output ofthe PMF there are two modulated signals delayed by , com-bined in orthogonal polarization. A 1: optical splitter is usedto replicate the PMF output signal, where is the number ofantenna elements of the PAA. Each output of the 1: opticalsplitter is connected to a polarization controller (PC) and a polar-izer. This allows performing a weighted addition of the polariza-tion combined signals, therefore resulting in a Mach–Zehnderinterferometer with tunable coupling ratio. The output signal ofeach polarizer is converted to the electrical domain using a pho-todetector (PD) and then fed to the respective antenna element.The photonic beamformer is thus composed by PDIs. Thecoupling ratio of each PDI can be tuned by appropriately set-ting the corresponding PC.

The analysis of the beamforming scheme is divided in twosteps. Firstly, the beamformer is theoretically described. Sec-ondly, beam squinting analysis is conducted.

A. Theoretical Analysis

The tuning principle of the tunable delay lines implementedwith PDIs is similar to the one presented in [11], with the dif-ference that in this work time delays are considered instead ofdispersive media.

The transfer function of one PDI is given by

(1)

where is the angular difference between the slow axes of thePMF and polarizer . The value of can be set by the PC . Theamplitude and group delay responses of the PDI can be derivedfrom (1)

(2a)

(2b)

Fig. 2. Amplitude (a) and group delay (b) responses of the PDI for differentvalues of� . As shown in (a), the carriers of the optical signal are centered atdifferent maxima of the PDI’s amplitude response.

Both responses are depicted in Fig. 2. A periodical behavior witha period of is obtained. Fig. 2(b) shows that the group delaycan be changed with , however at the cost of also changingthe amplitude response. There are two options for centering theoptical signal with the response of the PDI. The first one is toset the optical signal within only one period of the frequency re-sponse. This approach limits the bandwidth of the optical signalto be lower than . Moreover, the attenuation of the RF carriersrelatively to the optical carrier depends on . The other optionis to center the optical and RF carriers at different maxima of thePDI’s response, as depicted in Fig. 2(a). Mathematically this canbe written as

(3)

where is the frequency of the RF signal. In order to avoidsignal distortion, the bandwidth of the signal modulated ontothe RF carriers should be lower than . In comparison to thelatter option, this relaxes the PDI’s bandwidth requirement asthe bandwidth of the optical signal is usually much higher than


Fig. 3. Amplitude and group delay responses of the PDI for � � � �� .

the bandwidth of the signal modulated onto the RF carriers. An-other advantage is that the amplitude and group delay responsesare the same for all the carriers, for any value of . The varia-tion of the group delay and amplitude responses with for allcarriers is shown in Fig. 3.

The tunable time delay range required to achieve a beamsteering tuning range of 180 can be derived from the arrayfactor of a PAA, defined as

(4)

where is the spatial angle around the orientation axis of thePAA, is the amplitude of the RF signal, is thewavenumber of the RF signal, is the distance between twoadjacent antenna elements and is phase shift applied to therespective antenna element. In case of a TTD beamformer

(5)

where is the time delay. In order to achieve a phase shiftbetween 0 and for should have a tuning range of

. This corresponds to the maximum time delay betweentwo adjacent antennas. Therefore, the maximum delay requiredby the PAA is of

(6)

The tunable time delay range can be of , or of. In practice, the latter range only

requires an absolute maximum delay of . In

order to obtain positive delays one has, while for negative delays

, whereis a constant defined by the target beampointing angle

. The DGD of the PDI should therefore be of .This condition complies with (3). However, the exponentialterm in (1) assumes negative values for even values of , as

. In this case, the optical carrier is setat a minimum (maximum) of the amplitude response, whereasthe RF carriers are set at maxima (minima). This situationcan be resolved by considering a higher maximum delay of

. According to these considerations, the DGDof the PDI is

when is odd

when is even.(7)

Equation (7) shows that the required is proportional to thenumber of antenna elements. On the other hand, the bandwidthof the amplitude response is inversely proportional to . Assuch, there is a tradeoff between maximum delay and band-width. Nonetheless, the bandwidth of the optical signal mod-ulated onto the RF carrier usually increases at the cost of alsoincreasing the RF carrier frequency, which according to (7) de-creases the required . In this case, high signal bandwidths canbe accommodated by the beamformer, since a low is required.

B. Beam Squinting Analysis

TTD operation can be demonstrated using the same analysisas in [4]. The modulated optical double sideband (DSB) signalsat the input of the PDs, , can be written as a sum of threespectral lines,

(8)

where is the optical carrier frequency. After direct detectionof , the photocurrent can be expressed by

(9)

and (10), shown at the bottom of the page, where is thephase of . The phase of the photocurrent is given by(11), shown at the bottom of the page. As depicted in Fig. 2(a),

. From (1), it

(10)

(11)


Fig. 4. Group delay variation with � , considering different values of �� .The group delay variation is given by �� .

can be seen that .Hence, (10) and (11) are simplified to

(12)

(13)

Equation (13) shows that TTD operation is achieved acrossa given bandwidth, at which the approximation is valid. Theobtained time delay depends on the optical carrier frequency.Fig. 4 shows that group delay error increases with the frequencyshift, defined as , where is the frequencyat which a maximum of the interferometer response shouldbe located. The error also depends on , being maximumfor of about 20 and 70 . Therefore, a frequency shift of theamplitude response maximum from causes beam squinting.This can be particularly serious if the PDIs have differentfrequency shifts. The amount of squinting also depends on theconsidered values of .

Equation (12) shows that the amplitude of the RF signal de-pends on the optical carrier frequency, i.e., ,and thus also of . Even when there is no frequency shift of theamplitude response, the amplitudes of the RF signals can differup to 3 dB depending on . This is depicted in Fig. 3. Equation(4) shows that target beampointing angle only depends on thephase of the RF signals, . Hence, deviations ofthe RF amplitudes do not cause beam squinting, although beamshaping is obtained.

For RF signals with broad data bandwidth or with an RF fre-quency not compliant with (3) the approximation done in (13) isnot valid. Under such conditions, the phase of the photocurrentis given by

(14)

where is the frequency of the PDI’s amplitude responsemaximum nearest to . The phase of the detected RFsignal mainly depends on the group delay of the optical carrier,as . The first term corresponds to the phase

Fig. 5. Radiation diagrams for � of (a) 0 , (b) 20 , (c) 45 , and (d) 60 .The solid line represents the proposed beamformer; the dashed line representsan ideal TTD beamformer.

shift originated by the frequency detuning between the RF car-rier and . If such detuning is zero, thenand . Under such condition (14) can be writtenas (13).

Finally, it should be noted that the analysis performed on(12)–(14) is also valid for single sideband signals. By sup-pressing the RF carrier at , (12) and (13) become

(15)

(16)

III. SIMULATION RESULTS

The proposed beamformer was simulated considering a PAAwith antenna elements, GHz, and a distancebetween two adjacent antenna elements of mm.According to (7), ps. An optical DSB signalwas considered. Thus, the amplitude and phase of the RF signalswere obtained from (12) and (13). The radiation diagrams wereobtained using (4). Results considering different target beam-pointing angles are shown in Fig. 5. Beam squinting is not ob-served. However, beam shaping arises from the dependence ofthe RF amplitudes with the required time delays. Even thoughthe observed beam shaping is insignificant, it can be controlledusing variable optical attenuators at the outputs of the 1: op-tical splitter. Other option is to use RF amplifiers with variablegain after the PDs.


Fig. 6. Beam squinting considering different frequency shifts. Inset: radiationdiagrams for � � �� .

Fig. 7. Beam squinting considering different RF frequency detunings relativelyto � . The frequency detuning is normalized to the 3-dB cutoff frequency.Inset: radiation diagrams for � � �� .

Beam squinting and additional beam shaping occur whenthere is a frequency shift between the optical signal and thePDIs. Fig. 6 shows that beam squinting increases significantlywith . The beam squinting is the same for , as aresult of the symmetry of the PDI’s amplitude response with re-spect to its maxima. The additional beam shaping can severelydistort the radiation diagram, as shown in the inset of Fig. 6.It should be noted that these results consist on the worst-casescenario, which corresponds to having the same frequency shiftfor all the PDIs. Further simulations considering the samevalues of have shown that beam squinting increases withthe number of antenna elements. The increase of results inthe increase of the required . Thus, according to Fig. 4, a givenfrequency shift yields a higher group delay variation,which explains the increase of the beam squinting.

As mentioned in the last subsection, the approximation ofa PDI to an ideal delay line taken in (13) is only valid withina given bandwidth. This is particularly relevant for opticalRF signals with broad data bandwidths or with optical RFcarriers located away from the maxima of the PDI’s amplituderesponse. The impact of the RF frequency and data bandwidthwas assessed considering different RF frequencies. The RFphases were calculated using (14). The results are shown inFig. 7. Fig. 7 shows that the detuning of the RF frequencyresults in reduced beam squinting. However, the radiationdiagrams show that the power of the sidelobes increases. Theresults displayed in Figs. 6 and 7 show that a frequency shiftbetween the optical signal and the PDIs results in stronger beamdistortion in comparison to having an RF frequency differentfrom .

Fig. 8. Experimental setup. ONA—optical network analyzer; PL—polariza-tion locker.

Fig. 9. Group delay at the maxima of the amplitude response as function of theextinction ratio.

IV. EXPERIMENT

A proof-of-concept experiment with discrete componentswas performed in order to assess the operation principle of thePDI. The experimental setup is depicted in Fig. 8. A PMF witha DGD of 33.3 ps was used. PC 0 was a manual PC with threewave plates, used to set the state-of-polarization (SOP) of theinput signal at 45 relatively to the slow axis of the PMF. ThePC placed after the PMF output was a deterministic polarizationlocker (PL), which allowed transforming any input SOP to anyoutput SOP. An optical network analyzer (ONA) was used tomeasure the amplitude and group delay responses of the PDI.As the ONA can only measure relative values of the groupdelay, the absolute group delay was derived from the obtainedextinction ratio (ER) of the amplitude response. The ER is thedifference in dB between the optical powers of the maximaand minima of the amplitude response. Fig. 9 shows that thereare two possible group delays for the same ER. However, bothdelays can be distinguished as the corresponding group delayresponses are symmetric with respect to .This is depicted in Fig. 2(b).

The frequency response of the PDI was measured consid-ering different ERs. The results are shown in Fig. 10. The angle

increases from 0 up to 90 with the measurementnumber, thus proving that the group delay can be tuned from0 up to ps. A maximum ER of 29.5 dB was obtainedfor . The ER was limited to such value due to theinability of setting at exactly 45 , and also due to the lim-ited ER of the polarizer [12]. Fig. 10(b)–(d) shows that a goodagreement was obtained between the experimental and theoret-ical curves. Fig. 10(d) shows that the group delay response hadpositive and negative values at different minima of the ampli-tude response. As depicted in Fig. 2(b), slight variations of


Fig. 10. Measured (a) ERs and (b)–(d) selected amplitude and group delay responses.

Fig. 11. Temporal variation of the amplitude response’s frequency shift.

around 45 cause the group delay at the minima of the ampli-tude response to switch between positive and negative values.Although the employment of the PL resulted in a stable outputSOP, the short patch cord that connected the PL to the polarizerwas responsible for slight variations of that in turn resultedin the observed group delay variations.

As proved in Section II, frequency shifts on the interferom-eter response are deleterious since they lead to beam squinting.Instabilities on the experimental setup can result in frequencyshifts. In order to assess the stability of the experimental setupthe PL was firstly adjusted to obtain a maximum ER of 29.5 dB,and then unlocked. Measurements were taken during 30 min-utes, with a time interval of 5 minutes between them. The vari-ation of the ER during the 30 minutes was less than 1 dB. How-ever, as shown in Fig. 11 the frequency response drifted sig-nificantly over time. These results show that although wasstable, there was a random phase shift on the PDI’s re-sponse which can be modeled on (1) as

(17)

where is the Fourier transform of . Such phase shiftresulted from environmental instabilities at the patch cord thatconnected the PL to the polarizer. Therefore, stable operationwithout the need of polarization tracking can be achieved simplyby connecting the PL directly to the polarizer, without a patchcord.

V. CONCLUSION

In this paper, a novel photonic TTD beamforming system waspresented. The proposed system relies on tunable delay lines,which are based on Mach–Zehnder delay interferometers. Thetime delays are continuously tunable by changing the couplingratio of the interferometers. The optical bandwidth of the inter-ferometers needs only to be higher than the bandwidth of thedata signal modulated onto the RF carriers. This significantlyrelaxes the required bandwidth, as the spectral bandwidth of theoptical signal is usually much higher than the bandwidth of thedata signal. The proposed beamforming scheme implements theinterferometers on the polarization domain. As a result, it onlyrequires one optical source that needs not to be wavelength tun-able and a single piece of PMF.

A detailed mathematical analysis of the system and numericalsimulations have proven that squint-free operation is achievedover a significant bandwidth under frequency-stable operationof the interferometers. The optical carrier and frequency re-sponse of the interferometers should be properly centered inorder to avoid beam squinting and beam distortion.

A proof-of-concept experiment validated the operationprinciple of the proposed delay line. It was observed that polar-ization instabilities on the setup cause random frequency shifts.The polarization instabilities originated from the patchcordconnecting the PL and the polarizer. Hence, stable operationwithout the need of polarization tracking can be easily achievedwith the suppression of such connection.

The evolution of the proof-of-concept setup with discretecomponents into an implementation in integrated optics shouldbring additional benefits such as intrinsically stable operation,low tuning time, low power consumption, compactness and lowweight. The PCs can be implemented with polarization-sen-sitive phase modulators, which are basically polarizationmodulators. Furthermore, since all the waveguides can bebirefringent, an implementation in LiNbO technology ispromising.

REFERENCES

[1] R. C. Hansen, Phased Array Antennas. New York: Wiley, 1998.


[2] W. Ng, A. A. Walston, G. L. Tangonan, J. J. Lee, I. L. Newberg, and N.Bernstein, “The first demonstration of an optically steered microwavephased array antenna using true-time-delay,” J. Lightw. Technol., vol.9, no. 9, pp. 1124–1131, Sep. 1991.

[3] E. N. Toughlian and H. Zmuda, “A photonic variable RF delay linefor phased array antennas,” J. Lightw. Technol., vol. 8, no. 12, pp.1824–1828, Dec. 1990.

[4] B. Ortega, J. L. Cruz, J. Capmany, M. V. Andres, and D. Pastor,“Variable delay line for phased-array antenna based on a chirpedfiber grating,” IEEE Trans. Microw. Theory Tech., vol. 48, no. 8, pp.1352–1360, Aug. 2000.

[5] J. Yongqiang, B. Howley, S. Zhong, Z. Qingjun, R. T. Chen, M. Y.Chen, G. Brost, and C. Lee, “Dispersion-enhanced photonic crystalfiber array for a true time-delay structured X-band phased array an-tenna,” IEEE Photon. Technol. Lett., vol. 17, no. 1, pp. 187–189, Jan.2005.

[6] P. Q. Thai, A. Alphones, and D. R. Lim, “A novel simplified dualbeam-former using multichannel chirped fiber grating and tunable op-tical delay lines,” J. Lightw. Technol., vol. 26, no. 15, pp. 2629–2634,Aug. 2008.

[7] P. Q. Thai, A. Alphones, and D. R. Lim, “Limitations by group delayripple on optical beam-forming with chirped fiber grating,” J. Lightw.Technol., vol. 27, no. 24, pp. 5619–5625, Dec. 2009.

[8] P. A. Morton and J. B. Khurgin, “Microwave photonic delay line withseparate tuning of the optical carrier,” IEEE Photon. Technol. Lett., vol.21, no. 22, pp. 1686–1688, Nov. 2009.

[9] M. A. Piqueras, G. Grosskopf, B. Vidal, J. Herrera, J. M. Martinez, P.Sanchis, V. Polo, J. L. Corral, A. Marceaux, J. Galiere, J. Lopez, A.Enard, J. L. Valard, O. Parillaud, E. Estebe, N. Vodjdani, C. Moon-Soon, J. H. den Besten, F. M. Soares, M. K. Smit, and J. Marti, “Opti-cally beamformed beam-switched adaptive antennas for fixed and mo-bile broadband wireless access networks,” IEEE Trans. Microw. TheoryTech., vol. 54, no. 2, pp. 887–899, Feb. 2006.

[10] L. Jofre, C. Stoltidou, S. Blanch, T. Mengual, B. Vidal, J. Marti, I.McKenzie, and J. M. del Cura, “Optically beamformed widebandarray performance,” IEEE Trans. Antennas Propag., vol. 56, no. 6, pp.1594–1604, Jun. 2008.

[11] M. V. Drummond, R. N. Nogueira, P. Monteiro, M. Violas, C. Sterner,and P.-Y. Fonjallaz, “Novel opto-electrical tunable dispersion com-pensator for IM signals,” presented at the The 35th Eur. Conf. Opt.Commun., Vienna, Austria, 2009.

[12] M. V. Drummond, P. Monteiro, and R. N. Nogueira, “Photonic RF in-stantaneous frequency measurement system by means of a polarizationdomain interferometer,” Opt. Exp., vol. 17, no. 7, pp. 5433–5438, 2009.

Miguel V. Drummond (S’09) was born in Porto, Portugal, in 1984. He receivedthe Diploma degree in electronics and telecommunications engineering from theUniversity of Aveiro, Aveiro, Portugal, in 2007, where he is currently workingtoward the Ph.D. degree.

Since 2008, he has been with the Optical Communications Research Group,Institute of Telecommunications, Aveiro. His main research interests includeall-optical signal processing techniques applied to high bitrate optical commu-nication systems and microwave photonic devices.

Paulo P. Monteiro (M’06) was born in Coimbra, Portugal, in 1964. He receivedthe diploma degree in electronics and telecommunications from the Universityof Aveiro, Aveiro, Portugal, the M.Sc. degree from the University of Wales,Wales, U.K., and the Ph.D. degree in electronics and telecommunications fromthe University of Aveiro.

He is a Research Manager with Nokia Siemens Networks, Amadora, Por-tugal, and an Associate Professor with the University of Aveiro, where he hasbeen teaching courses of telecommunications. He is also a Researcher with theInstitute of Telecommunication, University of Aveiro. His main research inter-ests include high-speed optical communications for access and core networksand fixed-mobile convergence. He has acted as a reviewer for Electronics Let-ters, ETRI Journal, and the Journal of Optical Networking. He has participatedin several national and European projects and he is currently the Project Coor-dinator of the large-scale integrating project FUTON (FP7 ICT-2007-215533).He has authored or coauthored more than 15 patent applications and over 200refereed papers and conference contributions.

Dr. Monteiro has been a reviewer for the JOURNAL OF LIGHTWAVE

TECHNOLOGY.

Rogerio N. Nogueira (M’08) received the degree in physics engineering andPh.D. degree in physics from the University of Aveiro, Aveiro, Portugal, in 1998and 2005, respectively.

He is now an Assistant Researcher with the Institute of Telecommunicationswhere he has been working in the field of fiber optics since 1999, participating inseveral projects financed by national, EU organizations and private companies.In 2009, he joined Nokia Siemens Networks, Amadora, Portugal, as an R&Dexpert in the field of optical communications. He has authored or coauthoredone book chapter, more than 30 papers in international scientific journals, andmore than 100 papers in international conferences. His research interests includedesign and production of optical components, fiber optical communication sys-tems and fiber optical sensors.

Download - Photonic True-Time Delay Beamforming Based on Polarization-Domain Interferometers

Top Related