• 沒有找到結果。

An Optical Multiplexer for Pulse Train Generation

N/A
N/A
Protected

Academic year: 2021

Share "An Optical Multiplexer for Pulse Train Generation "

Copied!
21
0
0

加載中.... (立即查看全文)

全文

(1)

Chapter 3

An Optical Multiplexer for Pulse Train Generation

As mentioned in Chapter 1, the development of integrated optical

devices and all light wave circuit applied to communication transmission

has become the future trend. We propose optical splitter based sol gel glass

waveguide for optical time division multiplexer system in this chapter. This

chapter is organized as follows: Section 3-1 introduces the technique of

optical time division multiplexing. And we describe the theory of the

generation of pulse train based on optical time division multiplexer in

Section 3-2. In Section 3-3, we design optical light wave circuit system

based on MMI sol gel glass waveguide. The characteristics of our designed

system are also presented in Section 3-3. In the final Section 3-4, we give

the summary of our designed optical all light wave circuit time division

multiplexer systems.

(2)

3-1 Introduction of the optical time division multiplexing

Development of ultra high bit-rate telecommunication is requisite for demand for expanded transmission capacity. In order to satisfy these demands, high-speed and wide-band electronics in photonic transmitters and receivers for multi gigabit-per-second pulse-code modulated (PCM) systems is indispensable. The bit rate of electrical time-division-multiplexing (ETDM) system is currently limited to 100 Gbit/s. System speed is mainly limited by the speed of the electronic circuits we used. Although individual chips have been developed at higher data rate, these are only used in laboratory presently and methods of packaging have yet to be solved. One method to overcome this electronic speed bottleneck is to extend the well-known techniques of electrical multiplexing into the optical domain.

Optical time division multiplexing (OTDM) was first demonstrated as

early as 1968 [50], primarily as means to increase the capacity of an optical

link. And Optical multiplexing technology is developed in recently years,

and it offers some advantages, such as: enabling simple multiplexing,

demultiplexing and removing the need to electronically process all traffic

(3)

entering a node. The research of optical time division multiplexing and demultiplexing for very high bit-rate PCM systems has been demonstrated for more than three decades [51-54]. Technical progress has continued, with the world’s first 40 Gbit/s experiments carried out using OTDM [55].

Several techniques were reported to produce optical pulse trains with high quality. Although most of them operated at 10 GHz to achieve the 40 GHz repetition rate, source based on soliton are particularly attractive for their simplicity and stability. Soliton pulses can arise in the anomalous dispersion regime due to nonlinear fiber propagation, and it can be described by the nonlinear Schrodinger equation (NLSE). Soliton sources at 40 GHz were already demonstrated [56]. For OTDM applications, this source would need pulse compression since 160 Gbit/s requires pulsewidth values lower than around 2 ps. Indeed, pulse compression might be adopted by another soliton-based technique.

3-2 Theory of the generation of pulse train based on optical time division multiplexer

Extremely high bit-rate signal transmission can be achieved by optical

(4)

time-division multiplexing (OTDM). Demultiplexing of such signals has been reported using the Kerr effect in the optical fibers in a nonlinear optical loop mirror [57], [58] or four-wave mixing in nonlinear mediums, such as fibers [59], or semiconductor amplifiers [60]. Lasers generating ultra-short pulses at a high repetition rate or optical multiplexing of pulses with lower repetition rate are needed for ultra-high bit-rate optical signals.

All-optical pulse multiplexing can be achieved with bulk optics [61-62] or fibers [63]. Another way is to generate a high-repetition-rate pulse train by using one laser and a planar lightwave circuit (PLC). Such an experiment has been carried out by using X16 PLC multiplexer based on integrated silica single mode waveguides [64]. We report here the generation of a 40 GHz, 60 GHz, 80 GHz and 100 GHz optical pulse train by using a light wave circuit built on sol gel glass waveguide.

3-2-1 Basic principle of pulse train generation based on lightwave circuit

The basic principle of optical time-division multiplexing and

demultiplexing is that each of the baseband data streams is

(5)

allocated a series of time slots on the multiplexed channel. A multiplexer (MUX) assembles the higher bit-rate bit stream from several baseband streams and a demultiplexer (DEMUX) reconstructs bit streams at the original lower bit rate by separating bits in the multiplexed stream. The techniques for this process are well established for electrical time-division multiplexing and demultiplexing but are only now emerging in optical systems.

Fig.3-1. shows a schematic of 1x(2N+1) channel OTDM transmission system based on splitter we designed. A train of picosecond (ps) duration optical pulses from a suitable laser source is split 2N+1 ways by our designed power splitter. Each optical pulse train is individually modulated by an electrical data signal forming in N optical RZ format data channels.

Each branch is delayed by a fraction of the clock period and synchronized

to allow passive multiplexing to sum up an individual data stream. Here the

multiplexer we used is most simply implemented by using different length

fiber with appropriate optical delays between the channels. To avoid

crosstalk between these interleaved channels, the laser source must be able

to generate optical pulses of duration < 1/N of the clock period. We design

different length of fiber in each channel to delay the optical pulse trains. To

(6)

multiplex an optical signal with period T ps to channel 2N+1, the required time delay Δ τ

i

for each path is:

50( )

i

i T i ps

τ N

Δ = ⋅ = × , i =1,2,…, N-1 (3-1)

i i

L c

τ n

Δ = Δ , i =1,2,…, N-1 (3-2)

The timing scheme is shown on Fig.3-2. We can obtain the required difference of the fiber length Δ L

i

to cause the corresponding time delay in each path. In (3-1) and (3-2), the symbol Δ L

i

is the difference length for ith fiber. Δ τ

i

is the time delay for ith path, c is the speed of light in the vacuum, n is the refractive index of fiber core. The difference of the difference of the fiber length Δ L

i

and the time delay Δ τ

i

is in terms of the first path. For example, for multiplexing the optical pulse train of 10Gbit/s to 40Gbit/s, the period T is 25 ps, so the time delay is 3.125 ps, and the difference of the fiber length is 0.2mm.

Optical amplifiers are used to maintain the correct signal power to keep

sufficient signal to noise ratio for acceptable bit-error rate. In such systems,

fiber dispersion can be managed in many ways. One method is to balance

the average group velocity dispersion in the whole system such that it is

zero referred to dispersion management and another is to use soliton

(7)

transmission techniques. Demultiplexing and clock recovery allows the input optical signal to be split into discrete channels.

3-3 Design and analysis of the generation of pulse train based on optical time division multiplexing

The demonstrated optical multiplexer generates, from a single pulse at the input, a train of 2N+1 pulses uniformly temporally spaced at the output.

The input light pulse is split into 2N+1 balanced parts by a multimode interference (MMI) splitter and each part is coupled to 2N+1 fiber of designed different lengths. The propagation time difference between two consecutive paths is Δτ ps. Then, a 2N by 1 combiner gathers these 2N+1 pulses toward a unique single mode output waveguide.

3-3-1 Generation of pulse train based on 1x3 MMI power splitter

With 18-μm-wide and 492.3-μm-length multimode waveguide of 1x3

power splitter, and connecting the designed fiber length are L, L+ΔL,

(8)

L+2ΔL, respectively. According (3-1) and (3-2), we can calculate that ΔL

for the 40 GHz pulse train is 4557μm. And in other side, we use a 3x1 coupler to connect the different length fiber. After propagation in fiber, we notice that the light is perfectly split in three balanced parts. Fig.3-3 presents a BPM simulation of the 1x3 MMI power splitter. Following we input a continuous pulse train at repetition rate of 10 GHz to the optical time division multiplexer system as shown in Fig. 3-1. The signal is booted in an erbium doped amplifier (EDFA) before being split into three channels by a 1x3 power splitter. Each of the three 10 GHz repetition rate optical pulse trains is modulated individually. Before recombining the data streams by using a 3x1 power splitter, each path passes through a variable delay line interleave a single 40 Gbit/s RZ data stream with the correct delay. The receiver is consisted of an EDFA, an oscilloscopo. Fig. 3-4 shows the OTDM measurement trace. Fig. 3-5 shows the relation of bit-error rate and received power for each output ports. The receiver sensitivity for BER of 10

-9

is -37dBm and -36.7dBm respectively for channel 1 and channel 2.

And one of three output ports is employed as a monitor.

3-3-2 Generation of pulse train based on 1x5 MMI power

(9)

splitter

With 30-μm-wide and 1203-μm-length multimode waveguide of 1x5 power splitter, and connecting the designed single mode waveguide length are L, L+ΔL, L+2ΔL, L+3ΔL, L+4ΔL respectively. According (3-1) and (4-2), we can calculate that ΔL for the 50 GHz pulse train is 3646μm. And in other side, we use a 5x1 coupler to connect the different length fiber.

After propagation in fiber, we notice that the light is perfectly split in three balanced parts. Fig. 3-6 shows a BPM simulation of the 1x5 MMI power splitter. Following we input a continuous pulse train at repetition rate of 10 GHz to the optical time division multiplexer system as shown in Fig. 3-1.

The signal is booted in an erbium doped amplifier (EDFA) before being

split into three channels by a 1x5 power splitter. Each of the five 10 GHz

repetition rate optical pulse trains is modulated individually. Before

recombining the data streams by using a 5x1 power splitter, each path

passes through a variable delay line interleave a single 50 Gbit/s RZ data

stream with the correct delay. The receiver is consisted of an EDFA, an

oscilloscopo. Fig. 3-7 shows the OTDM measurement trace. Fig. 3-8 shows

the relation of bit-error rate and received power for each output ports. The

(10)

receiver sensitivity for BER of 10

-9

is -37.1dBm, -37.5dBm, -36.8dBm, -36dBm respectively for channel 1, channel 2, channel 3, channel 4. And one of five output ports is employed as a monitor.

3-3-3 Generation of pulse train based on 1x7 MMI power splitter

With 42-μm-wide and 2763-μm-length multimode waveguide of 1x7 power splitter, and connecting the designed single mode waveguide length are L, L+ΔL, L+2ΔL, L+3ΔL, L+4ΔL, L+5ΔL, L+6ΔL respectively.

According (3-1) and (3-2), we can calculate that ΔL for the 80 GHz pulse

train is 2278μm. And we use a 7x1 coupler to connect the different length

fiber. After propagation in these single mode waveguides, we notice that

the light is perfectly split in three balanced parts. Fig. 3-9 shows a BPM

simulation of the 1x7 MMI power splitter. Following we input a continuous

pulse train at repetition rate of 10 GHz to the optical time division

multiplexer system as shown in Fig. 3-1. The signal is booted in an erbium

doped amplifier (EDFA) before being split into three channels by a 1x7

power splitter. Each of the three 10 GHz repetition rate optical pulse trains

(11)

is modulated individually. Before recombining the data streams by using a 7x1 power splitter, each path passes through a variable delay line interleave a single 80 Gbit/s RZ data stream with the correct delay. The receiver is consisted of an EDFA, an oscilloscopo. Fig. 3-10 shows the OTDM measurement trace. Fig. 3-11 shows the relation of bit-error rate and received power for each output ports. The receiver sensitivity for BER of 10

-9

is -37dBm, -38dBm, -36.8dBm, -36.7dBm, -36.7dBm, -35.7dBm respectively for channel 1, channel 2, channel 3, channel 4, channel 5, channel 6. And one of seven output ports is employed as a monitor.

3-3-4 Generation of pulse train based on 1x9 MMI power splitter

With 54-μm-wide and 6000-μm-length multimode waveguide of 1x9

power splitter, and connecting the designed single mode waveguide length

are L, L+ΔL, L+2ΔL, L+3ΔL, L+4ΔL, L+5ΔL, L+6ΔL, L+7ΔL, L+8ΔL,

respectively. According (3-1) and (3-2), we can calculate that ΔL for the

100 GHz pulse train is 1823μm. And in other side, we use a 9x1 coupler to

connect the designed single mode waveguide. After propagation in these

(12)

single mode waveguides, we notice that the light is perfectly split in three balanced parts. Fig. 3-12 presents a BPM simulation of the 1X9 MMI power splitter. Following we input a continuous pulse train at repetition rate of 10 GHz to the optical time division multiplexer system as shown in Fig. 3-1. The signal is booted in an erbium doped amplifier (EDFA) before being split into three channels by a 1x9 power splitter. Each of the three 10 GHz repetition rate optical pulse trains is modulated individually. Before recombining the data streams by using a 9x1 power splitter, each path passes through a variable delay line interleave a single 100 Gbit/s RZ data stream with the correct delay. The receiver is consisted of an EDFA, an oscilloscopo. Fig. 3-13 shows the OTDM measurement trace. Fig. 3-14 shows the relation of bit-error rate and received power for each output ports.

The receiver sensitivity for BER of 10

-9

is -37.3dBm, -37.8dBm, -37.1dBm,

-36.3dBm, -35.7dBm, -37dBm, 35.7dBm, 35.9dBm respectively for

channel 1, channel 2, channel 3, channel 4, channel 5, channel 6, channel 7,

channel 8. And one of nine output ports is employed as a monitor.

(13)

3-4 Summary

Our designed optical multiplexer for pulse train generation based on

light wave circuit with 1x(2N+1) sol gel power splitter are successfully

demonstrated. This multiplexer is based on a MMI splitter, several fiber of

different length. And we successfully generate 40Gbit/s, 50Gbit/s, 80Gbit/s,

100Gbit/s optical pulse train by using an optical multiplexer based on

1x(2N+1) sol gel MMI splitters. In table 3-1, it shows that the value ofΔτ,

ΔL, we use and the generation of pulse frequency we get.

(14)

10 G H z pul se la ser 1x (2 N+1 ) splitter (2N+1)x1 co mb in er

AM AM AM AM

EDFA ED FA EA modul a to r Clock reco ver y

Pi n- PD low-pass fi lt er BER cou n ter

10 Gbit/s RZ, electri cal sig n a l monitor (a ) T ransmitter (b) Receiver

Figure. 3-1 (a) transmitter and (b) receiver of the OTDM system

(15)

Fig. 3-3 BPM simulation of the 1x3 MMI splitter operation

Fig. 3-2 Timing scheme for multiplexing in an n-channel OTDM

(16)

0 2E-011 4E-011 6E-011 0

4E-005 8E-005 0.00012

Fig. 3-5 BER versus received power for applying 1x3 optical Fig. 3-4 Simulation of pulse train trace for using 1x3 power splitter

-39 -38.5 -38 -37.5 -37 -36.5 -36 -35.5 -35 -34.5 -34 -20

-16 -12 -8 -4

-18 -14 -10

-6 ch1

ch2

log(BER)

Received power (dBm)

(17)

0 2E-011 4E-011 6E-011 8E-011 1E-010 0

1E-005 2E-005 3E-005 4E-005 5E-005

Fig. 3-6 BPM simulation of the 1x5 MMI splitter operation

Fig. 3-7 Simulation of pulse train trace for using 1x5 power splitter

(18)

Fig. 3-8 BER versus received power for applying 1x5 optical splitter

Fig. 3-9 BPM simulation of the 1x7 MMI splitter operation

-40 -39 -38 -37 -36 -35 -34 -33 -32

-20 -16 -12 -8 -4

-18 -14 -10

-6

ch1

ch2 ch3 ch4

log(BER)

Received power (dBm)

(19)

0 0.02 0.04 0.06 0.08 0

5E-006 1E-005 1.5E-005 2E-005 2.5E-005

Fig. 3-11 BER versus received power for applying 1x7 optical splitter

Fig. 3-10 Simulation of pulse train trace for using 1x7 power splitter

-42 -41 -40 -39 -38 -37 -36 -35 -34 -33 -32

-20 -16 -12 -8 -4

-18 -14 -10 -6

ch1 ch2 ch3 ch4 ch5 ch6

log(BER)

Received power

(20)

0 2E-011 4E-011 6E-011 8E-011 0

4E-006 8E-006 1.2E-005 1.6E-005

Fig. 3-12 BPM simulation of the 1x9 MMI splitter operation

Fig. 3-13 Simulation of pulse train trace for using 1x9 power

splitter

(21)

Splitter output ports

Δτ (ps) ΔL (μm) Input pulse frequency

(Gbit/s)

Output pulse frequency

(Gbit/s)

3 25 4557 10 40

5 20 3646 10 50

7 12.5 2278 10 80

9 10 1823 10 100

Table 3-1 The relation of splitters output posts for

Δτ, ΔL, and Output pulse frequency

-40 -39 -38 -37 -36 -35 -34 -33 -32

-20 -16 -12 -8 -4

-18 -14 -10 -6

ch1 ch2 ch3 ch4 ch5 ch6 ch7 ch8

log(BER)

Received power (dBm)

Fig. 3-14 BER versus received power for applying 1x9 optical

splitter

數據

Fig. 3-3 BPM simulation of the 1x3 MMI splitter operation Fig. 3-2 Timing scheme for multiplexing in an n-channel OTDM
Fig. 3-5 BER versus received power for applying 1x3 optical  Fig. 3-4 Simulation of pulse train trace for using 1x3 power splitter
Fig. 3-7 Simulation of pulse train trace for using 1x5 power splitter
Fig. 3-8 BER versus received power for applying 1x5 optical  splitter
+4

參考文獻

相關文件

You are given the wavelength and total energy of a light pulse and asked to find the number of photons it

In this paper, we have studied a neural network approach for solving general nonlinear convex programs with second-order cone constraints.. The proposed neural network is based on

Midpoint break loops are useful for situations where the commands in the loop must be executed at least once, but where the decision to make an early termination is based on

Additional Key Words and Phrases: Topic Hierarchy Generation, Text Segment, Hierarchical Clustering, Partitioning, Search-Result Snippet, Text Data

MTL – multi-task learning for STM and LM, where they share the embedding layer PSEUDO – train an STM with labeled data, generate labels for unlabeled data, and retrain STM.

This objective of this research is to develop water based sol-gel process to apply a protective coating on both optical molds and glass performs, which can effectively prevent glass

This research proposes a Model Used for the Generation of Innovative Construction Alternatives (MUGICA) for innovation of construction technologies, which contains two models:

Research on Analog and Mixed-Signal Processing Integrated Circuit Design for ISFET-Based Linear Sensor Array