Multi-Carrier Approaches for Next-Generation Transmission: Why, Where and How?

Transcription

1 Multi-Carrier Approaches for Next-Generation Transmission: Why, Where and How? Sander L. Jansen Nokia Siemens Networks, Germany Sander Jansen, OFC 2012 tutorial Nokia Siemens Networks Acknowledgment Colleagues Dirk van den Borne, Maxim Kuschnerov, Antonio Napoli and Bernhard Spinnler, Nokia Siemens Networks Students at NSN Susmita Adhikari, Christian Albrechts Universität Kiel Beril Inan, Yingkan Chen and Ozgur Karakaya, Technical University Munich Adriana Lobato, University of Federal Armed Forces Munich Mohammad Alfiad and Vincent Sleiffer, University of Technology Eindhoven Alberto Diaz, Polytechnic University of Catalonia and in addition I would like to acknowledge: Itsuro Morita, Masatoshi Suzuki, Hideaki Tanaka, Hidenori Takahashi and Wei-Ren Peng, KDDI R&D Laboratories Dayou Qian, NEC William Shieh and Abdullah Al Amin, University of Melbourne Arthur Lowery and Liang Du, Monash University Sebastian Randel, Jeffrey Lee and Peter Winzer, Alcatel Lucent Andrew Ellis and Naoise McSuibhne, Tyndall National Institute Alberto Bononi, Universita di Parma Maurice O'Sullivan, Ciena 2 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 1

2 Agenda: Multi-Carrier Approaches for Next-Generation Transmission Introduction How? Multi-carrier modulation formats and concepts Where? Single mode transmission systems Multi mode transmission systems Why? Conclusion: Applicability of multi-carrier modulation formats 3 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Spectrally efficient transmission Time domain Standard Single Carrier Nyquist WDM (Spectrally shaped Single Carrier) Multi-carrier Blockwise transmission N-1 N N+1 Frequency domain Spectral shaping IFFT time time time frequency frequency frequency Straightforward to generate Wide spectrum -not spectrally efficient Narrow spectrum TX/RX complexity: Requires low pass filtering or DAC-based waveform generation. Narrow spectrum TX/RX complexity: Requires blockwise FFT/IFFT processing in electrical or optical domain 4 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 2

3 Multi-carrier systems: the definition Strictly speaking all DWDM transmission systems are multi-carrier: Conventional DWDM system In this presentation we focus on multi-carrier approaches for each individual DWDM channel 3-subcarrier OFDM High-subcarrier OFDM 5 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Multi-carrier concept Time-domain Frequency-domain length of one symbol Ch # f4 f3 amplitude 1 n sin(x)/x f2 f1 Σ time 6 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks frequency Multi-carrier representation in the digital domain: The k-th complex timedomain sample x l, k = 1 N cl, N n= 1 FFT size Modulation Subcarrier number n 2π nk exp j N Discrete timerepresentation Page 3

4 Multi-carrier modulation concept Many different methods exist to generate subcarriers rf source ~ amplitude 2x 3x 4x B0 B1 B2B3 B4 Baseband TXs Mixers frequency f 2f 3f 4f For the creation of many subcarriers, a more efficient way is to generate the subcarriers in the digital domain using an FFT 7 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Multi-carrier detection concept (1/2) In this example, subcarrier 2 is detected Frequency-domain Ch # 1 n f4 Time-domain f3 frequency Demodulation of an OFDM signal: Step 1: Downconversion f2 f1 OFDM symbol length For the parallel demodulation of all subcarriers, a more efficient way is to detect the subcarriers in the digital domain using an FFT 9 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 4

5 Multi-carrier detection concept (2/2) In this example, subcarrier 2 is detected Frequency-domain Ch # 1 n f4 Time-domain Average f3 f2 frequency Demodulation of an OFDM signal: Step 1: Downconversion Step 2: Integration over one symbol length f1 OFDM symbol length For the parallel demodulation of all subcarriers, a more efficient way is to detect the subcarriers in the digital domain using an FFT 10 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Agenda: Multi-Carrier Approaches for Next-Generation Transmission Introduction to multi-carrier modulation formats How? Multi-carrier modulation formats and concepts Where? Single mode transmission systems Multi mode transmission systems Why? Conclusion: Applicability of multi-carrier modulation formats 11 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 5

6 Fiber-optic OFDM detection methods Optical spectrum Electrical spectrum Direct detected Optical OFDM (DDO-OFDM) S Carrier B S OFDM B 2 E S Carrier *S Carrier B S OFDM *S OFDM S Carrier*S OFDM B S OFDM S Carrier *S Carrier 2 E S OFDM *S OFDM Coherent detected Optical OFDM (CO-OFDM) S Carrier B Local oscillator 90 hybrid S Carrier *S OFDM B Note: The OFDM-OFDM and Carrier-Carrier mixing terms can be removed for CO-OFDM by using balanced photodiodes 13 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks OFDM detection Direct detected (DD) OFDM Least components required at the receiver -> most cost-effective solution. Guard band required. At least 50% of power required for optical carrier -> Inherently 3-dB OSNR penalty. Optical carrier B B DD-OFDM -> short reach applications Coherent optical (CO) OFDM Superior transmission performance. Polarization dependent. Most complex setup to realize (phase noise compensation required) B CO-OFDM -> long-haul transmission systems. The focus in this tutorial will be on OFDM with coherent detection for long haul transmission systems 14 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 6

7 Coherent OFDM realizations Coherent WDM / All optical OFDM / No guard band OFDM FFT in the analog domain and one subcarrier is modulated per (integrated) modulator Baseband TXs rf source Laser ~ Modulator INT Integrated Coherent optical OFDM (CO OFDM) / Digital OFDM FFT in the digital domain and all subcarriers modulated at the same time Note that only two subcarriers are shown, for each additional subcarrier an extra modulator is required DACs IFFT Laser Modulator 15 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Coherent WDM transmitter and receiver Transmitter Serial/Par rf source ~ Modulator INT IQ modulator TX out Integrated Receiver: Matched Filter option 1: Analog domain RX in T/2 (local oscillator) 90 hybrid real imag PD PD LPF LPF ADC ADC j Matched Filter option 2: Digital domain Bulk CD comp. T/2 MIMO Clk recovery Freq offset Digital signal processing Phase recovery Demapping Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 7

8 Matched filter for coherent WDM Exactly one period per time slot Optical spectrum of transmitter Transmitter Reference Channel Comb Generator Ch1 Ch2 T Ch3 T Combiner Ch1 Ch2 Ch3 2/T Receiver: Impulse response Decision points Filtering T time No crosstalk! Integration window T/2 Waveform after filter and PD By filtering with a matched filter at the receiver the crosstalk is minimized. Matched filter can be realized in both the Analog or digital domain A.D. Ellis et al, PTL, Vol. 17, No. 2, (2005) 18 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Coherent WDM transmitter 13.5 Tb/s (135 x 111 Gb/s/ch) No Guard Interval Coherent OFDM Transmission over 6,248 km Native dual-subcarrier transmitter INT Integrated At the transmitter dual carrier QPSK signals with a data rate of 111 Gb/s are generated using a native dual-subcarrier modulator H. Masuda et al., PDPB5, OFC Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 8

9 Coherent WDM receiver in the electrical domain After coherent detection RX-based DSP processing After frequency correction RX-1 After subcarrier separation Transmission performance after 6,248 km RX-2 T/2 SC-2 The receiver is implemented with standard coherent detection. The FFT is implemented in the electrical domain Transmission of 135 x 111 Gb/s is achieved over 6,248 km H. Masuda et al., PDPB5, OFC Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Coherent WDM receiver in the optical domain (1/2) N 1 mn DFT: X m = exp j2 π xn, m = 0,.., N 1 n= 0 N Conventional Implementation Complexity = # Phase Shifters CStd. = N log 2( N) M. E. Marhic, "Discrete Fourier transforms by single-mode star networks," Opt. Lett., vol. 12, no. 1, pp , January 1987 (rel. phase diff. per column is fixed) Simplified Scheme Complexity smaller! C N DI = 1 D. Hillerkuss et al., "Simple all-optical FFT scheme enabling Tbit/s real-time signal processing," Opt. Express 18, (2010). 21 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 9

10 Coherent WDM receiver in the optical domain (2/2) 26 Tbit/s 325 Carriers DP-16QAM 50 km SSMF 20% CP D. Hillerkuss et al., "26 Tbit s -1 line-rate super-channel transmission utilizing all-optical fast Fourier transform processing," Nature Photon. 5, (2011). 22 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks CO-OFDM transmitter and receiver Transmitter Serial/Par Mapping Training Sym. zeros IFFT Par/Serial Cyclic prefix real imag DAC DAC LPF LPF IQ modulator TX out Digital signal processing Receiver: Channel estimation RX in 90 hybrid real imag PD PD LPF LPF ADC ADC j PNC type 1 Bulk CD comp. Symbol Sync. CP removal Serial/Par FFT zeros TS removal 1-Tap EQ PNC type 2 Demapping Par/Serial (local oscillator) Digital signal processing PNC: Phase noise compensation (or sometimes referred to as carrier recovery) 23 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 10

11 Selected high capacity transmission experiments Recent high capacity and high spectral efficiency transmission experiments over single core, single mode fiber. Capacity (excl. 7% FEC overhead) Rate/Ch (Gb/s) Modulation Type SE (b/s/hz) Reach (km) Published (sorted by year of publication) 852-Gb/s 114 8PSK SC OFC2008 PDP1 at&t/nec 1.14-Tb/s QAM CO-OFDM OFC2008 PDP2 KDDI 17-Tb/s 114 8PSK SC ECOC2008 PDE2 at&t/nec 34-Tb/s 114 8QAM SC OFC2009 PDPB4 at&t/nec 64-Tb/s QAM SC OFC2010 PDPB9 at&t/nec 69-Tb/s QAM Coh. WDM OFC2010 PDPB7 NTT 11.2-Tb/s QAM Coh. WDM ECOC 2010 PD2.4 NTT Single ch QAM SC ECOC 2010 PD2.3 Tohoku U Tb/s QAM CO-OFDM OFC2011 PDPB5 NEC Highest capacity over single mode fiber has been obtained using (digital) OFDM Based on a slide by Dayou Qian from NEC 24 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks 107 Tb/s transmission over 165 km fiber 370 CW lasers with 25GHz channel space for C- and L-band. A total nominal capacity of 107 Tb/s is obtained by modulating each 25-GHz channel with: 107 Tb/s 370 = 289 Gb/s This provides a spectral efficiency of: 289 Gb/s 25 GHz = 11.6 bit/s/hz Modulation format: PDM-128QAM (achieves up to14 bits/symbol, allowing 20% margin for overhead) 0 OFDM allows because of its rectangular signal shape an extremely close channel spacing Power (dbm) Power (dbm) (f) (B A) B A (even Ch) Wavelength (nm) D. Qian, et al., OFC 2011, PDP B5 (D C) D C (odd Ch) Wavelength (nm) 25 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 11

12 Transmission Results All channels achieved BER below FEC threshold of (7% hard-decision FEC) Uneven noise floors of the transmitter-side EDFAs affect the performance of edge channels of C- and L-bands. D. Qian, et al., OFC 2011, PDP B5 26 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Algorithms to enable high constellation sizes In the 128 QAM transmission experiment reported by D. Qian three algorithms are used to enable such high constellations: IQ imbalance compensation Laser phase noise compensation Electrical phase noise compensation Electrical phase noise IQ imbalance Laser phase noise FIR DAC DAC LPF LPF ~ Baseband transmitter hxx hxy hyx hyy RF pilots generated by the DACs are used to comp. el. phase noise An FIR with static taps is used to compensate for IQ imbalance Laser phase noise is compensated for with an RF pilot tone D. Qian, et al., OFC 2011, PDP B5 27 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 12

13 IQ imbalance compensation IQ imbalance is an imbalance in phase and/or power between the in phase and quadrature component of the OFDM signal. Origin of IQ imbalance are unequal phase response in components: DACs, LPFs, amplifiers, etc or unequal path length for the I and Q path. In general IQ imbalance is static and can therefore be compensated for using an FIR filter in the digital domain, right before the DACs. FIR filter structure ~ Baseband transmitter hxx hxy hyx hyy DAC DAC LPF LPF IQ imbalance can be precompensated for in the digital domain Difference in the gain and phase response of the amps Different path length of the I and Q paths 28 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Laser phase noise: RF Pilot (RFP) phase noise comp. The 128 QAM constellation of the work by D. Qian is enabled by several performance enhancement techniques. RFP-based phase noise compensation is used as the basis for phase noise compensation. RF pilot RFP compensation technique OFDM signal Received Signal (after coherent detection) conj( ) S. L. Jansen, et al., JLT 2008, vol. 26, pp Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 13

14 Laser phase noise: Illustration of compensation algorithms Adapted from Wireless systems This figure shows the actual and the estimated phase as a function of time for CPE and RFP (laser linewidth is 200kHz). CPE assumes that the phase is constant during an OFDM symbol and as a result, a high BER penalty is present for wide laser linewidths RFP follows the phase of the LO within the OFDM symbol and can as a result track significanlty wider laser linewidths than CPE 31 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Laser phase noise: tolerance of compensation algorithms OSNR Penalty [db] FFT Size = 4096 FFT Size = 2048 FFT Size = 1024 FFT Size = 512 FFT Size = 256 FFT Size = 128 Type 1: RFP 128 FFT size FFT size LO linewidth [Hz] 1.5 Reference: 1 Method 1 (RFP) with FFT size LO linewidth [Hz] Parameters: 100G POLMUX OFDM with 4-QAM constellation size Type 1: RF-pilot phase noise compensation (RFP) High phase noise tolerance and requires only a small overhead More computationally complex than CPE Type 2: Common phase estimation (CPE) Proven concept from Wireless + least complex implementation Requires ~10% extra OFDM overhead for subcarrier pilots 32 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks OSNR Penalty [db] FFT Size = 4096 FFT Size = 2048 FFT Size = 1024 FFT Size = 512 FFT Size = 256 FFT Size = FFT size S. Randel et al., vol. 22, pp , 2010 Type 2: CPE 128 FFT size Page 14

15 Electrical phase noise compensation Receiver (schematic provided by D. Qian) Optical phase noise compensation (conventional RFP) Electrical phase noise compensation Input Down sampling Output BPF Conj Select RF-tone pilot Coarse carrier recovery BPF Compare with training signal Select pilot sub-carriers Fine carrier recovery Conj Electrical phase noise can be monitored by adding an RF pilot in the electrical domain with a DAC Compensation at the receiver is done the same way as optical RFP 33 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Electrical phase noise comp: SFO tolerance -10 Pilot Subcarriers 0-5 RF Pilot Without pilot base PNC 4 QAM: 70.2Gb/s Intensity (dbm) Frequency (GHz) Yingkan Chen et al., in proc. OFC ppm 100 ppm 200 ppm ppm ppm 1200 ppm 1300 ppm wo comp. 100 ppm OSNR (db) 34 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks BER OSNR (db) Sampling frequency offset correction is required for OFDM With pilot based PNC With the electrical phase noise compensation algorithm a large tolerance with respect to SFO can be obtained 1-dB OSNR penalty at 10e -3 allows for a sampling frequency offset of about 1000 ppm Page 15

16 Agenda: Multi-Carrier Approaches for Next-Generation Transmission Introduction to multi-carrier modulation formats How? Multi-carrier modulation formats and concepts Where? Single mode transmission systems Multi mode transmission systems Why? Conclusion: Applicability of multi-carrier modulation formats 35 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks CO-OFDM versus single carrier 1600 Maximum Distance [km] Launch Power [dbm] G. Bosco, et al, PTL, pp , 2010 A. Diaz et al., in IEEE Photonics, 2011 Clean comparison difficult to make as it is dependent on many parameters. Most investigations done so far agree that spectrally shaped single carrier outperforms CO-OFDM 36 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 16

17 How can the nonlinear tolerance of OFDM be improved? PAPR reduction B. Goebel, et al., OFC 2008, JWA58 B. S. Krongold, et al., ECOC 2008, P.4.13 Intensity-based SPM comp A. J. Lowery, PTL 2007 pp W. Shieh, et al., Opt. Expr. 2007, pp X. Liu, JLT 2009, pp RFP-based nonlinear compensation B. Inan, et al., ECOC 2010, Tu.4.A.6 L. Du, et al., ECOC 2011, Th.11.B Symbol rate optimization T. Kobayashi et al. JLT 2009, pp Tang, et al., OFC 2010, JThA6 Back propagation See reference list for full listing For instance: L. Du, et al., OFC 2010, OTuE2 Simple, but limited improvement Holy grail, but too complex 37 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Intensity based SPM compensation Signal Power Phase distortion TX Dispersion Optical channel Dispersion NLT NLT NLT Pre compensation RX Post compensation Q [db] A. Lowery, Optics Express, vol 15, Intensity-based SPM comp No comp Power / Wavelength [dbm] SPM induces an intensity dependent phase shift Pre- and post compensation is a simple form of SPM compensation using a 1-step intensity dependent phase shift compensator at TX and RX Dispersion changes the waveform and reduces efficiency of the algorithm 38 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 17

18 RFP-based nonlinear compensation B. Inan, et al, ECOC 2010, Tu.4.A.6 Power (dbm) Frequency (GHz) Pilot Phase (rad) 0 pi/4 Nonlinearities Nonlinearities + ASE -pi/ Time (OFDM Symbols) Signal in Signal out LPF ( ) * Select RF-Pilot Conj. XPM is a narrow band nonlinear effect that is monitored by the RF pilot in the middle of the OFDM spectrum It has been shown that XPM can partly be compensated for by RFP-based nonlinear compensation 39 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Nonlinear Effects on RF-Pilot Linear transmission (- 10 dbm launch power) Nonlinear transmission (- 3 dbm launch power) Electrical spectrum of the RF-aided pilot after 3200 km transmission at 1.25 khz resolution bandwidth. S. L. Jansen et al., JON, Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 18

19 Combination of intensity based SPM and RFP comp. Q [db] Intensity-based SPM comp only No comp Power / Wavelength [dbm] Q [db] Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Q [db] No comp SPM compensation + RFP No comp Power / Wavelength [dbm] Power / Wavelength [dbm] RFP only L. Du et al., Opt. Lett., vol. 36, pp , 2011 Intensity-based SPM compensation and RFP combined provide significant mitigation of nonlinearities Symbol rate optimization: Nonlinear tolerance vs subcarrier spacing Typical range for Digital OFDM Typical range for Analog OFDM Typically the subcarrier spacing of Digital OFDM is due to the relatively high FFT size significantly lower than that of analog OFDM. Especially for the dispersion managed maps (Map 1 and Map 2) a large penalty is observed for narrow sub-carrier spacing. How can the nonlinear tolerance of digital OFDM be improved?. A.D. Ellis et al., in proc OECC 2010, Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 19

20 OFDM in dispersion managed links (time domain explanation) Without inline dispersion compensation The chromatic dispersion changes the envelope of the waveform and with that the location of the peaks Chromatic dispersion de-correlated the phase shifts induced by SPM and XPM -> nonlinear phase shifts are spread over the OFDM symbol Transmitter Middle of the link Nonlinear Receiver Nonlinear phase shift Nonlinear phase shift phase shift With dispersion compensating modules The envelope of the waveform is the same for all nonlinear regions The phase kinks of SPM and XPM are highly correlated over all nonlinear regions -> coherent phase kinks destroy the OFDM symbol Nonlinear phase shift Nonlinear phase shift Nonlinear phase shift 43 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Symbol rate optimization: Subcarrier bandwidth optimization OSNR penalty (db) for 1e-3 BER Number of subcarriers per WDM channel Configuration: 55.5-Gb/s in a single polarization 7 WDM channels at 50-GHz spacing 2,000-km SSMF transmission: no inline DCMs 80-km spans Input powers Single channel: 2dBm WDM: 0dBm/ch When more than 32 subcarriers are used (subcarrier BW < 870 MHz) a steep increase in nonlinear penalty is observed How can the nonlinear performance of digital OFDM be improved where typically 500 subcarriers or more are used? T. Kobayashi, et al. in JLT, vol. 27, pp , Sander Jansen, OFC 2012 Tutorial. Nokia Siemens Networks Page 20

21 DFT-spread OFDM: Improving the nonlinear tolerance Standard OFDM transmitter Single carrier Multi carrier IFFT Adding an extra FFT at the transmitter inverts (cancels) IFFT operation Single carrier Single carrier FFT IFFT Combining smaller FFTs with one IFFT: flexible subcarrier bandwidth Single carrier FFT FFT IFFT Multi carrier Single carrier IFFT Multi carrier With DFT-spread OFDM the nonlinear tolerance of digital OFDM can be improved. The concept derived from wireless communication systems relies on the fact that by adding a set of FFTs at the transmitter the subcarrier bandwidth can be increased. 45 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Multi-band DFT-Spread OFDM for Nonlinearity Mitigation Transmission perform of MB-DFT-S-OFDM at 400 Gb/s over 1000-km SSMF fiber Q (db) dbm launch power Number of Bands Q (db) (b) 3-channel WDM MB-DFT-S-OFDM SC MB-C-OFDM Fiber Launch Power (dbm) MB-DFT-S-OFDM: Multiband DFT-Spread OFDM SC: Single-Carrier MB-C-OFDM: Multiband Conventional OFDM DFT-spread OFDM in each subband generates a single-carrier signal in each subband, reducing PAPR in each subband. This, in return, improves the nonlinearity performance. Y. Tang et al., in proc OFC 2010, JThA6 46 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 21

22 Agenda: Multi-Carrier Approaches for Next-Generation Transmission Introduction to multi-carrier modulation formats How? Multi-carrier modulation formats and concepts Where? Single mode transmission systems Multi mode transmission systems Why? Conclusion: Applicability of multi-carrier modulation formats 47 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Multi-mode transmission systems MMF core >> SMF core Nonlinear tolerance MMF is significantly larger than that on SMF. Main challenge for MMF: Compensation of differential mode delay and mode mixing 1-mode transmission on multi-mode fiber (MMF) F. Ferreira et al., in PTL, to be published Digital ID: /LPT TX 1 TX 2 TX N Multi-mode fiber (MMF) RX 1 RX 2 RX N N x N MIMO processing Illustration of modal delay and mode mixing in multi-mode fiber 48 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 22

23 Channel estimation Blind channel estimation Training symbol channel estimation Modulation Coherent det h xx h xy h yx h yy IFFT Modulation Coherent det FFT h xx h xy h yx h yy Multi-tap equalizer Single tap equalizer Many different algorithms exist for channel estimation and equalization. These algorithms can be divided into blind channel estimators and one that use training symbols Blind channel estimation Equalization in the time domain Most common algorithm for single carrier and coherent WDM Training symbol channel estimation Equalization in the Frequency domain Most common algorithm for CO-OFDM 49 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Complexity comparison for single-mode fiber 25 Single TDE TDE / TDE FDE / TDE # complex mults per bit Blind channel estimation Training symbol channel estimation CD [ps/nm] Single FDE OFDM Fixed equalizer type All training symbol channel estimation algorithms are less complex than blind channel estimation algorithms. The difference in complexity is up to a factor of 3x Spinnler, JSTQE, vol. 16, no. 5, Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Dynamic equalizer type Configuration: Net datarate: 100Gb/s QPSK modulation DGD tolerance: 100ps Oversampling rate: 1.5x Page 23

24 Maximum Reach with TS based Approach 10 5 Blind channel estimation 10 5 Training symbol channel estimation # complex mults per bit Blind equalization 3-mode MMF Blind equalization SMF Distance [km] Distance [km] The modal dispersion assumed for the FMF is 27 ps/km (the lowest modal dispersion reported to date in [xxx] For blind equalization a 10-fold+ increase in complexity is observed whereas with training symbols the increase in complexity is ~2x The max transmission distance for OFDM is 689 km due to the 10% overhead constraint. A lower modal disp. is required for further reach. 51 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Training symbol equalization 3-mode MMF B. Inan et al., in proc. OFC 2012 Max reach OFDM on FMF (with 10% OFDM-overhead) Training Symbol equalization SMF Conclusion: Why use multi-carrier modulation formats? Single-mode fiber (SMF) Because of its well defined spectral shape CO-OFDM is the modulation format used to realize the highest capacity transmission over single mode fiber For long-haul transmission systems OFDM does not offer significant advantages in terms of nonlinear performance without the use of nonlinear compensation techniques Various interesting schemes exist to significantly improve the nonlinear tolerance of multi-carrier modulation formats. Multi-mode fiber (MMF) MMF transmission systems are limited by mode-group dispersion rather than nonlinearities Training symbol based modulation formats provide a significant reduction in complexity. OFDM is therefore ideally suited for MMF transmission systems 52 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 24

25 Thank you for your attention Sander Lars Jansen Website: 53 Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks References (1/3) (sorted by publication date) CO-OFDM (digital OFDM) W. Shieh, et al., Electronics Letters, vol 43, pp , S.L. Jansen, et al., Gb/s PDM-8QAM-OFDM Transmission at 4-b/s/Hz Spectral Efficiency, in PTL, vol 21, pp , 2009 R. Dischler and F. Buchali, Transmission of 1.2 Tb/s Continuous Waveband PDM-OFDM-FDM Signal with Spectral Efficiency of 3.3 bit/s/hz over 400 km of SSMF, OFC 2009, PDPC2. X. Liu, et al, Single coherent detection of a 606-Gb/s CO-OFDM signal with 32-QAM subcarrier modulation using 4 80-Gsamples/s ADCs, in proc. ECOC 2010, PD2.6 D. Qian, et al., Tb/s ( Gb/s) PDM-128QAM-OFDM Transmission over 3 55-km SSMF using Pilot-based Phase Noise Mitigation, OFC 2011, PDP B5 Coherent WDM / All-optical OFDM H. Sanjoh, et al, Optical orthogonal frequency division multiplexing using frequency/time domain filtering for high spectral efficiency up to 1 bit/s/hz, in Proc. OFC 02, Anaheim, 2002, Paper ThD1. A. D. Ellis and F. C. G. Gunning, Spectral density enhancement using coherent WDM, IEEE Photon. Technol. Lett., vol. 17, no. 2, pp , A. Sano, et al, No-Guard-Interval Coherent Optical OFDM for 100-Gb/s Long-Haul WDM Transmission, Journal of Lightwave Technology, Vol. 27, pp , S. Chandrasekhar and Xiang Liu, Experimental investigation on the performance of closely spaced multi-carrier PDM-QPSK with digital coherent detection, Optics Express, Vol. 17, pp , D. Hillerkuss et al., "Simple all-optical FFT scheme enabling Tbit/s real-time signal processing," Opt. Express 18, , 2010 J. Zhao, A. Ellis, Electronic Impairment Mitigation in Optically Multiplexed Multi-Carrier Systems, Journal of Lightwave Technology, vol. 29, pp , 2011 Direct detection OFDM: different forms of direct detection OFDM I.B. Djordjevic, et al., Opt. Expr., vol 14, pp , 2006 Arthur Lowery, Liang Du, and Jean Armstrong, Performance of Optical OFDM in Ultralong-Haul WDM Lightwave Systems, Journal of Lightwave Technology, Vol. 25, pp , M. Schuster, S. Randel, C.-A. Bunge, S.C.J. Lee, F. Breyer, B. Spinnler, and K. Petermann. Spectrally Ecient Compatible Single-Sideband Modulation for OFDM Transmission with Direct Detection, Photonics Technology Letters, Vol. 20, W.R. Peng, X. Wu, V.R. Arbab, K-M Feng, B. Shamee, L.C. Christen, J-Y Yang, A.E. Willner, and S. Chi, Theoretical and Experimental Investigations of Direct-Detected RF-Tone-Assisted Optical OFDM Systems, in Journal of Lightwave Technology, Vol. 27, pp , A. Lowery and L. Du, Optical orthogonal division multiplexing for long haul optical communications: A review of the first five years, Volume 17, 2011, Pages Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 25

26 References (2/3) (sorted by publication date) Phase noise compensation L. Tomba, On the Effect of Wiener Phase Noise in OFDM Systems, Trans. on Com, Vol 46, pp , 1998 Xingwen Yi, et al., Phase Estimation for Coherent Optical OFDM, Photonics Technology Letters, vol. 19, 2007 S.L. Jansen, et al., Coherent Optical 25.8-Gb/s OFDM Transmission over 4,160-km SSMF, Journal of Lightwave Technology, vol. 26, pp. 6-15, X. Yi, W Shieh, and Y. Ma, Phase Noise Effects on High Spectral Efficiency Coherent Optical OFDM Transmission, Journal of Lightwave Technology, Vol. 26, pp , 2008 S. Randel, et al., "Analysis of RF-Pilot-based Phase Noise Compensation for Coherent Optical OFDM Systems," Photonics Technology Letters, vol. 22, pp , 2010 G. Colavolpe, T. Foggi, E. Forestieri, and M. Secondini, Impact of Phase Noise and Compensation Techniques in Coherent Optical Systems, Journal of Lightwave Technology, vol. 29, pp , Nonlinear tolerance A.J. Lowery, et al., Calculation of power limit due to fiber nonlinearity in optical OFDM systems, Optics Express, vol. 15 no. 20, pp , Arthur Lowery, Fiber nonlinearity pre- and post-compensation for long-haul optical links using OFDM, Optics Express, Vol. 15 Issue 20, pp (2007) K. Forozesh, et al., "The influence of the dispersion map in coherent optical OFDM transmission systems," in Digest of the IEEE/LEOS Summer Topical Meetings, (Acapulco, Mexico, 2008), pp Liang Du and Arthur Lowery, Fiber Nonlinearity Compensation for CO-OFDM Systems with Periodic Dispersion Maps, in proc OFC 2009, OTuO1 T. Kobayashi, et al. Over 100 Gb/s Electro-Optically Multiplexed OFDM for High-Capacity Optical Transport Network, in Journal of Lightwave Technology, vol. 27, pp , 2011 Yan Tang, et al. Fiber Nonlinearity Mitigation in 428-Gb/s Multiband Coherent Optical OFDM Systems in proc. OFC 2010 B. Inan, et al., "Pilot-tone-Based Nonlinearity Compensation for Optical OFDM Systems," in Proc. European Conference on Optical Communications (ECOC) 2010, Tu.4.A.6. G. Bosco, A. Carena, V. Curri, P. Poggiolini, and F. Forghieri, Performance Limits of Nyquist-WDM and CO-OFDM in High-Speed PM-QPSK Systems, Photonics Technology Letters, vol. 22, pp , 2010 L. B. Du and A. J. Lowery, "Pilot-based cross-phase modulation compensation for coherent optical orthogonal frequency division multiplexing long-haul optical communications systems," Opt. Lett., vol. 36, pp , Y. Chen, et al., Pilot-aided Sampling Frequency Offset Compensation for Coherent Optical OFDM, in proc OFC Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks References (3/3) (sorted by publication date) Processing complexity of optical OFDM S.J. Savory, Digital Signal Processing Options in Long Haul Transmission, in proc. OFC 2008, OTuO3. P. Poggiolini, et al., Evaluation of the computational effort for chromatic dispersion compensation in coherent optical PM-OFDM and PM-QAM systems Optics Express, Vol. 17, pp , B. Spinnler, Equalizer Design and Complexity for Digital Coherent Receivers, JSTQE, vol. 16, no. 5, B. Inan, et al., Equalizer Complexity of Coherent Receivers for Three Mode Fiber Systems, in proc OFC History of OFDM Chang, R. W. Synthesis of band-limited orthogonal signals for multi-channel data transmission, Bell System Technical Journal, Vol. 46, , S. Weinstein and P. Ebert, Data Transmission by Frequency-Division Multiplexing Using the Discrete Fourier Transform, IEEE Transactions on Communication Technology, Vol. 19, pp , 1971 L. Cimini, Analysis and Simulation of a Digital Mobile Channel Using Orthogonal Frequency Division Multiplexing, IEEE Transactions on Communication Technology, Vol. 33, pp , 1985 OFDM principles, understanding OFDM R. van Nee and R. Prasad, "OFDM for Wireless Multimedia Communications", Artech House, T.C.W. Schenk, RF Imperfections in High-rate Wireless Systems, Springer B. Shieh and Ivan Djordjevic, OFDM for Optical Communications, Elsevier OFDM online Optical OFDM online Sander Jansen, OFC 2012 Tutorial Nokia Siemens Networks Page 26