JP3019756B2 - ADD / DROP method and synchronization method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a time division multiplex communication node using an optical 2.times.2 switch for switching the route of a time division multiplex signal at a very high speed, and to a network constituted by the nodes.

[0002]

[Prior art]

<ADD / DROP Function and Loop Network> In a network composed of nodes communicating with each other, each node extracts information addressed to its own node from a transmission signal passing therethrough (DROP) or newly transmits information. A switch operation for adding (ADD) is performed. This operation is called an ADD / DROP function. A network in which a plurality of nodes having this function are connected in a loop and further hierarchized is generally used today.

<Necessity of ADD / DROP Function in Optical Signal Area> ADD / DROP is a core function of a node. In recent years, since the capacity of an optical transmission link has increased, the scale of an electric circuit that provides this function has increased dramatically, making it difficult to realize the function, and the problems of increasing costs have become apparent.

As one of the solutions, there is a method of dividing a switch operation for a signal into two layers of an optical signal level and an electric signal level. From the large-capacity line passing through the optical fiber, a signal bundle containing the signal to be accessed at that node is extracted in the optical signal area, only that is converted to an electrical signal, signal processing is performed, and again an optical signal The signal is converted and sent to the optical fiber transmission line together with the remaining optical signal. Thereby, the load of signal processing by the electric circuit can be reduced.

[0005] As a specific means for performing the switch operation in the optical signal area, a switch in a wavelength multiplexed signal space or an optical time division multiplexed signal space has been proposed and an experimental report has been made so far. This operation principle is the same as the switch operation of the frequency multiplex signal space or the time division multiplex signal space by the electric circuit which has been widely used conventionally.

<Optical 2 × 2 Switch> The present invention relates to switch operation in a time-division multiplexed signal space. Next, an optical switch which is a core of the present invention will be described.

A 2 × 2 switch has two inputs and two outputs as shown in FIGS. 9A and 9B, and has a cross and a bar according to an external control signal.
Is a switch for switching the state of. This typical configuration is a two-beam interferometer. The operation will be described with reference to FIG. A phase modulator is provided on one arm of the interferometer, and the brightness of the interference fringe is controlled by the presence or absence of a control signal. In the system using interference, two input ports and two output ports are provided in principle.

<Typical Specific Example of Optical 2 × 2 Switch> A typical example of the optical 2 × 2 switch currently generally available is a LiNbO ₃ waveguide interferometer type switch. A typical configuration is Mach-Zehnd.
er) An interferometer, which uses the Pockets effect (a phenomenon in which the refractive index changes according to an applied voltage) for phase modulation.
In this type of optical switch, the signal that controls the course of the optical signal is an electric signal, and is therefore referred to as an “electric / optical switch”.

On the other hand, "optical / optical switches" which enable switching of optical signals by optical control signals have been actively proposed and studied in recent years. A typical example is a device that uses interference like an electric / optical switch, performs phase modulation using the Kerr effect (a phenomenon in which a refractive index changes in accordance with light intensity) instead of the Pockets effect, and performs a switch operation.
Ultra-high-speed optical multiplexed signal demultiplexer (Demultiplex)
The application as xer) is being actively studied.

In order to operate a switch in an optical time division multiplex space, an optical switch that operates at high speed is required. Switches controlled by optical signals can operate faster than switches by electrical signals for two main reasons. 1
One reason is that some optical nonlinear phenomena such as the Kerr effect have a very high response time constant compared to electronic circuit elements, and the other is that the optical fiber signal line has a wide band. The ability to generate and use ultrashort pulses due to the nature.

<ADD / DRO of Optical Time Division Multiplexing / Demultiplexing System>
Application to P Function> There have been several experimental reports of the ADD / DROP function in which the optical switch is used to switch the route of a time-division multiplexed optical signal as it is.

For example, the first one held in September 1993
9th European Conference on Optical Communications (19th European Co
nference on Optical Communication), Proceedings, Vol. 282-184.
Page (Vol. 2, pp 281-284). M. Patrick et al., "Variable Bit Rate All-Optical Demultiplexing (Bi
t-Rate Flexible All-Optical De-Multiplexing using
a Nonlinear Optical Loop Mirror) "
A. D. Ellis et al., “Three-node 40Gb / s OTDM network experiment using an electric / optical switch.
electro-optic switches) "(electronics letters, vol.30, No.1
6, 1994, pp1333-1334).

FIG. 11A and FIG. 11B show these configuration diagrams. The configuration difference between these two reports is that, in FIG. 11A, the received signal is dropped and the transmitted signal is
11B, while the received signal light is 2
DROP and ADD are performed independently for each signal, and an additional erasing function is required for ADD. For the optical switch, FIG.
A uses an optical / optical switch, and FIG. 11B uses an electrical / optical switch. Incidentally, neither uses a 2 × 2 switch for light.

<Network Synchronization> In order to perform the ADD / DROP processing by the time division multiplex switch operation at the node, it is necessary that the timings of a plurality of signals to be switched are aligned. That is, synchronization between nodes communicating with each other is required. This is called network synchronization. In general, each node has an accurate clock, and its frequency does not greatly shift, but fluctuations in the phase level always occur due to the influence of vibration, temperature change, etc. while the signal is transmitted on the transmission line. . Therefore, phase adjustment is indispensable.

At present, a buffer having a semiconductor memory as a core is used as the phase adjusting means. This buffer memory is an Elastic Store.
cStore), which allows a write operation and a read operation to be operated by different clocks. As shown in FIG. 12, by writing in synchronization with the clock of the transmission line and reading out in synchronization with the clock of the node, the phase fluctuation can be buffered. Conversely, a signal synchronized with the node clock can be replaced with the transmission line clock.

[0016]

[Problems to be solved by the invention]

<Inventions of Claims 1, 2 and 3: Problems of Conventional ADD / DROP Function> The conventional method has the following disadvantages.

First, there are the following difficulties in timing adjustment.

In order to perform ADD / DROP, it is necessary to adjust the timing of dropping a pulse in a desired portion and the timing of adding a pulse to a desired portion. A. D. In the report of Ellis et al., It is necessary to further adjust the timing for erasing. These timing adjustment precisions correspond to a high bit rate (bit rate).
Obviously, it becomes more severe in e). In the experimental reports given in the prior art, no measures are taken against this, and this is a problem for practical application.

Second, the waveform of the pulse to be added is AD
In some cases, the pulse waveform of the data signal at the destination D must be the same. In the time division multiplexing method, it is necessary to narrow the pulse width as the bit rate becomes higher. But usually A
The bit rate of the DD signal is lower than the bit rate of the ADD destination signal. As a result, the pulse width of the signal to be added must be smaller than the pulse width corresponding to the bit rate.

For example, in the above-mentioned report, the bit rate is 40 G [b / s] for the ADD destination signal and 10 G [b / s] for the ADD signal, and each pulse width is about 10 [p / p].
s]. Therefore, the ADD signal includes a general 1
0G [b / s] Minimum pulse width of about 1 transmitted by the optical transmitter
00 [ps] NRZ code (NRZ: Non Retu)
Since an optical signal of rn to Zero cannot be used, an RZ code (RZ: Ret) having a pulse width of about 10 [ps] is used.
A special 10G that sends out a "urn to Zero"
[B / s] An optical transmitter is required. As described above, this method has a disadvantage that a special optical transmitter is required.

Third, since the number of parts is large, the cost is high, and since the configuration is complicated and the number of adjustment points is large, it is difficult to maintain stable operation.

An object of the present invention is to provide an ADD / DROP method which overcomes the above-mentioned disadvantages of the prior art.

<Invention of Claim 4: Problem of Conventional Network Synchronization>
Although the speed of optical communication has been remarkably improved in recent years, the operating speed of semiconductor memories has hardly improved. Therefore, the phase synchronization operation of the ultra-high-speed signal to be optically transmitted is divided into a number of low-speed signals and then processed in parallel.

To give a concrete example, 10G is currently commercially available.
Although a trunk line with a speed of [b / s] is about to be introduced, the operating speed of the elastic store buffer is about 50 M [b / s] at maximum, so it is separated into at least about 200 low-speed signals. Processing in parallel. In order to withstand large phase fluctuations, a large-capacity memory is required, and the cost is extremely high. ("SDH transmission method", page 128, supervised by Shimada Yoshijin, Ohmsha, 199
3).

By the way, even if a high-speed transmission signal is separated into a large number of low-speed signals, the target of the ADD / DROP processing is generally only a small part of the whole signal, and most of the signal is not processed. Is also multiplexed with the high-speed signal again without being processed and sent to the next node. That is, a large cost is required only for the phase adjustment.

[0026]

SUMMARY OF THE INVENTION ADD / DRO of the present invention
The P method includes a two-input two-output switch (hereinafter, an optical 2 × 2 switch) for an optical signal in a network node transmitting and receiving an optical signal by a time division multiplexing method, and two input ports of the optical 2 × 2 switch. Includes a transmitted trunk optical signal and a transmitted optical signal from its own node, ie, AD
D optical signal is input, and two output ports of the optical 2 × 2 switch are respectively connected to a trunk optical signal to be transmitted to the next node and a received optical signal addressed to the own node, that is, a DROP optical signal. Are output, and the timing between the trunk optical signal and the ADD optical signal input to the optical switch and the timing of the switching operation of the optical switch are set to a desired portion of the trunk optical signal. The ADD of
The optical signal is adjusted so as to be ADDed, and the timing of the switching operation of the optical switch is adjusted so that the DROP optical signal is DROPed from a desired portion of the trunk optical signal, and the optical signal is switched as necessary. The ADD process and the DROP process are performed simultaneously.

In the ADD / DROP method, the optical 2 × 2 switch is controlled by an optical signal.

In the ADD / DROP method, the optical 2 × 2 switch is controlled by an electric signal.

Alternatively, according to the present invention, in an optical network including a plurality of nodes and at least one optical fiber transmission line interconnecting the nodes, the node includes one parent for supplying a clock of the transmission line signal. A node and a plurality of other child nodes, and the child node is ADD /
A DROP function and an elastic store buffer having a capacity corresponding to the capacity of each of the ADD signal and the DROP signal are provided. The elastic store buffer performs write and read operations based on different clocks. The phase fluctuation of the ADD signal with respect to the transmission line clock can be performed by writing the ADD signal into the ADD signal buffer in synchronization with the clock of the child node and reading out the ADD signal in synchronization with the clock of the transmission line. And the phase fluctuation of the DROP signal with respect to the child node clock is
A synchronization method characterized in that a signal is written into the DROP signal buffer in synchronization with a clock of a transmission line and read out in synchronization with a clock of the child node, thereby being absorbed.

(Operation) The above-mentioned problem of the prior art is solved by the AD according to the first or second or third aspect of the present invention.
This is overcome by the D / DROP method and the synchronization method of claim 4 for effectively utilizing the method.

<Inventions of Claims 1, 2 and 3> An optical ADD / DROP method using the present invention will be described.

FIG. 1 shows a basic configuration. Here, the configuration of one node connected to the loop transmission path is shown. The transmitted main optical signal and the transmitted optical signal from the own node, that is, the ADD optical signal, are input to the two input ports of the optical 2 × 2 switch, respectively. A trunk optical signal transmitted to the node and a received optical signal addressed to the own node, that is, a DROP optical signal are output. The operation of the main optical signal, the ADD optical signal, and the operation of the optical switch are synchronized. As shown in Table 1, the signal going to the trunk line and the DRO
An operation of switching the signal to be performed is performed. The logic of the switches in Table 1 may be reversed. Thus, according to this method, ADD
/ DROP can be realized with a very simple configuration.

[0033]

[Table 1]

<Invention of Claim 4> When a network is formed by using the optical ADD / DROP method of the invention of claim 1, 2, or 3, the method of achieving phase synchronization becomes a problem. As described in the related art, at present, the phase difference is absorbed by using a variable-length buffer using an electric circuit. Unfortunately, constructing a similar variable-length buffer for optical signals is currently unrealistic.

Therefore, if the following synchronization method of the invention of claim 4 is adopted, a network that can effectively utilize the invention of claim 1 can be constructed.

FIG. 7 shows the configuration. One parent node is placed in the loop, which has a master clock that generates the clock for the loop transmission path. The other plurality of child nodes perform ADD / DROP switching operation in synchronization with the clock of the transmission path.

FIG. 8 shows an example of the child node signal processing. An optical signal of B _H [b / s] passes through the optical fiber transmission line. This optical signal is a signal obtained by time-division multiplexing a signal having a capacity of B _L [b / s] in an optical signal area. In the figure, four B _L
The signals are bundled to form a _BH signal, and one _BL signal is ADD / DROP-processed.

One DR-dropped B _L signal is generally composed of signals of a plurality of capacitances B _E [b / s], and is cross-connected to the _BE signal of the child node in units of the B _L signal. After that, it is multiplexed again with the _BL signal and ADDed.

All the processing so far is synchronized with the clock of the transmission line. Therefore, the phase difference with the clock of the child node need only be absorbed for the low-speed signal before ADD and after the DROP, and the scale can be greatly reduced as compared with the conventional method. Will be fully feasible.

[0040]

BEST MODE FOR CARRYING OUT THE INVENTION

Embodiment 1 <Inventions of Claims 1, 2 and 3> A representative Sagnac (Sa)
gnac) an interferometer type optical / optical switch, a non-linear optical loop mirror using an optical fiber as a non-linear medium (NO
LM: Nonlinear Optical Loop
Mirror) will be described with reference to an embodiment incorporating the present invention.

(Structure) The structure is shown in FIG. First, light 2 ×
The description will be started from two switches.

The optical fiber 50 and the 3 dB coupler 10 are optically connected to each other and form the basis of a Sagnac interferometer. This optical fiber 50 is a silica-based single mode optical fiber (length: 10 [km], zero dispersion wavelength: 1550 [n]
m], the total loss: about 3 dB), and serves as an optical waveguide constituting the Sagnac interferometer and also serves as an optical nonlinear medium.

The control light introducing 3 dB coupler 20 has a configuration in which the operation of the Sagnac interferometer is not substantially affected.
Is inserted. The signal light has excessive loss (3
dB), but only as much as the losses due to other factors.

The ports 10a and 10b of the coupler 10
3 dB coupler 3 for separating lights having different traveling directions
1, 41 are connected, and further, optical isolators 311 and 312 for blocking light going back to the input port are connected.

Next, the wavelength setting will be described.

In this embodiment, in order to prevent interference between two input signal lights, the wavelengths thereof are made different. As a result, light of three wavelengths is used together with the control light.

The input trunk optical signal 101 has a wavelength λ ₁ = 15.
This is an optical data signal of an RZ code having 53.5 [nm], a pulse width of 40 [ps], and a bit rate of 10 G [b / s]. A
The DD optical signal 102 has a wavelength λ ₂ = 1552.5 [nm],
This is an optical data signal of an NRZ code having a bit rate of 622 M [b / s]. The control light 120 has a wavelength λ _C = 1547.
[Nm], pulse width 40 [ps], repetition frequency 622
This is an optical clock signal of M [Hz]. ADD optical signal 10
2 and the clock phase of the control light 120 were matched with the input trunk optical signal 101 using a phase shifter. This is not synchronous, but there is no difference from the synchronous state as long as the clock phase does not fluctuate.

FIG. 4 shows the wavelength settings of these three light waves.

The parabolic curve represents the relative delay time at each wavelength with reference to the zero dispersion wavelength 1550 [nm] of the optical fiber 50. Since the control light power required for switching is smaller when the relative delay is smaller, these three light waves are set so that the relative delay times are minimized with each other, and are separated from each other to such an extent that they can be separated by an optical filter.

The optical filters 301 and 302 have band pass characteristics, and have a bandwidth of about 1 [nm]. As shown in FIG. 4, the optical filter 301 removes λ _C and passes λ ₁ and λ ₂ , and the optical filter 302 removes λ _C and λ ₂ and passes λ _1. I have.

The optical receiver 320 has a bit rate of 622M.
[B / s] optical receiver, which is connected to receive the DROP optical signal 112 obtained by filtering the light coming out of the coupler 41 with the optical filter 302. The optical receiver 320 allows the DROP optical signal 112
Is converted to a DROP electrical signal 340.

The optical transmitter 321 outputs the ADD electric signal 341
Is converted to an ADD optical signal 102 and transmitted at a bit rate of 622 M [b / s].

Each of the polarization controllers 331, 332, and 333 is a polarization converter capable of converting an arbitrary polarization state.

(Operation) Next, the operation will be described with reference to the block diagram of FIG.

10 G [b / s] input trunk optical signal 101
Passes through the isolator 311 and the coupler 31 and enters the Sagnac interferometer. On the other hand, the ADD optical signal 102 also enters the Sagnac interferometer through the isolator 312 and the coupler 41. At this time, port 3 of couplers 31 and 41
Optical signals are also emitted from 1d and 41d, but are discarded.
Termination to prevent adverse effects such as reflection.

The light incident on the Sagnac interferometer is split into two equal parts by the coupler 10, goes around the loop, returns to the coupler 10 again, and interferes. When there is no control light 120, each interference output light returns to the port where it was incident. When the control light is present in an appropriate amount to cause the phase difference π,
Each interference output light is output from a port opposite to the input port.

The interference output light always includes λ _C , and also includes λ ₁ or λ ₂ depending on the presence or absence of the control light. I will explain each of them.

First, the light output from the port 31a of the coupler 31 is controlled by the optical filter 301 into control light (λ _C ).
Is removed. As a result, the output trunk optical signal 111 becomes the input trunk optical signal 101 (λ ₁ ) when there is no control light,
When there is control light, it becomes an ADD optical signal 102 (λ ₂ ).

Next, the light output from the port 41a of the coupler 41 is controlled by the optical filter 301 into control light (λ _C ).
And the ADD optical signal 102 (λ ₂ ) is removed. As a result, the DROP optical signal 112 is not output when there is no control light, and is not output when there is control light.
(Λ ₁ ).

When the above operation was performed, a good basic operation was confirmed.

FIG. 5 shows the eye pattern. In order from the top, the input trunk optical signal 101, the ADD optical signal 102, and the control light 12
0, output trunk optical signal 111 and DROP optical signal 112. In the input trunk optical signal 101 of B _H = 10 G [b / s], one B _L = 622 M [b / s] signal channel is replaced with the ADD optical signal 102 transmitted from the own node. In this embodiment, since B _H / B _L = 16, replacement is performed every 16 bits. here,
ADD / DROP is performed on the eighth bit slot.

In this embodiment, the optical 2 × 2 switch is
Although the NOLM which is an optical / optical switch is used, the present invention is not limited to this, and it is apparent that the same operation can be realized by using another optical 2 × 2 switch.

In this embodiment, since the Sagnac interferometer type switch is used, it is necessary to separate the lights traveling in different directions at the ports 10a and 10b, and the 3 dB couplers 31 and 41 are connected. Instead, an optical circulator is used. For example, a loss of 3 dB in the 3 dB coupler can be reduced.

In the present embodiment, the ADD signal is an optical signal of the NRZ code, but this may be an optical signal of the RZ code.

There are the following methods for avoiding the influence of interference between two signal lights at the coupler of the switch input section.

One method is to make the wavelength of the signal light slightly different. However, if the frequency is far apart, Ke
Since there is a fear that the performance may be degraded due to the wavelength dispersion of the rr medium, an appropriate frequency difference is selected so that the difference frequency does not enter the band of the photodetector. This embodiment also uses this method.

Another method is to input two signal lights orthogonally to each other, make the control light circularly polarized, and make the switching efficiency constant for all polarized signal lights ( See JP-A-4-19717).

In the case where the wavelengths are made different, the wavelengths of the light entering the node and the light exiting the node are partially different. In this case, when there are a plurality of child nodes, there is a possibility that wavelength setting may be difficult. In order to avoid this, as shown in FIG. 6, the output light of the 2 × 2 switch may be wavelength-converted to completely return to the original wavelength. For example, M. Schilling et al., “Wavelength converter based on integrated using a photoactive integrated 3-port Mach-Zehnder interferometer.
all-active three-port Mach-Zehnder interferomete
r) "(Electronics Letters, electronics lette
rs, vol. 30, No. 25, 1994, pp. 2128-2130) is suitable for such an application.

In this embodiment, the simplest bit multiplexing is used as the frame configuration of the trunk optical signal. However, the present invention is not limited to this, and other frame configurations, for example, in units of bytes or several bytes are used. Cell unit.

In this embodiment, a loop is used in the shape of the trunk transmission line network. However, the present invention is not limited to this.
For example, the present invention can be applied to a bus-type network.

[0071]

Embodiment 2 <Invention of Claim 4> One parent node and a plurality of claims 1
ADD / DROP nodes (child nodes) are connected by optical fiber transmission lines to form a loop network as shown in FIG. Bit rate of main optical signal B _H = 10G
[B / s], ADD / DROP bit rate B _L =
622 M [b / s]. In the sending child node, 6
A 22M [b / s] signal is transmitted every 16 bits by the optical ADD / DROP function, and the child node on the receiving side converts the signal to 16b by the optical ADD / DROP function.
It is received every other it.

FIG. 2 shows a detailed configuration of the child node. Light 2
As the × 2 switch and the optical transceiver, those described in the embodiment of the first invention were used.

Synchronization is realized as follows. At each child node, a part of the trunk optical signal is branched and
The intensity modulated signal component in the [Hz] band is received by the photodetector, and Q
10G using an electric tank filter with a value of about 800
[Hz] clock recovery is performed, and the
z] in the electric circuit. This clock is called a transmission line clock.

The child node on the transmitting side initially replaces the ADD signal synchronized with the clock of its own node with the transmission line clock using an elastic store. Further, in order to perform ADD to an appropriate bit slot, an appropriate delay is given to the transmission line clock using a phase shifter, and light 2 ×
Driving two switches.

In the child node on the receiving side, in order to DROP an appropriate bit slot, the optical 2 × 2 switch is driven by giving an appropriate delay to the transmission line clock using a phase shifter. By this phase adjustment, the bit slot to be accessed is specified. Further, at first, the DROP signal synchronized with the transmission line clock is replaced with the clock of the own node by using an elastic store.

When the above configuration was operated, good basic operation was confirmed. An operation waveform similar to that of FIG. 5 was obtained, and the operation continued stably even if the length of the optical fiber transmission line was slightly varied using an optical variable delay device. The capacity of the elastic store required for synchronization is greatly reduced to 1/16 compared to the case where all trunk signals are synchronously processed.

In order to secure the operation stability, at least one of the following functions is required in the loop. This is not a limitation that applies only to the present invention, but is a common technical requirement for loop networks.

First, a signal regenerator is required. This prevents a failure due to signal degradation even when the signal goes around the loop many times.

Further, it is necessary to provide a buffer for absorbing fluctuations in the time required for the signal to make one round of the loop.

In this embodiment, these functions are provided to the parent node, and a buffer is realized by an electric circuit as in the prior art.

The present invention does not particularly mention the medium access method, because the switch operation and the network synchronization method of the present invention can be combined with various medium access methods. The medium access method is closely related to the technology provided by the present invention, but has a different hierarchy.

In this embodiment, a high-speed photodetector and an electronic circuit are used to reproduce the transmission line clock, and the operation speed may limit the speed of the entire system.
As a method for avoiding this, for example, JP-A-6-30
A clock recovery system using optical signal processing as described in Japanese Patent No. 3216 may be used.

In this embodiment, the optical / optical switch described in the first embodiment of the present invention is used as the optical 2 × 2 switch. However, the present invention is not limited to this. Good.

In the above description, the parent node and the child node are clearly separated for the sake of explanation. However, the only restriction is that only one master clock must exist in the loop. Yes, as long as this is maintained, the configuration of the parent node and the child node may be the same. In this way, if a failure occurs in the parent node, another child node can become the parent node, and there is also an advantage that the restoration of the network becomes easy.

[0085]

The optical ADD / D according to the first, second and third aspects of the present invention.
According to the ROP function, strict timing conditions of ADD and DROP are automatically satisfied, so that the realization is easy.

The pulse width of the ADD signal may correspond to the bit rate, and may be an NRZ code signal, so that a general optical transmitter can be used.

Further, since the configuration is simplified, it is possible to expect an improvement in operation stability and reliability and a reduction in cost.

Further, by using the optical / optical switch, the speed limitation of the electric signal is removed, and the capacity can be further increased.

According to the network synchronization method of the fourth aspect of the present invention, it is only necessary to partially perform the phase synchronization of an ultra-high-speed and large-capacity optical signal, so that simple and low-cost network synchronization can be realized and a simple configuration. In addition, a large-capacity optical ADD / DROP type network can be realized.

[Brief description of the drawings]

FIG. 1 is an explanatory view of a first invention.

FIG. 2 is a configuration diagram of a fourth invention.

FIG. 3 is a configuration diagram of an embodiment of the first invention.

FIG. 4 is a wavelength setting diagram according to the embodiment of the first invention.

FIG. 5 is an explanatory diagram of operation waveforms according to the embodiment of the first invention.

FIG. 6 is a typical configuration example of an embodiment of the first invention.

FIG. 7 is an explanatory view of an embodiment of the fourth invention.

FIG. 8 is an explanatory view of an embodiment of the fourth invention.

FIG. 9 is an explanatory diagram of a 2 × 2 switch.

FIG. 10 is an explanatory diagram of an interferometer type optical switch.

11A is an explanatory diagram of a first conventional method of the ADD / DROP function, and FIG. 11B is an explanatory diagram of a second conventional method of the ADD / DROP function.

FIG. 12 is an explanatory diagram of an operation of an elastic store.

[Explanation of symbols]

 10, 20, 31, 41 Coupler 50 Optical fiber 101 Input trunk optical signal 102 ADD optical signal 111 Output trunk optical signal 112 DROP optical signal 120 Control light 301, 302 Optical filter 311, 312 Isolator 320 Optical receiver 321 Optical transmitter 331 , 332, 333 Polarization controller 340 DROP electric signal 341 ADD electric signal

Continuation of the front page (51) Int.Cl. ⁷ identification symbol FI H04Q 3/52 101 H04L 11/00 330 (58) Investigated field (Int.Cl. ⁷ , DB name) H04Q 3/52 H04Q 11/00- 11/08 H04B 10/02 H04J 3/00 H04L 7/00 H04L 12/42

Claims (6)

(57) [Claims] An optical network, comprising a plurality of nodes and at least one optical fiber transmission line interconnecting the nodes, for transmitting an optical signal by a time division multiplexing method, wherein the node comprises a transmission line. A parent node for supplying a signal clock; and a plurality of other child nodes. The child node has an ADD / DROP function and an irradiator having a capacity corresponding to the capacity of each of the ADD signal and the DROP signal. A storage buffer and a two-input two-output switch for optical signals (hereinafter referred to as an optical 2 × 2 switch). The elastic store buffer performs write and read operations based on different clocks. The phase fluctuation of the ADD signal with respect to the transmission line clock can be performed by converting the ADD signal into a buffer for the ADD signal. It is absorbed by writing in synchronization with the clock of the child node and reading out in synchronization with the clock of the transmission line. The phase fluctuation of the DROP signal with respect to the child node clock is generated by transmitting the DROP signal to the buffer for the DROP signal. It is absorbed by writing in synchronization with the clock and reading out in synchronization with the clock of the child node. The two input ports of the optical 2 × 2 switch are respectively connected to the transmitted trunk optical signal and from the own node. The optical 2 × 2 switch has two output ports respectively connected to a trunk optical signal to be transmitted to the next node and a receiving optical signal to the own node. Signal, that is, a DROP optical signal, and the switching operation timing of the optical 2 × 2 switch is synchronized with the clock of the transmission line. Cage, the timing of the switching operation timing and the optical switch between the stem line optical signal and the ADD optical signal input to the light switch, desired of the A to the desired portion of the stem line optical signal
The DD optical signal is adjusted so as to be ADDed, and the switching operation timing of the optical switch is adjusted so that the DROP optical signal is DROPed from a desired portion of the trunk optical signal. An ADD / DROP method, wherein the ADD process and the DROP process are performed simultaneously. 2. The ADD according to claim 1, wherein the switching of the optical 2 × 2 switch is performed using a control signal.
/ DROP method. 3. The apparatus according to claim 2, wherein when there is no control signal, the transmitted trunk optical signal is output as a signal to be transmitted to a next node, and the DROP optical signal is not output. ADD / DROP method. 4. When the control signal is present, the transmitted trunk optical signal is output as a DROP optical signal, and the ADD optical signal is output as a trunk optical signal to be transmitted to a next node. A according to claim 2 or 3,
DD / DROP method. 5. The A according to claim 1, wherein said 2 × 2 switch is controlled by an optical signal.
DD / DROP method. 6. The ADD / DROP method according to claim 1, wherein said 2 × 2 switch is controlled by an electric signal.