Teletraffic Capacity Performance of WDM/ Direct Sequence ... · teletraffic theory to compare the maximum value of teletraffic capacity of WDM/DS-OCDMA system. This results in designing

Abstract— Passive optical network (PON) has been widely

accepted to provide broadband services to end users of the access network. In this paper, we study teletraffic capacity performance of a WDM/DS-OCDMA PON. This hybrid configuration is proposed to increase the user multiplexing capacity and to be a cost effective method per wavelength of a WDM PON. The architecture of the WDM/DS-OCDMA PON is presented and its capacity performance is analyzed using different direct sequence optical code division multiple access (DS-OCDMA) codes. Since multiple access interference (MAI) noise is dominant noise, the number of available WDM/DS-OCDMA channels is highly dependent to the code correlation properties. For each code type, an appropriate simultaneous capacity threshold has been measured in a manner to ensure good transmission performance (BER~10-9). As this capacity threshold is inferior to the nominal multiplexing capacity of the considered code, a soft blocking of the system has been determined even if resources in the system are still available. The teletraffic capacity of the WDM/DS-OCDMA system has been then analyzed under the maximum of soft blocking probability constraint for different prime codes. Finally, using extended quadratic congruent code (EQC) has been demonstrated to lead to better teletraffic capacity in a WDM/DS-OCDMA PON.

Index Terms—Code-division multi-access (CDMA), optical fiber communication, queueing analysis.

I. INTRODUCTION HE conventional data, voice and the new video services such as video sharing, teleteaching, and also Internet

access with high data rate encourage network operators to build up a next step toward a future access network generation. Passive optical network (PON) has been largely recognized as an attractive solution to the last mile bottleneck, providing broadband access networks to end users [1]. PON is a point to multipoint optical network, connecting an optical line terminal (OLT) on the service provider side to the

Manuscript received February 3, 2009. This work was supported in part by ARAL project at Institute Telecom.

M. Gharaei and P. Gallion are with the Telecom ParisTech, Ecole Nationale Supérieure des Télécommunications, CNRS LTCI, 46 rue Barrault, 75634 Paris Cédex 13, France (phone: +33-145-818-085; fax: +33-145-804-036; e-mail: [email protected]; philippe. [email protected]).

C. Lepers is with Telecom & Management SudParis, 9 rue Charles Fourier, 91011 Evry Cedex, France (email: [email protected]).

O. Affes was with the Telecom Lille1, Cité Scientifique - Rue Guglielmo Marconi, BP 20145 - 59653 Villeneuve d'Ascq Cedex, France. (email: [email protected]).

multiple optical network units (ONUs) on the subscriber side via a non active star coupler (SC). Time division multiplexing PON (TDM PON) is the commonly accepted PON for network operators, sharing network bandwidth temporally between users. Though TDM PON benefits low installation and maintenance cost, it does not exploit the huge bandwidth of optical fiber. In order to increase optical bandwidth per user, wavelength division multiplexing PON (WDM PON) was proposed to create point to point links between OLT and users [2]. Each user is assigned a dedicated wavelength to enjoy a large bandwidth for communication. WDM PON also benefits guaranteed quality of service (QoS), protocol transparency and channel independency even though, provided bandwidth by WDM PON is too large for user applications. Moreover WDM PON cost is very high compared with TDM PON. Therefore hybrid WDM/TDM PON was proposed to exploit optical bandwidth efficiently with sharing the cost between access network subscribers [3], [4]. In other words, bandwidth sharing technique leads to share the optical resources cost between users. Temporal multiplexing per wavelength leads to increase the number of end users in access network. Yet WDM/TDM PON does not supply with high data rate per user. Also it requires a contention management protocol to avoid the situation in which more users access the channel at the same time. To resolve this contention situation, optical code division multiple access (OCDMA) was proposed [5]-[7]. Using asynchronous OCDMA as a quasi contentionless multiple access technique prevents from contention management protocols [8]. Various OCDMA techniques have been proposed such as temporal encoding, which is known as direct sequence encoding (DS-CDMA) [9], spectral amplitude coding (SAC-OCDMA) [10] and/or phase encoding [11].

Taking into consideration the compatibility of another dimension to be added by wavelength in WDM system, DS-OCDMA is a convenient technique which uses temporal encoding. WDM/DS-OCDMA is a multiplexing technique which each user data bit is encoded with a given sequence of pulses in temporal axes. Then, for a given wavelength there are different DS-OCDMA encoders/decoders who communicate concurrently with a huge optical bandwidth. This approach is deployed for all wavelengths to establish a hybrid WDM/DS-OCDMA network.

We study here, WDM/DS-OCDMA PON where each user gains the whole bandwidth usage per wavelength which

Teletraffic Capacity Performance of WDM/ Direct Sequence-OCDMA PON

Mohammad GHARAEI, Catherine LEPERS, Member, IEEE, Olfa AFFES, and Philippe GALLION, Senior Member, IEEE

T

acknowledges efficient utilization of optical resources. In other words, data rate per bandwidth utilization or spectral efficiency of WDM/DS-OCDMA PON is superior to WDM/TDM PON. Furthermore the number of users is extremely increased by bringing into play the higher capacity codes. Thus WDM/DS-OCDMA is considered as a well-matched technique in physical layer to increase the user capacity of access networks.

However, transmission performance of WDM/OCDMA system is limited by optical source noise, channel nonlinearities and channel noise including multiple access interference (MAI) [12], [13]. These noises degrade optical bit-error-rate (BER) performance in the system. MAI noise is considered as the dominant cause of BER degradation in OCDMA systems.

With the request of high broadband services by huge number of users, determination of system user capacity plays an important role in designing network architecture. Generally, nominal resource capacity is defined as maximum number of resources in the different systems; these are wavelengths in WDM system and assigned codes in OCDMA system. On the other hand, transmission performance of WDM/DS-OCDMA is limited by MAI noise which makes its nominal resource capacity unachievable. Thus we define simultaneous user capacity in OCDMA system as maximum simultaneous active users at acceptable BER performance.

Generally lack of resources in a system causes blocking in that system. It means that for a new user arrival, there would be no resource any more for transmission. In WDM/DS-OCDMA PON, there are two types of blocking probability. The first one is hard blocking probability which represents the determined blocking such as reaching to nominal resource capacity in WDM system. The second is soft blocking probability which demonstrates the flexible blocking as it is observed in OCDMA systems when calculating simultaneous user capacity.

To make a relationship between system capacity and grade-of-service, teletraffic becomes a tool by which investments can be planned [14]. Teletraffic capacity provides information

on how a network operator should control data admission to the network so as to ensure a more robust quality of service. It demonstrates also foreseeable capacity measurement with different probability rates. Degradation of transmission quality affects the teletraffic capacity of CDMA systems [15]. The teletraffic performance of OCDMA system has been already determined and compared with WDM system by Goldberg et al [16]. The authors have demonstrated that OCDMA is well suited to applications where conventional hard blocking is undesirable.

In this paper, we investigate that the teletraffic capacity performance of OCDMA system is dependent to the code family type. We then analyze teletraffic capacity performance of WDM/DS-OCDMA PON using different prime codes. Initially, we measure the number of simultaneous active users employing different prime codes in DS-OCDMA under system conventional . This measurement leads to bring out the notion of simultaneous capacity threshold. Since different prime codes have different correlation properties; they support different simultaneous capacity threshold value. Afterward, these values for different prime codes are used in teletraffic theory to compare the maximum value of teletraffic capacity of WDM/DS-OCDMA system. This results in designing proposed system with a defined grade of services and a high system capacity as cost effectively as possible. The teletraffic capacity of proposed system architecture demonstrates the performance measurement of WDM/DS-OCDMA system using different prime codes. At last, comparing this result with WDM PON system leads to build another step toward the future broadband access network.

The rest of this paper is organized as follows. In section II, the proposed network architecture is introduced and simultaneous capacity performance of OCDMA system using different prime codes is defined. In section III, soft blocking probability of DS-OCDMA is measured for different codes. In section IV, by exploiting the physical limits of OCDMA system in teletraffic capacity, we compare the teletraffic capacity of proposed hybrid WDM/DS-OCDMA PON using

Fig. 1 Architecture of hybrid WDM/DS-OCDMA PON in downstream direction.

different prime codes. Finally we conclude this paper in section V.

II. SYSTEM CAPACITY PERFORMANCE OF WDM/DS-OCDMA PON

A. Architecture of WDM/DS-OCDMA PON Hybrid WDM/OCDMA PON approach profits building

overlay on top of existing WDM PON. Fig. 1 shows the proposed architecture of WDM/DS-OCDMA PON. This network architecture is proposed for optical access network where the distance between OLT and ONUs does not exceed 20 km.

In this architecture, different optical pulse generators are used in OLT to create short pulses from to . Then each pulse is forked to K branches in accordance with the code capacity. These pulses are modulated with user data. After that, the modulated data are coded using temporal OCDMA encoders and subsequently coupled and multiplexed to transport in the access feeder fiber. At the reception part, the filtered wavelengths from to are divided to the same coding branches with the adapted decoder in each branch. Finally, the autocorrelation pulse of the decoder is detected via a receiver. Based on OCDMA approach, transmitted encoded data to each user is not decoded by other users since they do not possess the matched decoder. The presence of multiple users affects the BER performance of the system.

B. Simultaneous user capacity threshold of WDM/DS-OCDMA PON In this subsection, we compare system capacity

performance of proposed WDM/DS-OCDMA system by means of nominal user multiplexing capacity and by determining simultaneous user capacity threshold applying different prime codes. It seems at prior that hybrid WDM/DS-OCDMA PON is capable to increase the system capacity of WDM PON by G×N. This value is the nominal capacity of the system where G is the number of wavelengths and N is the user multiplexing capacity. However all code resources in each wavelength could not be utilized at the same time due to MAI limitation. Thus the maximum capacity of WDM/DS-OCDMA PON is practically assumed to be G× where denotes the simultaneous user capacity threshold of OCDMA system.

It has been revealed in [17], [18] that WDM systems suffer from nonlinear effects. They affect the performance of system by unwanted crosstalk. One of the major dominant effect is four wave mixing (FWM). Our system is designed in access network where the distance between OLT and ONU does not exceed 20 km. Therefore, the FWM in a system using standard fiber type and low power transmitting source can be neglected. It is the reason why we will focus on impairments resulting from OCDMA system. The performance of OCDMA system is strongly limited by MAI noise [12]. As the number of simultaneously users increases in OCDMA system, bit-error rate (BER) performance degrades.

BER performance of OCDMA systems is highly dependent

to the code family. Accordingly, we compare the properties of different DS-OCDMA prime codes. First of all, the prime codes are defined as (p,ω,L) where p is the prime number, ω is the weight of the code and L denotes the length of the code sequence. As prime number p is increased in a code type, the weight ω and subsequently nominal user multiplexing capacity of code N is respectively enlarged. Prime codes performances are evaluated based on their correlation properties. The autocorrelation and cross correlation

of prime codes are given by

(1)

where and are two different code sequences, and and are the maximum auto/cross correlation levels. The comparison of correlation properties of prime sequence (PS), extended prime sequence (EPS), quadratic congruent (QC) and extended quadratic congruent (EQC) codes is presented in Table I.

With the aim of understanding better the effect of MAI

noise in BER performance, we assume K different users for each wavelength. The signal-to- noise ratio (SNR) is given by the ratio of the autocorrelation peak squared to the variance of the amplitude of the interference. There are (K-1) users’ signals which interfere with desired signal. These interferences are supposed to be uncorrelated and to have an identical variance. Therefore the signal-to-noise ratio (SNR) for the code sequence is given by [19]:

(2)

where is the variance of additive noise power. Based on SNR ratio, the number of simultaneous active users is defined in accordance with BER performance.

The variance of additive noise power is identified as average variance of the cross correlation amplitude [20]. It is calculated over all possible cross correlation pairs in a given code [20], [21]. Asynchronous OCDMA system has demonstrated a lower bound on the performance [5], [6]. We

TABLE I COMPARISON OF CORRELATION PROPERTIES OF DIFFERENT PRIME

SEQUENCE FAMILY CODES. Code

Family L N

PS p2 p p p-1 2 EPS p(2p-1) p p p-1 1

QC p2 p p-1 2 4 EQC p(2p-1) p p-1 1 2

take into account this assumption for PS and EPS codes. However for QC and EQC codes, we follow the congruent operator to estimate variance of noise power [22]. Fig. 2 shows the number of simultaneous active users in a system using different prime codes which can communicate with a defined BER performance. The degradation of BER performance over increased simultaneous users is observed for all possible OCDMA systems. The curves deviate since different prime codes possess different correlation properties and consequently different BER performances. Hence, we define a threshold value of simultaneous active users at conventional for different prime codes. The number of simultaneous active users is fixed at which represents the simultaneous user capacity threshold. The simultaneous threshold values for PS, EPS, QC and EQC code with p=31 are respectively 17, 19, 28 and 30. Let us note that value is increased as a function of prime number in this system. Beyond this simultaneous threshold value, the quality of service (QoS) of the transmission will not be guaranteed.

III. SOFT BLOCKING PROBABILITY OF DS-OCDMA The aim of this section is to determine the soft blocking

probability when the resource utilization of the system is increased. The state probabilities are here directly used to model the performance measurements of the system. In the case where the number of users is greater or equal to the number of channels, the generalized Engset loss model is used [14]. The Engset system is characterized by the following parameters: = offered traffic per user [23], K= number of users, and = number of active WDM or/and DS-OCDMA channels. The number of active channels is supposed to be a stationary random process over some temporal interval which follows the binomial distribution [24]. Thus the probability distribution in this limited capacity system for a

new arrival channel demand concerning about its stochastic random process is given by [14]

(3)

This formula represents the probability of a new channel demand rejection corresponding to the case where the system is blocked strictly for a new user channel order. In WDM system model with G users, It denotes that neglecting nonlinearities, a WDM system can support a maximum number of users as equal to the number of wavelengths. It has a hard blocking probability since there is not any wavelength available after blocking.

However in OCDMA system, which means that simultaneous user capacity threshold is the maximum value for active OCDMA channel. As soft blocking takes place, there are still at maximum available codes. That’s why it is named as soft blocking probability. Soft blocking probability has its own probability distribution. In order to verify it, binomial distribution has to be defined. Binomial distribution has probability density function of

It is a trial which is a

"success" outcome and is a "failure" outcome. Subsequently the number of simultaneous active channels M follows the probability distribution which is derived from the total binomial distribution and active channel probability distribution. Therefore soft blocking probability is written by:

(4)

DS-OCDMA system model has a soft blocking probability which means that there would be still accessible channel resource after the simultaneous threshold value but without anymore QoS guarantee.

Let us note that parameter is not directly measurable. Just an average transmission rate of a source is observed. So is defined as the probability of channel activity. The product of (4) results in defining the activity rate of a channel which is represented by: The probability of the number of simultaneously active DS-OCDMA channel should be less or equal to the threshold value of As a result, the soft blocking probability in proposed architecture is given by:

(5)

When WDM/DS-OCDMA channel demands are exceeded

simultaneous user capacity threshold (channel demands > ), new demand is admitted however it does not maintain QoS constraint such as BER performance.

Fig. 2. BER performance versus number of simultaneous active users for different prime codes with p=31.

To obtain the average number of DS-OCDMA channels supported by the network, the offered channel load should be aggregated. When there is no blocking, the aggregated carried load is equal to aggregated offered load which is given by:

(6)

Fig. 3 compares the soft blocking probability of DS-

OCDMA system exploiting different prime codes versus the aggregated offered load. These simulations are done with different channel activity rates. It confirms that low channel activity rate cause low rate of soft blocking probability. These results demonstrate that EQC code permits to reach higher aggregated offered load with lower soft blocking penalty. Consequently the best aggregated offered load corresponds to the code with better correlation properties.

IV. TELETRAFFIC CAPACITY OF WDM/DS-OCDMA PON In this section, we measure teletraffic capacity of

WDM/DS-OCDMA PON. The teletraffic capacity with the specification of quantitative measurements for grade of service, forecast the traffic demand and the system teletraffic capacity requirement is one of the objectives of network operators.

We define the maximum value of the teletraffic capacity (Max ) such that the blocking probability does not exceed the blocking constraint. The blocking constraint is estimated based on system properties such as channel resources and/or commutation loss. To determine the maximum teletraffic capacity in proposed system, we take into account constraint (line inserted in Fig. 3).

Fig. 4 shows the teletraffic capacity of WDM/DS-OCDMA architecture using different prime codes. In this simulation, we consider 32 users per wavelength with totally 8 wavelengths and the prime code with p=31. It is demonstrated that using codes with better correlation properties approaches the

Fig. 3. Soft blocking probability versus aggregated offered load with different channel activity ( ) from 0.05 to 1, for a) PS code, b) EPS code, c) QC code and d) EQC code. We consider k=32 users and prime code with p=31.The dot lines show the Max blocking constraint at 10-6.

teletraffic capacity to the nominal teletraffic capacity corresponding to the nominal resource user capacity. For instance in this simulation, WDM/EQC-OCDMA PON can stay at maximum nominal teletraffic capacity with a channel activity of 0.55. It is also deducted that teletraffic capacity of WDM system is independent of channel activity and its hard blocking probability just depends on the number of active channels N. It is depicted as a constant line near zero in Fig. 4.

It is also verified from Fig. 4 that in low channel activity rate, the teletraffic capacity is equal to the nominal teletraffic resource capacity. We define the shooting point as the point which the teletraffic capacity initiates to deviate from the nominal teletraffic resource capacity. Fig. 5 demonstrates the shooting points of different prime codes when the Maximum blocking constraints change. This figure states how appointing maximum blocking constraint value is important in teletraffic capacity of the network. For example, WDM/EQC-OCDMA PON teletraffic capacity might equal to the nominal resource capacity even with channel activity of 0.8 while we fix a maximum blocking constraint at 10-2.

Finally, the results shows that EQC code which have the better correlation properties are more efficient to be used in WDM/DS-OCDMA architecture. Therefore, the DS-OCDMA teletraffic capacity performance using different prime codes leads to define a foreseeable capacity of the system. This foreseeable capacity measurement is primordial for network operators to adapt their statistical channel demands with physical layer network architecture.

V. CONCLUSION The capacity of optical access networks plays an important

role in designing the future FTTH architecture. In this paper, we have proposed a hybrid PON architecture using WDM/DS-OCDMA multiplexing technique. Based on code characteristics, the performance of the simultaneous active users for different prime codes in WDM/DS-OCDMA system is defined and consequently, the simultaneous capacity

threshold is estimated for each code. Then, the soft blocking probability of DS-OCDMA is determined with different channel activity versus the aggregated offered load. At last, teletraffic performance of WDM and DS-OCDMA system with considering the blocking constraint was compared. This comparison leads to justify the increased multiplexing capacity of hybrid WDM/DS-OCDMA system. We have demonstrated that WDM/DS-OCDMA PON using the extended quadratic congruent (EQC) code leads to increase the multiplexing capacity. It is also worth noticing that these analyses can be developed by employing more wavelengths.

REFERENCES [1] F. Effenberger, D. Cleary, O. Haran, G. Kramer, R. D. Li, M. Oron, T.

Pfeiffer, “An introduction to PON technologies,” IEEE Commun. Magazine, vol. 45, no. 3, pp. S17 - S25, Mar. 2007.

[2] S. J Park, C. H. Lee, K. T. Jeong, H. J Park, J. Gyun Ahn; K. –H. Song, “Fiber-to-the-home services based on wavelength-division-multiplexing passive optical network,” J. Lightw Technol., vol. 22 , no. 11, pp 2582-2592, Nov. 2004.

[3] F.T. An, D. Gutierrez, K. S. Kim, J. W. Lee and L. G. Kazovsky, “SUCCESS-HPON : A next generation optical access architecture for smooth migration from TDM-PON to WDM-PON,” IEEE Commun. Magazines, vol. 43, no. 11, pp s40-s47, Nov. 2005.

[4] G. Talli, P. D. Townsend, “Hybrid DWDM-TDM long-reach PON for next-generation optical access,’’ IEEE Journal of Lightwave Technology, vol. 24, no. 7, pp 2827-2834, Jul. 2006.

[5] J. A. Salehi, “Code division multiple access techniques in optical fiber networks-Part I: Fundamental principles,” IEEE Trans. Commun., vol. 37, no. 8, pp. 824–833, Aug. 1989.

[6] J. A. Salehi, and C. A. Brackett, “Code division multiple access techniques in optical fiber networks-Part II: Systems Performance Analysis,” IEEE Trans. Commun., vol. 37, no. 8, pp. 834–842, Aug. 1989.

[7] A. Stok, E. H. Sargent, “The role of optical CDMA in access networks,” IEEE Commun. Magazine, vol. 40, no. 9, pp S83-S87, Sep. 2002.

[8] R. Prasad, CDMA for wireless personal communication, Norwood, MA, Artech House 1996.

[9] I. Fsaifes, C. Lepers, M. Lourdiane, P. Gallion, V. Beugin and P. Guignard, “Source Coherence Impairments in a Direct Detection DS-OCDMA system,” J. Appl. Opt., vol. 46, no. 4, pp 456-462, Feb. 2007.

[10] J. Penon, Z. A, El-Sahn, L. A. Rusch, S. LaRochelle, “Spectral-Amplitude-Coded OCDMA Optimized for a Realistic FBG Frequency Response,” J. Lightw Technol., vol. 25, no. 5, pp.1256-1263, May 2007.

Fig. 4. Teletraffic capacity of WDM/DS-OCDMA PON using different prime codes with different probability of channel activity.

Fig. 5. Shooting points of WDM-OCDMA-PON teletraffc capacity using different prime codes versus different probability of channel activity.

[11] S. Galli, R. Menendez, P. Toliver, T. Banwell, J. Jackel, J. Young, S. Etemad, “DWDM-Compatible Spectrally Phase Encoded Optical CDMA,” IEEE GLOBCOM., vol. 3, pp. 1888-1894, Dec. 2004.

[12] S. Shen, A. M. Weiner, G. D. Sucha, and M. L. Stock, “Bit error rate performance of ultrashort-pulse optical CDMA detection under multi-access interferences,” Electron. Lett., vol. 36, no. 21, pp. 1795-1797, Oct 2000.

[13] B. Ramamurthy, D. Datta, H. Feng, J. P. Heritage, and B. Mukherjee, “Impact of transmission impairments on the teletraffic performance of wavelength-routed optical networks,” J. Lightw. Technol., vol. 17, no. 10, pp. 1713-1723, Oct. 1999.

[14] ITU D Study Group 2. (2005, Jan.). Handbook: Teletraffic Engineering, [Online]. Available: http://www.tele.dtu.dk/teletraffic/handbook/ telehook.pdf.

[15] J. S. Evans and D. Everitt, “On the teletraffic capacity of CDMA cellular networks,” IEEE Trans. Veh. Technol., vol. 48, no. 1, pp. 153–165, Jan. 1999.

[16] S. Goldberg and P. R. Prucnal, “On the teletraffic capacity of optical CDMA,” IEEE Trans. Commun., vol. 55, no. 7, pp. 1334-1343, Jul. 2007.

[17] M. Shtaif, M. Eiselt, and L. D. Garrett, “Cross-Phase Modulation distortion measurements in multispan WDM systems,” IEEE Photon. Technol. Lett., vol. 12, no. 1, pp 81-90, Jan. 2000.

[18] E. A. Golovchenko, V. J. Mazurczyk, D. G. Duff, and S. M. Abbott, “ Four-Wave Mixing Penalties in Long-Haul WDM Transmission Links,” IEEE Photon. Technol. Lett., vol. 11, no. 7, pp 821-823 Jul. 1999.

[19] G. –C. Yang, and W. C. Kwong, Prime codes with applications to CDMA optical and wireless networks. Norwood, MA, Artech House, 2002.

[20] S. V. Maric, M. D. Hahm and E. L. Titlebaum, “Construction and performance analysis of a new family of optical orthogonal codes for CDMA fiber-optic networks,” IEEE Trans. Commun., vol. 43, no 2/3/4, pp 485-489, Feb. 1995.

[21] P. R. Prucnal, M. A. Santoro, and T. R. Fan, “Spread spectrum fiber-optic local area network using optical processing,” J. Lightw. Technol., vol. 4, no. 5, pp. 547-554, May 1986.

[22] T. Pu, Y. Q. Li and S. W. Yang, “Research of algebra congruent codes used in two-dimensional OCDMA system,” J. Lightw. Technol., vol. 2, no. 11, pp 2557-2564, Nov. 2003.

[23] D. Gross, J. F. Shortle, J. M. Thompson and C. M. Harris, Fundamentals of queueing theory, Hoboken, New Jersy: John Wiley & Sons, Inc. 2008.

[24] G. Giambene, Queuing theory and telecommunications networks and applications. New York, NY Springer Science+Business Media, Inc. 2005.

Mohammad Gharaei received the B.Sc degree in electrical engineering from Azad University, Iran in 2005. In September 2006 he received M.Sc degree in digital telecommunication systems from Université Pierre et Marie Currie (Paris6), France. He is currently pursuing a Ph.D. degree in optical Networks in Institute Telecom, Telecom ParisTech, France. His research interests include the theory,

design and development of new concepts and architecture of future optical access networks, with focus on optical CDMA. Catherine Lepers received the Ph.D. degree in physics from the University of Lille, France, in 1993 and the diploma “Habilitation à Diriger des Recherches” (HDR) from the University of Lille, France, in 2007. In 1993, she joined the Nonlinear Optic Group, Laboratoire de Physique des Lasers,

Atomes et Molécules (PhLAM), University of Lille, as an Associate Professor and worked on Nonlinear Dynamics of Lasers. In 2000, she joined the Photonique Group in PHLAM to do some research on Fiber Bragg Gratings and on the characterization of Photonic Crystal Fibers. From 2002 to 2005, she has built collaboration between PHLAM and Institute Telecom, Telecom ParisTech France, performing advanced optical digital communications on code division multiple access (CDMA) techniques. Since 2006, she has some responsibility in SUPERCODE, an ANR contract concerning the generation of a Supercontinuum Source for OCDMA Application. In October 2008, she joined Telecom & Management SudParis, Institute Telecom, France where she is currently a Full Professor and Responsible of the group “Optics and Optoelectronics”. Her present research topics concern new concepts of optical access networks including PON OCDMA, hybrid PON WDM/OCDMA and Radio over Fiber networks. She is author or coauthor of 60 technical papers. Olfa Affes received the Engineer degree in telecommunication systems from ISI Ecole, Tunisia, Tunis. From February to July 2008 she realized her final project in optical Networks in Institute Telecom, Telecom Lille1, France.

She's currently working as an engineer at the company OWLIANCE. Philippe Gallion (M’82–SM’93) received the Doctorat de Troisième Cycle degree from the University of Reims, Reims, France, in 1975 and the Doctorat d’Etat degree from

the University of Montpellier, Montpellier, France, in 1986. He joined the Ecole Nationale Supérieure des Télécommunications (Telecom ParisTech), Paris, France, in 1978, where he is presently a Full Professor. He has made pioneering contributions on laser noise, injection locking, semiconductor laser

modulation chirp and tuning, coherent systems and optical devices, and digital optical communications systems and networks. His present research topics include theory, conception, modeling, and characterization of functional devices and their applications in advanced optical digital communication systems and networks. He is author or coauthor of more than 150 technical papers and

communications, and he has acted as Supervisor for more than 40 Ph.D. degree theses. Prof. Gallion is a member of the Optical Society of America. He is the Chairman of the IEEE Laser and Electro Optics Society (LEOS) French Chapter. He serves on the Editorial Board and Scientific Committee of several technical publications and as a member of Program or Steering Committee of international scientific meetings.