Flexible Bandwidth Terabit Coherent Optical Flexible ......Evolution of Optical Networking Exponential increases in data traffic have transformed the landscape of modern networks

October 2011 IEEE PhotonIcs socIEty nEWsLEttER 5

Research Highlights

Flexible Bandwidth Terabit Coherent Optical Communication Networks by Optical Arbitrary

Waveform Generation and MeasurementS. J. Ben Yoo

UNIVERSITY OF CALIFORNIA, DAVIS, CA 95616

1

Flexible Bandwidth Terabit Coherent Optical Communication Networks by Optical Arbitrary Waveform Generation and Measurement

S. J. Ben Yoo

University of California Davis, CA 95616

Evolution of Optical Networking Exponent ia l increases in data t raf f ic have transformed the landscape of modern networks. In the latter part of the 1990s, the exponential growth in capacity demands were successfully met by the wavelength-division-multiplexing (WDM) technology which had become commercially available by then.This initial phase of optical networking, often referred to as the first generation optical networking deployed in 1997, focused on point-to-point link capacity increases. Beyond the simple capacity increase, the true benefit of optical networking may arise from the reconfigurabil i ty of such vast bandwidths directly in the optical layer without involving electronics in the data plane. The second-generation optical networking achieves reconfiguration of optical wavelength circuit paths (lightpaths) by properly configuring the optics within the optical network elements. It also supports format and protocol transparency, and simplifies the hardware requirements in the data plane. With the on-going deployment of reconfigurable optical add drop multiplexers (ROADMs) and optical crossconnects (OXCs), we anticipate the third-generation optical networking to address dynamic reconfiguration of even higher capacity traffic. The third generation optical networking is expected to appear in two possible configurations. One is a packet- agile optical networking technology and the other is a flow-agile optical networking technology. In both configurations, extremely high data rates must be supported on each strand of fiber. In practical implementations, usable bandwidth of a single mode fiber with amplification is limited by the bandwidth of amplification technologies, which is on the order of 5~10 THz for commonly used erbium-doped fiber amplifiers (EDFAs). In order to achieve high capacity communications in the limited spectral bandwidth, the recent trend in optical communications has been to adopt advanced modulation formats that achieve high spectral

efficiency together with dense WDM (DWDM). The historical trend of this evolution appears in Figure 1 and Figure 2 [1].

Channel Bitrate

Num

ber o

f WD

M C

hann

els

TOTAL CAPACITY128

64

32

16

8

4

2

1

10 Mb/s 100 Mb/s 1 Gb/s 10 Gb/s 100 Gb/s 1 Tb/s

'89

ONTC-1 '94

ONTC-2 '95

AT&T '85

Bellcore '88

NTT '90

AT&T '94

AT&T '94

'87 '91'86'85'83'80

Transmission

Networking

Prods.

NortelAlcatel

CienaLucent

Cambrian

HitachiMCI/Hitachi ‘97

Fujitsu ‘97AT&T ‘97

NTT ‘97

Ciena

Lucent

CienaLucent

Tellium 2000

Lucent 2001

Lucent 2001NEC ‘00NEC ‘97

Figure 1. Commercial and research grade optical systems

(Courtesy of W. Anderson et al).

1986 1990 1994 1998 2002 200610

100

1

10

100

Syst

em c

apac

ity

Single channel

Gb/

sTb

/s

(ETDM)

Multi-

chan

nel

2010

0.01

0.1

1

10

Spec

tral

eff

icie

ncy

[b/s

/Hz]

Spectral efficiency

WD

M c

hann

els

Figure 2. (Top) Historical evolution of record capacity and (bottom) spectral efficiency of “hero experiments” in fiber-1


S. J. Ben Yoo


Evolution of Optical Networking Exponential rises in the data traffic have transformed the landscape of modern networks. In the latter part of the 1990s, the exponential growths in capacity demands were successfully met by the wavelength-division-multiplexing (WDM) technology which became commercially available by then. This initial phase of optical networking, often referred to as the first generation optical networking deployed in 1997, focused on point-to-point link capacity increase. Beyond the simple capacity increase, the true benefit of optical networking may rise from the reconfigurability of such vast bandwidths directly in the optical layer without involving electronics in the data plane. The second-generation optical networking achieves reconfiguration of optical wavelength circuit paths (lightpaths) by properly configuring the optics within the optical network elements. It also supports format and protocol transparency, and simplifies the hardware requirements in the data plane. With the on-going deployment of reconfigurable optical add drop multiplexers (ROADMs) and optical crossconnects (OXCs), we anticipate the third-generation optical networking to address dynamic reconfiguration of even higher capacity traffic. The third generation optical networking is expected to be in two possible configurations. One is a packet agile optical networking technology and the other is a flow agile optical networking technology. In both configurations, extremely high data rates must be supported on each strand of fiber. In practical implementations, usable bandwidth of a single mode fiber with amplification is limited by the bandwidth of amplification technologies, which is on the order of 5~10 THz for commonly used erbium-doped fiber amplifiers (EDFAs). In order to achieve high capacity communications in the limited spectral bandwidth, the recent trend in optical communications has been to adopt advanced modulation formats that achieve high spectral efficiency together with dense WDM (DWDM).

The historical trend of this evolution appears in Figure 1 and Figure 2 [1].

Channel Bitrate

Num

ber o

f WD

M C

hann

els

128

64

32

16

8

4

2

1


'89

ONTC-1 '94

ONTC-2 '95

AT&T '85

Bellcore '88

NTT '90

AT&T '94

AT&T '94

'87 '91'86'85'83'80

Transmission

Networking

Prods.

NortelAlcatel

CienaLucent

Cambrian

HitachiMCI/Hitachi ’97

Fujitsu ’97AT&T ’97

NTT ’97

Ciena

Lucent

CienaLucentTellium ’00

Lucent ’01

Lucent ’01NEC ‘00NEC ‘97

Tyco ’03

Lucent ’02

ZTE ’11

256

512

Tyco ’10

NTT’10

ATT ’09

NTT ’09Lucent ’09

NTT’07



1986 1990 1994 1998 2002 200610

100

1

10

100

Syst

em c

apac

ity

Single channel

Gb/

sTb

/s

(ETDM)

Multi-

chan

nel

2010

0.01

0.1

1

10

Spec

tral

eff

icie

ncy

[b/s

/Hz]

Spectral efficiency

WD

M c

hann

els

Figure 2. (Top) Historical evolution of record capacity and (bottom) spectral efficiency of “hero experiments” in fiber‐

1


S. J. Ben Yoo


Evolution of Optical Networking Exponent ia l increases in data t raf f ic have transformed the landscape of modern networks. In the latter part of the 1990s, the exponential growth in capacity demands were successfully met by the wavelength-division-multiplexing (WDM) technology which had become commercially available by then.This initial phase of optical networking, often referred to as the first generation optical networking deployed in 1997, focused on point-to-point link capacity increases. Beyond the simple capacity increase, the true benefit of optical networking may arise from the reconfigurabil i ty of such vast bandwidths directly in the optical layer without involving electronics in the data plane. The second-generation optical networking achieves reconfiguration of optical wavelength circuit paths (lightpaths) by properly configuring the optics within the optical network elements. It also supports format and protocol transparency, and simplifies the hardware requirements in the data plane. With the on-going deployment of reconfigurable optical add drop multiplexers (ROADMs) and optical crossconnects (OXCs), we anticipate the third-generation optical networking to address dynamic reconfiguration of even higher capacity traffic. The third generation optical networking is expected to appear in two possible configurations. One is a packet- agile optical networking technology and the other is a flow-agile optical networking technology. In both configurations, extremely high data rates must be supported on each strand of fiber. In practical implementations, usable bandwidth of a single mode fiber with amplification is limited by the bandwidth of amplification technologies, which is on the order of 5~10 THz for commonly used erbium-doped fiber amplifiers (EDFAs). In order to achieve high capacity communications in the limited spectral bandwidth, the recent trend in optical communications has been to adopt advanced modulation formats that achieve high spectral

efficiency together with dense WDM (DWDM). The historical trend of this evolution appears in Figure 1 and Figure 2 [1].

Channel Bitrate

Num

ber o

f WD

M C

hann

els

TOTAL CAPACITY128

64

32

16

8

4

2

1


'89

ONTC-1 '94

ONTC-2 '95

AT&T '85

Bellcore '88

NTT '90

AT&T '94

AT&T '94

'87 '91'86'85'83'80

Transmission

Networking

Prods.

NortelAlcatel

CienaLucent

Cambrian

HitachiMCI/Hitachi ‘97

Fujitsu ‘97AT&T ‘97

NTT ‘97

Ciena

Lucent

CienaLucent

Tellium 2000

Lucent 2001

Lucent 2001NEC ‘00NEC ‘97



1986 1990 1994 1998 2002 200610

100

1

10

100

Syst

em c

apac

ity

Single channel

Gb/

sTb

/s

(ETDM)

Multi-

chan

nel

2010

0.01

0.1

1

10

Spec

tral

eff

icie

ncy

[b/s

/Hz]

Spectral efficiency

WD

M c

hann

els

Figure 2. (Top) Historical evolution of record capacity and (bottom) spectral efficiency of “hero experiments” in fiber-

2

optic communication systems. (Courtesy of R. J. Essiambre [1])

As Figure 2 [1] indicates, the spectral efficiency gradually increased from 0.1 b/s/Hz to 10 b/s/Hz [2] in a little over one decade. Such high spectral efficiency is achieved by adopting advanced modulation formats carried on either a single carrier or multiple carriers. Table 1 shows the summary of a) single-carrier modulation with variable bit and symbol rate; b) multicarrier modulation with variable bit and subcarrier for providing data rate B. In the orthogonal frequency division multiplexing (OFDM) [3], coherent WDM (CoWDM) [4] and optical arbitrary waveform generation (OAWG) [5-11] methods we will discuss in the next section, multiple carriers are used to substantially reduce the symbol modulation rate necessary for achieving the given data rate. In OFDM, the frequency spacing between the carriers is matched to be exactly 1/TS, where TS is the symbol duration, synchronized across the individual carriers to assure orthogonality.

ModulationLevel

QPSK 16-QAM 64-QAM

Symbol rate(Hz)

B/2 B/4 B/6

Number of carriers

1 (const.) 1 (const.) 1 (const.)

Bits per symbol 2 4 6

Data rate per polarization (b/s)

B (const.) B (const.) B (const.)

(a)

ModulationLevel

OFDM-QPSKOAWG-QPSK

OFDM-16QAMOAWG-16QAM


Symbol rate(Hz)

B/2n (const.) B/2n (const.) B/2n (const.)

Number of carriers

n n/2 n/3




(b)

Table 1. Parameters in tunable spectral width modulation format: a) single‐carrier modulation with variable bit and symbol rate; b) multicarrier modulation with variable bit and subcarrier (adapted from [12]).

While the single carrier based advanced modulation schemes are interesting on their own, (for instance, in realizing optical packet switching networks with high-spectral efficiency), the multi-carrier based methods (OFDM, Co-WDM, and OAWG) scalable to THz bandwidths and beyond can be particularly useful for realizing a new class of networks called flexible bandwidth networks [13] [12, 14-18] [19, 20], which can possibly form a new direction for the third generation optical networks.

Flexible Bandwidth Networking

Flexible Bandwidth Network

Conventional DWDM Network

f

TRX

MuxTRX

TRX

Mux

TRX

TRX

TRX

f

FlexBWTRX

WSS

WSS

FlexBWTRX

TRX = Transmitter and Receiver

Flex BW TRX = Flexible Bandwidth Transmitter and Receiver

Figure 3. Comparison between a conventional dense WDM network vs. a flexible bandwidth network.

Flexible bandwidth networking has recently been proposed as a spectrally efficient networking technology that effectively supports dynamically varying traffic demands [21]. Figure 3 illustrates a flexible bandwidth network in comparison to a conventional DWDM network. Instead of using DWDM wavelength channels on a standard wavelength grid (e.g. ITU-T) through a specific wavelength assignment process (e.g. RWA), a flexible bandwidth network can utilize flexible and elastic wavebands (or SLICEs) for a spectrum assignment process (called RSA for routing and spectrum assignment). Once such a network level RSA becomes possible, a large scale network can be deployed as Figure 4 illustrates. The advantage of flexible bandwidth networking can be summarized as (a) elimination of stranded

2



ModulationLevel

QPSK 16-QAM 64-QAM

Symbol rate(Hz)

B/2 B/4 B/6

Number of carriers





(a)

ModulationLevel

OFDM-QPSKOAWG-QPSK



Symbol rate(Hz)


Number of carriers

n n/2 n/3




(b)

Table 1. Parameters in tunable spectral width modulation format: a) single-carrier modulation with variable bit and symbol rate; b) multicarrier modulation with variable bit and subcarrier (adapted from [12]).

While the single-carrier based advanced modulation schemes are interesting on their own, (for instance, in realizing optical packet switching networks with high spectral efficiency), the multi-carrier based methods (OFDM, Co-WDM, and OAWG) scalable to THz bandwidths and beyond can be particularly useful for realizing a new class of networks called flexible bandwidth networks [13] [12, 14-18] [19, 20], which can possibly form a new direction for the third generation optical networks.



Flex BW TRx =Flexible Banswidth Transmitter and Receiver


Flexible bandwidth networking has recently been proposed as a spectrally efficient networking technology that effectively supports dynamically varying traffic demands [21]. Figure 3 illustrates a flexible bandwidth network in comparison to a conventional DWDM network. Instead of using DWDM wavelength channels on a standard wavelength grid (e.g. ITU-T) through a specific wavelength assignment process (e.g. RWA), a flexible bandwidth network can utilize flexible and elastic wavebands (or SLICEs) for a spectrum assignment process (called RSA for routing and spectrum assignment). Once such a network level RSA becomes possible, a large scale network can be deployed as Figure 4 illustrates.

6 IEEE PhotonIcs socIEty nEWsLEttER October 2011

2



ModulationLevel

QPSK 16-QAM 64-QAM

Symbol rate(Hz)

B/2 B/4 B/6

Number of carriers





(a)

ModulationLevel

OFDM-QPSKOAWG-QPSK



Symbol rate(Hz)


Number of carriers

n n/2 n/3




(b)








2



ModulationLevel

QPSK 16-QAM 64-QAM

Symbol rate(Hz)

B/2 B/4 B/6

Number of carriers





(a)

ModulationLevel

OFDM-QPSKOAWG-QPSK



Symbol rate(Hz)


Number of carriers

n n/2 n/3




(b)








2



ModulationLevel

QPSK 16-QAM 64-QAM

Symbol rate(Hz)

B/2 B/4 B/6

Number of carriers





(a)

ModulationLevel

OFDM-QPSKOAWG-QPSK



Symbol rate(Hz)


Number of carriers

n n/2 n/3




(b)








3

The advantage of flexible bandwidth networking can be summarized as (a) elimination of stranded spectrum between the wavelength grids, (b) easily accommodating subwavelength granularity traffic, (c) easily accommodating superwavelength traffic, etc. [13]. Key enabling technologies for flexible bandwidth networks are the flexible bandwidth transmitters and receivers (Flex BW Trx) that can scale up to THz bandwidth and beyond, and flexible bandwidth wavelength (spectrum) selective switches (Flex BW WSS) that can switch variable spectral bands.

Figure 4. Metro area and wide area networking based on flexible bandwidth networking technology.

Multicarrier solutions such as coherent wavelength division multiplexing (CoWDM) [22] and orthogonal frequency division multiplexing (OFDM) [23-26] have been proposed as possible implementations of flexible bandwidth networking. These solutions rely on the generation of many low speed subcarriers to form broadband data waveforms using lower speed modulators. CoWDM maintains orthogonality between closely packed subcarriers by individually modulating each tone from a set of coherent subcarrier tones, and setting the subcarrier symbol rate equal to the subcarrier spacing. OFDM systems utilize an inverse Fourier transform at the transmitter and a Fourier transform at the receiver to ensure orthogonality between subcarriers. Often a guard band is necessary to compensate for chromatic dispersion at the cost of a slight spectral efficiency penalty, but techniques such as no-guard-band OFDM can eliminate the need for guard bands [27]. Both CoWDM and OFDM systems can change the modulation format of individual subcarriers, but lack the ability for arbitrary control over subcarrier symbol rate and spacing with a single physical architecture.

A more general method for broadband waveform generation is based on dynamic OAWG [7, 8]. The generated arbitrary optical waveforms can include data waveforms in both single carrier modulation formats, and multicarrier modulation formats such as CoWDM and OFDM. Here, ‘dynamic’ refers to continuous waveform generation, as opposed to line-by-line pulse shaping [28-32], which has time duration limitations typically on the order of tens of picoseconds. In particular, spectral-slice based dynamic OAWG can create continuous, high-fidelity waveforms that overcome the limitations of rapidly updating the modulations to a line-by-line pulse shaper [33, 34]. Spectral-slice dynamic OAWG utilizes the parallel synthesis and coherent combination of many lower bandwidth spectral slices to create broadband data waveforms [35, 36]. In contrast to multi-carrier systems, the spectral slice bandwidth is not related to the subcarrier bandwidth of generated waveforms. This removes any restrictions on the subcarrier bandwidth and its modulation format, and is only limited by the total operational bandwidth of the OAWG transmitter. The parallel nature of this transmitter structure enables bandwidth scalability without increasing the bandwidth demand on the supporting electronics. The complementary receiver is optical arbitrary waveform measurement (OAWM), in which a broadband, continuous bandwidth waveform is divided into many spectral slices for parallel measurement using independent digital coherent receivers [37]. In the following section, we will discuss OAWG/ OAWM based Flex BW Trx and resulting flexible bandwidth networks. Compared to commonly used Co-WDM[22] or CO-OFDM [38] techniques which utilize subcarriers with frequency spacing ∆fS with a baud rate of exactly ∆fs , OAWG can generate truly arbitrary waveforms (including Co-WDM, CO-OFDM signals) using optical frequency comb spacings that can differ from that at the receiver. In addition, arbitrary spectral bands of arbitrary modulation formats can be generated and also detected using the OAWG technique [37, 39]. This facilitates multi-vendor interoperability and network evolutions across heterogeneous subnetwork domains.

2



ModulationLevel

QPSK 16-QAM 64-QAM

Symbol rate(Hz)

B/2 B/4 B/6

Number of carriers





(a)

ModulationLevel

OFDM-QPSKOAWG-QPSK



Symbol rate(Hz)


Number of carriers

n n/2 n/3




(b)








2



ModulationLevel

QPSK 16-QAM 64-QAM

Symbol rate(Hz)

B/2 B/4 B/6

Number of carriers





(a)

ModulationLevel

OFDM-QPSKOAWG-QPSK



Symbol rate(Hz)


Number of carriers

n n/2 n/3




(b)









3





3





3





4

OAWG Transmitter Technology

Spectral Demultiplexer

Am

Am

Am

Am

Parallel GHz Rate Intensity and Phase Modulation

Spectral Multiplexer

Bandwidth-Scalable THz Rate Optical Arbitrary Waveform

Frequency

Optical Frequency

Comb Generator

Figure 5. OAWG Transmitter principle.

A flexible bandwidth transmission system based on a dynamic OAWG transmitter and an optical arbitrary waveform measurement (OAWM) receiver can coherently generate and receive data waveforms by dividing the total waveform bandwidth into spectral slices of manageable bandwidth. This enables, using currently available technology in a bandwidth scalable manner,operation over large amounts (>1 THz) of continuous bandwidth. In this fashion, a dynamic OAWG transmitter creates large bandwidth waveforms through the parallel generation and coherent combination of many lower speed spectral slices (e.g., ~10 GHz optical bandwidth) [40]. Similarly, an OAWM receiver coherently divides the data waveform into spectral slices (e.g., ~40 GHz optical bandwidth) that are individually detected with parallel digital coherent receivers [37]. For example, a 100 GHz transmission system could be implemented using 10 × 10 GHz spectral slices at the transmitter and 4 × 25 GHz spectral slices at the receiver.

Figure 5 shows how dynamic OAWG can generate N spectral slices, each with bandwidth ∆fG, to form an aggregate output waveform with a total bandwidth of N × ∆fG. Dynamic OAWG begins with a coherent optical frequency comb (OFC), which is spectrally demultiplexed with narrow passbands placing each comb line at a separate spatial location. A set of in-phase and quadrature-phase modulators (I/Q modulators) each with a bandwidth of ∆fG apply temporal I/Q modulations to broaden the comb lines to create the spectral slices. Coherently combining the spectral slices using a gapless spectral multiplexer with broad overlapping passbands ensures a continuous bandwidth output waveform. Also, incorporating compensation for the multiplexer transmission as a pre-emphasis of the

modulation signals ensures high-fidelity waveform generation after the multiplexer [40].

Integrated Photonic 100 GHz OAWG Photonic integration is extremely important for practical and realistic implementations of flexible bandwidth networking systems. As Figure 6 shows, our group has achieved monolithic integration of InP based 100 GHz bandwidth (10 × 10 GHz) an OAWG transmitter chip including spectral demux, spectral mux, and arrays of 10 amplitude modulators and 10 phase modulators driven by 10 GHz electronic drivers shown in Figure 7.

Figure 6. InP based monolithically integrated 100 GHz bandwidth (10 x 10 GHz) OAWG transmitter chip including spectral demux, spectral mux, and arrays of 10 amplitude modulators and 10 phase modulators.

Figure 7. Ten 10 GHz amplitude modulator drivers and ten 10 GHz phase modulator drivers on a driver board coupled to the InP based integrated OAWG transmitter chip shown in Figure 6.

4



Am

Am

Am

Am




Frequency

Optical Frequency

Comb Generator








4



Am

Am

Am

Am




Frequency

Optical Frequency

Comb Generator









4



Am

Am

Am

Am




Frequency

Optical Frequency

Comb Generator








4



Am

Am

Am

Am




Frequency

Optical Frequency

Comb Generator








5

Integrated Photonic 1 THz OAWG

Figure 8. Fabricated monolithically integrated InP 100-channel × 10-GHz OAWG device [41].

The InP 10 channel × 10 GHz OAWG devices described above proved challenging not only to fabricate, but also to couple RF driver signals to the InP chips. Therefore, the second generation 100-channel devices [42, 43] incorporate optically driven phase modulators which are controlled by the intensity of remote, fiber-pigtailed lasers. This all-optical cross modulation of optical phase takes place in the waveguide, with or without multiple quantum wells, within the short (< 50 µm)

absorption length of the propagating 1310-nm control laser wavelengths. As shown in , the 100-channel, 10-GHz OAWG exploits a reflection-mode geometry to utilize a compact 50-mm diameter wafer substrate, to allow easy access from the optical control signals to the phase modulators, and to guarantee spectral alignment of the demultiplexer and multiplexer for all 100 channels. Figure 8 shows the fabricated monolithically integrated InP 100-channel × 10-GHz OAWG device with overall dimensions of 30 mm × 35 mm [41]. The device’s 1550-nm input/output on the left cleaved-facet uses a multimode-interference (MMI) 2×1 coupler to separate the input OFC from the output OAWG signals. The AWG contains 200 arrayed waveguides, each with an electro-optic Mach-Zehnder modulator (MZM) for passband shaping. Each of the 100 AWG outputs has a Michelson interferometer consisting of a 2×1 MMI splitter/combiner. The device includes 1200 independently addressable active devices, and is one of the largest-scale integrated photonic devices in the world.

OAWM Receiver Technology The working principle of optical arbitrary waveform measurement (OAWM) is quite analogous to that of OAWG except for the fact that it will do coherent detection instead of coherent generation at each spectral slice. The optical comb and arbitrary optical waveform will propagate in the opposite direction in OAWM compared to OAWG.

5




absorption length of the propagating 1310-nm control laser wavelengths. As shown in the 100-channel, 10-GHz OAWG exploits a reflection-mode geometry to utilize a compact 50-mm diameter wafer substrate, to allow easy access from the optical control signals to the phase modulators, and to guarantee spectral alignment of the demultiplexer and multiplexer for all 100 channels. Figure 8 shows the fabricated monolithically integrated InP 100-channel × 10-GHz OAWG device with overall dimensions of 30 mm × 35 mm [41]. The device’s 1550-nm input/output on the left cleaved-facet uses a multimode-interference (MMI) 2×1 coupler to separate the input OFC from the output OAWG signals. The AWG contains 200 arrayed waveguides, each with an electro-optic Mach-Zehnder modulator (MZM) for passband shaping. Each of the 100 AWG outputs has a Michelson interferometer consisting of a 2×1 MMI splitter/combiner. The device includes 1200 independently addressable active devices, and is one of the largest-scale integrated photonic devices in the world.


4



Am

Am

Am

Am




Frequency

Optical Frequency

Comb Generator








4



Am

Am

Am

Am




Frequency

Optical Frequency

Comb Generator








5






5







5






6

Figure 10(b) illustrates how an OAWM receiver characterizes waveforms through the coherent detection of M spectral slices, each with bandwidth ∆fM. For the receiver, a reference OFC with M-lines spaced at ∆fM provides a reference tone for the detection of each spectral slice [37]. The reference comb lines are isolated using a spectral demultiplexer with narrow and discrete passbands, and the signal is divided into spectral slices using a separate gapless spectral demultiplexer that has strongly overlapping passbands. Each reference comb line is then used to detect the corresponding spectral slice using a standard digital coherent receiver [44]. At this point, Digital Signal Processing (DSP) enables recombination of the spectral slices after electronic detection. In this transmission system, ∆fG can be different from ∆fM as long as the total measurement bandwidth (M × ∆fM) is greater than the generated waveform’s bandwidth (N × ∆fG). The use of spectral slices enables independent optimization at the transmitter and receiver for the exact bandwidth of available electronics, and also allows utilization of the transmitters and receivers across heterogeneous network domains.

640GHzIntegratedPhotonicOAWM

Figure 10. Silica planar lightwave integrated circuit (PLC) with 640 GHz spectral bandwidth (16 x 40 GHz) containing two spectral demultiplexers and 16 sets of optical hybrids for coherent detection. Balanced detectors are placed external to the silica PLC.

Figure 10 illustrates silica planar lightwave integrated circuit (PLC) with 640 GHz spectral bandwidth (16 x 40 GHz) containing two spectral demultiplexers and 16 sets of optical hybrids for coherent detection. Commercial balanced receivers are placed external to the silica PLC.

(a) (b) Figure 9. A schematic of the dynamic (a) OAWG and (b) OAWM techniques for bandwidth scalable generation and detection of arbitrary modulation format data transmissions. BPD: balances photodiodes, FQD: Four‐quadrature detector.

6






6

Figure (b) illustrates how an OAWM receiver characterizes waveforms through the coherent detection of M spectral slices, each with bandwidth ∆fM. For the receiver, a reference OFC with M-lines spaced at ∆fM provides a reference tone for the detection of each spectral slice [37]. The reference comb lines are isolated using a spectral demultiplexer with narrow and discrete passbands, and the signal is divided into spectral slices using a separate gapless spectral demultiplexer that has strongly overlapping passbands. Each reference comb line is then used to detect the corresponding spectral slice using a standard digital coherent receiver [44]. At this point, Digital Signal Processing (DSP) enables recombination of the spectral slices after electronic detection. In this transmission system, ∆fG can be different from ∆fM as long as the total measurement bandwidth (M × ∆fM) is greater than the generated waveform’s bandwidth (N × ∆fG). The use of spectral slices enables independent optimization at the transmitter and receiver for the exact bandwidth of available electronics, and also allows utilization of the transmitters and receivers across heterogeneous network domains.





9

6






6







6






7

Monolithically integrated OAWM chip development including the balanced receivers are currently in the plans based on recently completed designs.

Flexible Bandwidth Variable Modulation Format Signal Generation and Detection

OO

KD

PSK

Figure 11. OOK and DPSK waveforms generated and detected at 360 Gb/s using line‐by‐line amplitude and phase manipulation of optical frequency comb lines. The agreement between the targeted (dots) vs. measured (solid) data are reasonably good.

Target Measured(a)

(b) (c) Figure 12. (a) 1.2 Tb/s 16 QAM signal with 3 b/s/Hz spectral efficiency. The constellation generated using a limited number of bits (120 bit) showed a reasonably good agreement between (b) the targeted constellation and (c) the measured constellation.

Using the OAWG and OAWM technologies, we generated and detected waveforms of various modulation formats, data rates, and spectral widths. Figure 11 shows OOK and DPSK waveforms generated and detected at 360 Gb/s data rates. The agreement between the targeted (dots) vs. measured (solid) data are reasonably good. Figure 12 shows 1.2 Tb/s 16 QAM signal with 3 b/s/Hz spectral efficiency using a 120 bit sequence.

Flexible Bandwidth Networking Testbed

Node A

Node CNode B

100 GH z

300 GH z Node D

400 GHz

400 GHz

BPSK, QPSK, 8PSK, 16 QAM

(a)(b)

25 km

25 km 25 km

15 km

Network Control & Management

Figure 13. (a) Flexible bandwidth networking experimental testbed arrangement with (b) experimental system node) with variable modulation formats and bandwidths.

−30 −25 −20 −15Received Power (dBm)

log(

Sign

al B

ER)

-3

-4

-5

-7

-9

-6

-8

-10-11

BPSKQPSK

8PSK

Figure 14. BER curves for BPSK, QPSK, and 8PSK modulation formats in the Flex BW networking testbed.

We have recently established a four-node flexible bandwidth networking testbed using OAWG, OAWM, Flex BW WSS (Nistica) devices. Figure 13 (a) shows the simple topology of the testbed used for the experimental demonstration, and Figure 13 (b) shows a schematic of the network element. One of the experiments recently conducted was to demonstrate impairment-aware networking by experimenting self adaptation of modulation format while the physical layer impairment (OSNR) increases. The network testbed was able to control and to minimize the spectral bandwidth occupied by the adaptive lightpaths (flexpaths) while supporting acceptable bit error rate defined by quality of service (QoS) requirements. Each flexpaths utilized one of the variable modulation formats (BPSK, QPSK, 8PSK, 16QAM, etc) As Figure 14

7



OO

KD

PSK


Target Measured(a)




Node A

Node CNode B

100 GH z

300 GH z Node D

400 GHz

400 GHz


(a)(b)

25 km

25 km 25 km

15 km




log(

Sign

al B

ER)

-3

-4

-5

-7

-9

-6

-8

-10-11

BPSKQPSK

8PSK



8

shows the BER curves for various modulation formats at 360 Gb/s BPSK, QPSK, and 8 PSK, which exhibit higher signal-to-noise ratios for modulation schemes for higher spectral efficiency. The testbed experiment successfully showed transition of modulation formats when OSNR degradation caused the supervisory channel BER to alert the control plane to downgrade the modulation format from QPSK to BPSK while minimizing the stranded spectral bandwidth.

ConclusionDynamic OAWG and OAWM can generate and

detect arbitrary waveforms in any modulation format across their operating bandwidth. Flexible bandwidth networking can greatly benefit from this versatility of OAWG and OAWM to enable Flex BW Trx scalable to THz and beyond.

As a result, in addition to being able to operate using a single carrier modulation format such as binary phase shifted keying (BPSK) and quaternary phase shifted keying (QPSK), it is also possible to utilize multicarrier modulation formats such as CoWDM and OFDM using the same transmitter and receiver structure. Furthermore, multicarrier waveforms can be generated with each carrier having a unique bandwidth and modulation format.

Adaptive network control and management working with OAWG/OAWM based Flexible Bandwidth Network can provide efficient and robust operation under time varying physical layer impairments.

Acknowledgement The author is grateful for the work from the Next Generation Networking Systems Group at UC Davis, in particular David Geisler, Ryan Scott, Nick Fontaine, Xinran Cai, Binbin Guan, Francisco Soares, Jonghwa Baek, and collaborators at Multiplex, Royal Institute of Technology, Sweden, and Nistica. The author acknowledges support in part by DARPA and SPAWAR under OAWG contract HR0011–05–C–0155, under NSF ECCS grant 1028729, under CISCO URP Award, and under a Google Research Award.

References [1] R. J. Essiambre, et al., "Capacity Limits of

Optical Fiber Networks," Journal of

Lightwave Technology, vol. 28, pp. 662-701, Feb 2010.

[2] M. Nakazawa, "Challenges to FDM-QAM coherent transmission with ultrahigh spectral efficiency," in OpticalCommunication, 2008. ECOC 2008. 34th European Conference on, 2008, pp. 1-2.

[3] S. L. Jansen, et al., "Coherent optical 25.8-Gb/s OFDM transmission over 4160-km SSMAF," J. Lightw. Technol., vol. 26, pp. 6-15, 2008.

[4] A. D. Ellis and F. C. G. Gunning, "Spectral density enhancement using coherent WDM," Ieee Photonics Technology Letters, vol. 17, pp. 504-506, Feb 2005.

[5] B. Jalali, et al., "Photonic arbitrary waveform generator," in LEOS 2001. 14th Annual Meeting of the IEEE Lasers and Electro-Optics Society (Cat. No.01CH37242). IEEE. Part vol.1, 2001, pp.253-4 vol.1. Piscataway, NJ, USA., 2001.

[6] S. T. Cundiff and A. M. Weiner, "Optical arbitrary waveform generation," Nature Photonics, vol. 4, pp. 760-766, Nov 2010.

[7] D. J. Geisler, et al., "Bandwidth scalable, coherent transmitter based on the parallel synthesis of multiple spectral slices using optical arbitrary waveform generation," Optics Express, vol. 19, pp. 8242-8253, Apr 2011.

[8] N. K. Fontaine, et al., "Dynamic optical arbitrary waveform generation and detection in InP photonic integrated circuits for Tb/s optical communications," OpticsCommunications, vol. 284, pp. 3693-3705, Jul 2011.

[9] N. K. Fontaine, et al., "32 phase 32 amplitude optical arbitrary waveform generation," Opti. Lett., vol. 32, pp. 865-867, 1 Apr. 2007.

[10] D. J. Geisler, et al., "Modulation-Format Agile, Reconfigurable Tb/s Transmitter Based on Optical Arbitrary Waveform Generation," Opt. Express, vol. 17, pp. 15911-15925, August 2009.

[11] T. He, et al., "Modulation-Format Transparent Optical Arbitrary Waveform Generation Based Optical-Label Switching Transmitter with All-Optical Label Extraction Using FBG," presented at the LEOS, 2009.

7



OO

KD

PSK


Target Measured(a)




Node A

Node CNode B

100 GH z

300 GH z Node D

400 GHz

400 GHz


(a)(b)

25 km

25 km 25 km

15 km




log(

Sign

al B

ER)

-3

-4

-5

-7

-9

-6

-8

-10-11

BPSKQPSK

8PSK



7



OO

KD

PSK


Target Measured(a)




Node A

Node CNode B

100 GH z

300 GH z Node D

400 GHz

400 GHz


(a)(b)

25 km

25 km 25 km

15 km




log(

Sign

al B

ER)

-3

-4

-5

-7

-9

-6

-8

-10-11

BPSKQPSK

8PSK



7



OO

KD

PSK


Target Measured(a)




Node A

Node CNode B

100 GH z

300 GH z Node D

400 GHz

400 GHz


(a)(b)

25 km

25 km 25 km

15 km




log(

Sign

al B

ER)

-3

-4

-5

-7

-9

-6

-8

-10-11

BPSKQPSK

8PSK



7



OO

KD

PSK


Target Measured(a)




Node A

Node CNode B

100 GH z

300 GH z Node D

400 GHz

400 GHz


(a)(b)

25 km

25 km 25 km

15 km




log(

Sign

al B

ER)

-3

-4

-5

-7

-9

-6

-8

-10-11

BPSKQPSK

8PSK



7



OO

KD

PSK


Target Measured(a)




Node A

Node CNode B

100 GH z

300 GH z Node D

400 GHz

400 GHz


(a)(b)

25 km

25 km 25 km

15 km




log(

Sign

al B

ER)

-3

-4

-5

-7

-9

-6

-8

-10-11

BPSKQPSK

8PSK




8




















8




















8




















8




















9

[12] M. Jinno, et al., "Distance-adaptive Spectrum Resource Allocation In Spectrum-sliced Elastic Optical Path Network," IEEE Commun. Mag., pp. 138-45, 2010.

[13] M. Jinno, et al., "Spectrum-Efficient and Scalable Elastic Optical Path Network: Architecture, Benefits, and Enabling Technologies," IEEE Commun. Mag., vol. 47, pp. 66-73, 2009.

[14] M. Jinno, et al., "Virtualized Optical Network (VON) for Future Internet and Applications," Ieice Transactions on Communications, vol. E93B, pp. 470-477, Mar 2010.

[15] B. Kozicki, et al., "Optical Path Aggregation for 1-Tb/s Transmission in Spectrum-Sliced Elastic Optical Path Network," IeeePhotonics Technology Letters, vol. 22, pp. 1315-1317, Sep 2010.

[16] B. Kozicki, et al., "Experimental demonstration of spectrum-sliced elastic optical path network (SLICE)," OpticsExpress, vol. 18, pp. 22105-22118, Oct 2010.

[17] B. Kozicki, et al., "Distance-Adaptive Path Allocation in Elastic Optical Path Networks," Ieice Transactions on Communications, vol. E94B, pp. 1823-1830, Jul 2011.

[18] Y. Sone, et al., "Bandwidth Squeezed Restoration in Spectrum-Sliced Elastic Optical Path Networks (SLICE)," Journal of Optical Communications and Networking, vol. 3, pp. 223-233, Mar 2011.

[19] O. Gerstel, "Flexible use of Spectrum and Photonic Grooming," presented at the Photonics in Switching,, Monterey, CA, paper PMD3, 2010.

[20] K. Wen, et al., "Dynamic On-demand Lightpath Provisioning Using Spectral Defragmentation in Flexible Bandwidth Networks," in ECOC 2011, Geneva, Switzerland, 2011, p. Mo.2.K.4.

[21] M. Jinno, et al., "Distance-Adaptive Spectrum Resource Allocation in Spectrum-Sliced Elastic Optical Path Network," IEEE Commun. Mag., vol. 48, pp. 138-145, 2010.

[22] P. Frascella, et al., "Unrepeatered field transmission of 2 Tbit/s multi-banded coherent WDM over 124 km of installed SMF," Opt. Express, vol. 18, pp. 24745-24752, 2010.

[23] W. Shieh and I. Djordjevic, OFDM for Optical Communications. Burlington, MA: Elsevier, 2010.

[24] H. Takahashi, et al., "Highly Spectrally Efficient DWDM Transmission at 7.0 b/s/Hz Using 8 x 65.1-Gb/s Coherent PDM-OFDM," J. Lightw. Technol., vol. 28, pp. 406-414, 2010.

[25] D. Hillerkuss, et al., "Single Source Optical OFDM Transmitter and Optical FFT Receiver Demonstrated at Line Rates of 5.4 and 10.8 Tbit/s," in Proc. of OFC 2010, 2010.

[26] B. Kozicki, et al., "Experimental demonstration of spectrum-sliced elastic optical path network (SLICE)," Opt. Express, vol. 18, pp. 22105-22118, 2010.

[27] A. Sano, et al., "No-Guard-Interval Coherent Optical OFDM for 100-Gb/s Long-Haul WDM Transmission," J. Lightw. Technol., vol. 27, pp. 3705-3713, 2009.

[28] Z. Jiang, et al., "Optical arbitrary waveform processing of more than 100 spectral comb lines," Nature Photon., vol. 1, pp. 463-467, 2007.

[29] K. Takiguchi, et al., "Flexible pulse waveform generation using silica-waveguide-based spectrum synthesis circuit," Electron. Lett., vol. 40, pp. 537-538, 2004.

[30] Z. Jiang, et al., "Line-by-line pulse shaping control for optical arbitrary waveform generation," Opt. Express, vol. 13, pp. 10431-10439, 2005.

[31] P. J. Delfyett, et al., "Optical frequency combs from semiconductor lasers and applications in ultrawideband signal processing and communications," J. Lightw. Technol., vol. 24, pp. 2701-2719, 2006.

[32] N. K. Fontaine, et al., "32 phase 32 amplitude optical arbitrary waveform generation," Optics Letters, vol. 32, pp. 865-867, 2007.

[33] J. T. Willits, et al., "Theory of rapid-update line-by-line pulse shaping," Opt. Express, vol. 16, pp. 315-327, 2008.

[34] M. Akbulut, et al., "Dynamic line-by-line pulse shaping with GHz update rate," Opt. Express, vol. 18, pp. 18284-18291, 2010.

[35] N. K. Fontaine, "Optical Arbitrary Waveform Generation and Measurement," Ph.D.


9

























9

























10

dissertation, University of California, Davis, Davis, CA, 2010.

[36] R. P. Scott, et al., "Continuous, real-time, full-field waveform measurements via spectral slicing and parallel digital coherent detection," 2010 Conference on Optical Fiber Communication - OFC 2010. Collocated National Fiber Optic Engineers Conference OFC/NFOEC 2010, p. 3 pp., 2010.

[37] N. K. Fontaine, et al., "Real-time full-field arbitrary optical waveform measurement," Nature Photon., vol. 4, pp. 248-254, 2010.

[38] W. Shieh, et al., "Coherent optical OFDM: theory and design," Optics Express, vol. 16, pp. 841-859, Jan 2008.

[39] N. K. Fontaine, et al., "Demonstration of high-fidelity dynamic optical arbitrary waveform generation," Opt. Express, vol. 18, pp. 22988-22995, 2010.

[40] D. J. Geisler, et al., "Bandwidth scalable, coherent transmitter based on the parallel synthesis of multiple spectral slices using optical arbitrary waveform generation," Opt.Express, vol. 19, pp. 8242-8253, 2011.

[41] F. M. Soares, et al., "Monolithically integrated InP wafer-scale 100-channel * 10-GHz AWG and Michelson interferometers for 1-THz-bandwidth optical arbitrary waveform generation," 2010 Conference on Optical Fiber Communication - OFC 2010. Collocated National Fiber Optic Engineers Conference OFC/NFOEC 2010, p. 3 pp., 2010.

[42] F. M. Soares, et al., "Monolithically integrated InP wafer-scale 100-channel × 10-GHz AWG and Michelson interferometers for 1-THz-bandwidth optical arbitrary waveform generation," in OpticalFiber Communication Conference (OFC), 2010.

[43] F. M. Soares, et al., "Monolithic InP 100-channel × 10-GHz optical arbitrary waveform generation device," IEEEPhotonics Technology Letters.

[44] E. Ip, et al., "Coherent detection in optical fiber systems," Opt. Express, vol. 16, pp. 753-791, 2008.

9

























Documents

Flexible Bandwidth Terabit Coherent Optical Flexible ......Evolution of Optical Networking Exponential increases in data traffic have transformed the landscape of modern networks