**7. Conclusion**

20 Selected Topics on Optical Amplifiers in Present Scenario

results clearly show the correlation between the penalty and the chirp. The latter has more pronounced effects over long SMF. In Figure 16, a comparison between single and bielectrode RSOA over 100 km transmission is shown and a difference of 4.1 dB is obtained at the FEC limit. We can clearly see that the eye diagram starts to be closed due to the chirp on single-electrode devices over long distances. This effect is much reduced when using bielectrode RSOAs which confirms BER measurements. The proposed network design allows the use of Dense-WDM (DWDM) which means 62 wavelengths considering the 50 nm optical bandwidth of the RSOA. By considering the passive splitter (four clients), 248 potential suscribers can be feeded. At the FEC level, a variable attenuation of 4 dB is obtained which can be use as a splitter in order to design two parallels WDM PONs

It was shown that the large gain of the RSOA and also the low chirp allows a reach extension of the link from standard 20km to 100 km. We demonstrated that penalties due to the transmission over 100 km SMF at 2.5 Gbit/s are reduced using an optimized multielectrode device and a BER below the FEC limit was achieved. We also believe this effect

Active research on high bit rate RSOA has led to 10 Gbit/s operation with EDC (Torrientes et al., 2010), OFDM (Duong et al., 2008) or electronic filtering (Schrenk et al, 2010). Bandwidth improvement to 7 GHz small-signal bandwidth with dual-electrode devices have been obtained but no large signal operation (Brenot et al, 2007). However the modulation bandwidth of one-section RSOA is limited to 2 GHz and increasing the modulation bandwidth of RSOA is still a challenge. Since carrier lifetime is mainly governed by stimulated emission rate, we have decided to increase the length of our RSOA to increase photon density, hence reducing carrier lifetime (de Valicourt et al., 2011). This device was chosen because an open eye diagram was obtained when the RSOA was driven by a 27-1 PRBS at 10 Gbit/s (Figure 17. (a)). The experimental set-up used for the 10 Gbit/s modulation is the same as represented in Figure 13. An ECL is used to launch a 4.5 dBm CW signal into the system through an optical circulator (OC). The signal is coupled into the RSOA which is modulated and generates the upstream signal. The RSOA is driven by a 27-1 PRBS at 10 Gbit/s, with a DC bias of 160 mA. The upstream signal propagates on various SMF lengths. A variable optical attenuator is placed in front of the receiver in order to analyze the performance of the system as a function of the received power. BER measurements are done using an APD receiver and an error analyzer. BER measurements

With optical injection, BER values below the FEC limit in BtB and after 2km transmission are obtained (Figure 17. (b)). Error-free operation can either be obtained with FEC codes, or under certain optical injection regimes. However we can clearly see that the eye diagram tends to be closed due to the chirp over long distances. Multi-electrode devices can be used

As described in section 4, the modulation speed of RSOA is limited by the carrier lifetime. In the large signal regime, the slow decay is probably governed by the no-stimulated recombination process, which increases the carrier lifetime. A 3 GHz modulation bandwidth can be obtained with 850 µm long RSOA, which has led to the first eye-opening of a RSOA at 10 Gbit/s without electrical equalization or passive electronic filtering. Limitation due to

in order to compensate for this effect as demonstrated in the previous section.

will be even more pronounced when 10 Gbit/s RSOA will be used.

without ECL have led to a BER floor of 10-6 (ASE regime).

**6.2 Reaching 10 Gbit/s modulation without any electronic processing** 

(2\*248=496 customers) over 100Km.

Nowadays, research in the telecom area is partly focused on passive optical network architecture. WDM-PON seems to be a promising approach allowing high data bit rate and flexibility. WDM techniques used in long-haul systems are now mature, however the shift to local access networks is more challenging. New requirements appear such as cost reduction, the need for new key devices at the ONU and compatibility with the existing network. Colourless ONU are necessary to obtain cost effective architecture and RSOA is one potential solution. In this chapter, we focused on these devices.

The SOA theory has been discussed and applied to Reflective SOA devices. We underline several physical mechanisms that are responsible for the carrier density variation. The stimulated, radiative and non-radiative recombination rates are described. A model has been developed, taking into account several longitudinal sub-sections of the active guide. RSOAs exhibit a non-homogeneous carrier density profile which strongly affects the overall gain. At the input/output of the device, a strong saturation effect is observed. Therefore the net gain needs to be carefully integrated along the device taking in account this nonhomogeneity. All these results confirm the presence of key parameters such as the length and the optical confinement which should lead to design rules.

To assess the RSOA dynamics, the carrier lifetime is estimated. The E/O modulation bandwidth mainly depends on this parameter, for instance shorter carrier lifetime induces larger 3 dB E/O modulation bandwidth. The reduction of the carrier lifetime is required to obtain high speed RSOAs.

Next Generation of Optical Access Network Based on Reflective-SOA 23

Connelly M. J., "Wide-Band Steady-State Numerical Model and Parameter Extraction of a

D'Alessandro D., Giuliani G., & Donati S. (2001), "Spectral gain and noise evaluation of SOA

de Valicourt G., Maké D., Fortin C., Enard A., Van Dijk F., & Brenot R. (2011), "10Gbit/s

de Valicourt G., Make D., Lamponi M., Duan G., Chanclou P. , & Brenot R. (2010a), "High

de Valicourt G., Pommereau F., Poingt F., Lamponi M., Duan G.H., Chanclou P., Violas M.

de Valicourt G., Lamponi M., Duan G.H., Poingt F., Faugeron M., Chanclou P., & Brenot R.

de Valicourt G., Maké D., Landreau J., Lamponi M., Duan G.H., Chanclou P. , & Brenot R.

Feuer M., Wiesenfeld J., Perino J., Burrus C., Raybon G., Shunk S., & Dutta N. (1996),

Healey P., Townsend P., Ford C., Johnston L., Townley P., Lealman I., Rivers L., Perrin S. , &

Iezekiel S., Elamaran B., & Pollard R.D. (2000), "Recent developments in lightwave network

Kao K. C. & Hockham G. A. (1996), Dielectric-fibre surface waveguides for optical

Koponen J.J., & Söderlund M.J. (2004), "A duplex WDM passive optical network with 1:16

Lee H-H., Cho S-H., Lee J-H., Jung E-S., Yu J-H., Kim B-W., Lee S-S., Lee S-H., Koh B-H.,

architecture," *in Proc. ECOC*, Th.3.F, Sep. 2008, Brussels, Belgium

SOAs", *Electron. Lett*., Vol. 37, No. 19, Sept. 2001, pp. 1181 – 1182

frequencies, *Proc. IEE*, Vol. 113, No. 7, July 1966, pp. 1151-1158

Anaheim, USA

pp. 125-130

Torino, Italy

247-257

San Diego, USA

*Electroni.*, Vol. 43, No. 1, Jan. 2007, pp. 47-56

*technology letters*, Vol. 22, No. 3, Feb. 2010, pp. 191-193

OThT2, 2011, Los Angeles, USA

No. 19, Oct. 1, 2010, pp. 1425-1427

Vol.8, No.9, Sept. 1996, pp.1175-1177

March 2004, Los Angeles, USA

modulation at 2.5 and 5 Gbit/s for WDM PON", *in Proc. OFC*, OWD1, 2007,

Tensile-Strained Bulk Semiconductor Optical Amplifier," *IEEE J. Quantum* 

and SOA-based switch matrix", *IEE Proc.-Optoelectron*., Vol. 148, No. 3, June 2011,

modulation of Reflective SOA without any electronic processing", *in Proc. OFC'11*,

Gain (30 dB) and High Saturation Power (11dBm) RSOA Devices as Colourless ONU Sources in Long Reach Hybrid WDM/TDM -PON Architecture", *Photonics* 

A., & Brenot R. (2010b) "Chirp Reduction in Directly Modulated Multielectrode RSOA Devices in Passive Optical Networks", *Photonics technology letters*, Vol. 22,

(2010c), " First 100 km uplink transmission at 2.5 Gbit/s for hybrid WDM/TDM PON based on optimized bi-electrode RSOA", *in Proc. ECOC'10*, Tu.5.B.6, 2010,

(2009), "New RSOA Devices for Extended Reach and High Capacity Hybrid TDM/WDM -PON Networks", *in Proc. ECOC'09*, P9.5.2, 2009, Vienna, Austria Duong T., Genay N., Chanclou P., Charbonnier B., Pizzinat A., & Brenot R., "Experimental

demonstration of 10 Gbit/s upstream transmission by remote modulation of 1 GHz RSOA using adaptively modulated optical OFDM for WDM-PON single ber

"Single-port laser-amplifier modulators for local access", *Photonics technology letters*,

Moore R. (2001), "Spectral slicing WDM-PON using wavelength-seeded reflective

analysis", *Engineering Science and Education Journal*, Vol. 09, No. 06, Dec. 2000, pp.

power split using reflective SOA remodulator at ONU", *in proc. OFC'04*, MF 99,

Sung J-S., Kang S-J., Kim J-H., & Jeong K-T. (2009), "First Commercial Service of a Colorless Gigabit WDM/TDM Hybrid PON System", *in Proc. OFC*, PDPD9, 2009,

Potential cost effective solutions for next generation of access network could be based on RSOA devices. Therefore, research on RSOA devices is driven by WDM-PON applications. It is of prime interest to solve issues related to this application. RSOAs as colourless ONU have been investigated for access network. High performances RSOAs enable an upstream transmission of 8 WDM channel at 2.5 Gbit/s over 45 km. A high optical budget (36 dB) was demonstrated.

The chirp remains one major limiting factor as well as the modulation speed. 2-section RSOAs were used to overcome the first drawback. The use of 2-section RSOAs allows a 100 km transmission below the FEC limit at 2.5 Gbit/s. Finally, long RSOA allows performing the first direct 10 Gbit/s modulation with open eye diagram thanks to the E/O modulation bandwidth increase.

Therefore, as a general conclusion, RSOAs show a great potential as a next generation of optical transmitter. It is a colorless device which can be used in WDM access networks. However the modulation speed is still limited and 10 Gbit/s modulation needs to be realised over a minimum of ten kilometres.

Tech-eco analysis has to be performed in order to evaluate the different technologies for WDM-PON and a trade-off between performances and cost will determine the future of optical access network. RSOA are still limited in terms of performances and architecture but new approach such as self-seeding could overcome these main issues.
