Optical signal processing device

Abstract

To provide an optical signal processing device capable of providing a transponder aggregation function which has a small signal loss in a path from an input to an output, and is also provided in a low cost.SOLUTION: A TPA 4 having: a port P of N transponder 41; and M cross-connect ports 48, comprises: K optical switches 42 connected to K ports P of N ports; L optical switches 52 connected to (N-K) optical switches of the N optical switches; M optical confluence directional couplers 43 connected to each optical switch 42 in a mesh like; L optical confluence directional couplers 44 connected with the optical switches 52 in the mesh like; a matrix switch 46 connected to each optical confluence directional coupler 44; and M optical confluence directional couplers 47 connected to each optical confluence directional coupler 43, the matrix switch 46, and each cross-connect port 48.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] The present invention relates to an optical signal processing device. [Background technology]

[0002] With the expansion of data communication networks such as the Internet, optical communication networks are expected to have ever-increasing capacity. To meet this growing demand for networks, wavelength division multiplexing (WDM) communication has been put into practical use. Recently, there has also been an increasing demand for wavelength selective switches (WSS), which switch optical signal nodes by wavelength without converting optical signals to electrical signals. A node configuration using wavelength selective switches is called a reconfigurable optical add / drop multiplexing (ROADM) system. Non-Patent Document 1 discloses a node with a configuration called a multi-degree ROADM, which processes optical signals from multiple nodes. Non-Patent Document 2 discloses a configuration in such a system that uses a multicast switch (MCS), which functions as a transponder aggregator (TPA), connecting optical signals from any node to any transponder.

[0003] FIG. 1 is a schematic functional block diagram for explaining the MCS disclosed in Non-Patent Document 2, showing the configuration of a multi-degree optical node. As shown in FIG. 1, the multi-degree optical node includes a cross-connect unit (WXC: Wavelength Cross Connect) 101 composed of a multicast switch 102 combining an SPL (splitter) and a WSS 103, and a WSS 103. WDM signals are input to the optical node from M input fibers 111 to 11M. Furthermore, among the input signals, signals to be added or dropped at the optical node are sent to a drop TPA 104a, and transmitted from the TPA 104a to one of N transponders 131 to 13N. Similarly, optical signals generated in transponders 131 to 13N are output to one of output fibers 121 to 12M via an add TPA 104b and the WSS 103 of the cross-connect unit 101. In this specification, a node where an optical signal is input (received) or output (transmitted) with respect to the element of interest as the reference will hereinafter also be referred to as a "path".

[0004] 2 is a functional block diagram illustrating TPA 104a, which is a drop TPA among the TPAs ​​shown in FIG. 1. However, since TPA 104b, which is an add TPA, has a configuration similar to that of TPA 104a, the description of TPA 104a will be substituted for that of TPA 104b in this specification. As shown in FIG. 2, TPA 104a includes M 1-input, N-output (hereinafter also referred to as 1×N) splitters 201 to 20M, each having an input corresponding to input fibers 111 to 11M, an M×1 switch 211 that receives and outputs an optical signal from each of the 1×N splitters 201 to 20M, and N optical filters 221 to 22N that receive and filter the output of M×1 switch 211. The optical filters 221 to 22N output the filtered signals to transponders 131 to 131N shown in FIG. 1.

[0005] However, MCS involves the splitting of input optical signals, which in principle results in optical signal loss. That is, in MCS, an optical signal is split into N parts by 20M from the 1×N splitter 201, so the optical signal strength is reduced to 1 / N. For example, when eight transponders are connected, a theoretical loss of 9 dB occurs, and when 16 transponders are connected, a theoretical loss of 12 dB occurs. Generally, a transponder requires an input of optical power equal to or greater than the minimum receiving sensitivity (the minimum optical power at which a signal can be received), so there is an upper limit to the number M that can be split. To alleviate the restriction imposed by this upper limit, an optical amplifier such as an erbium-doped fiber amplifier (EDFA) may be inserted between the multicast switch 102 and the TPA shown in FIG. 1.

[0006] However, while EDFAs increase the intensity of optical signals, they also generate amplified spontaneous emission (ASE) noise, which degrades the optical signal-to-noise ratio (OSNR). To solve this problem, it is possible to remove ASE by inserting optical filters 221 to 22N between the transponder and MCS, as shown in Figure 2. However, inserting optical filters increases the number of components and also increases the signal loss that was supposed to be compensated for by the EDFA.

[0007] Patent Document 1 describes a method for realizing a loss-free TPA function (M×N WSS) by combining multiple wavelength selective switches (WSS) and non-wavelength selective switches to eliminate optical signal loss. Figure 3 is a diagram for explaining the configuration described in Patent Document 1, showing an M×N WSS using an optical system called a Spatial and Planar Optical Circuit (SPOC), which combines optical waveguides and free-space optics. The configuration shown in Figure 3 can be said to replace the M 1×N splitters in the MCS shown in Figure 2 with M 1×N WSSs for selecting transponders connected via the free-space optics.

[0008] In FIG. 3, a WDM signal is input from In (WDM side) (three inputs a, b, and c in the figure) connected to the cross-connect unit 101 shown in FIG. 1. The WDM signal is converted into collimated light via an SBT (Space Beam Transformer) circuit having a lens function integrated in an optical waveguide, and is output to the free-space optical system. At this time, the input waveguides a, b, and c have different emission directions of the light waves. In the free-space optical system, the WDM signal is wavelength-separated by a diffraction grating Gr in the direction perpendicular to the paper surface, and then the WDM signal is passed through a plurality of lenses L SP , L DP The light is incident on the liquid crystal element LC, which is a switching element, at different positions in the y-axis direction for each wavelength via the optical waveguides a, b, and c. At this time, the liquid crystal element LC is incident on different positions A, B, and C in the x-axis direction depending on the input waveguides a, b, and c. The liquid crystal element LC is composed of multiple pixel elements that can individually modulate the optical phase two-dimensionally. The liquid crystal element LC reflects and deflects each wavelength in the vertical direction (x-axis direction) of the paper. In other words, the optical signals demultiplexed in the direction perpendicular to the paper are deflected and reflected in different directions, independently for each input waveguide a, b, and c. The reflected optical signals are coupled to one of SBT11_1 to 11_4 and input to the optical waveguide optical system. An optical signal input to the optical waveguide optical system propagates through one of the connecting waveguides 12_1-a to 12_1-c, 12_2-a to 12_3-c, ..., 12_4-a to 12_4-c depending on the input waveguide a, b, or c, passes through N M×1 switches 13_1 to 13_3, is output from one of the output waveguides 1 to 4, and is input to the transponder.

[0009] The M×N WSS described above does not generate losses due to branching. In addition, the WSS function, which is configured with a spatial optical system, provides a wavelength filter function, eliminating noise components of wavelengths other than the signal wavelength. In this way, the M×N WSS configuration has superior functionality and performance compared to MCS in that it has no fundamental losses and automatically includes wavelength selection functionality. [Prior art documents] [Patent documents]

[0010] [Patent Document 1] Japanese Patent Application Laid-Open No. 2016-212128 [Non-patent literature]

[0011] [Non-Patent Document 1] M. Fukutoku, “Next Generation ROADM technology and applications”, paper M3A.4, OFC, 2015 [Non-patent document 2] T. Watanabe et al., “Silica-based PLC Transponder Aggregators for Colorless, Directionless, and Contention less ROADM”, paper OThD3.1, OFC, 2012 [Non-patent document 3] Y. Ikuma et al., “Low-loss transponder aggregator using spatial and planar optical circuit”, Journal of Lightwave Technology, Vol.34, No. 1, 2016 [Non-patent document 4] PD Colbourne et al. “Contentionless Twin 8x24 WSS with Low Insertion Loss”, paper Th4A.1, OFC, 2018 Summary of the Invention [Problem to be solved by the invention]

[0012] However, in an MxN WSS, it is necessary to install as many expensive WSSs as there are connected routes, which poses a problem in terms of scalability. For example, for eight routes, eight WSSs are required, and for 16 routes, 16 WSSs are required. The maximum number of WSSs that can be integrated in one optical system is limited by the area S of the liquid crystal element LC, M×N ∝ S There is a relationship between

[0013] The relationship between the maximum number of integrations and the area S of the liquid crystal element LC is described in, for example, Non-Patent Document 3 and Non-Patent Document 4. Non-Patent Document 3 and Non-Patent Document 4 show an example of a figure of merit when M = 8 and N = 24, i.e., M × N = 192. This is due to the limitations of the liquid crystal element LC, and it is difficult to expand the scale any further. When such limitations are imposed, for example, if an attempt is made to realize a 16-way TPA with M = 16, the number of connectable output Ns is limited to 12, even though 16 WSSs must be prepared.

[0014] Furthermore, the WSS of the M×N TPA is difficult to manufacture because it requires the spatial optical system and the waveguide optical system to be formed as a single unit. In particular, the M×1 switch described in Patent Document 1, which utilizes the thermo-optic effect, has reliability issues because the alignment of the optical system changes due to the generated heat.

[0015] The present invention has been made in view of the above points, and relates to providing an optical signal processing device that has small signal loss in the path from input to output and also provides a low-cost transponder aggregation function. [Means for solving the problem]

[0016] In order to achieve the above-mentioned object, one form of the optical signal processing device of the present invention is an optical signal processing device having N first ports and M second ports, and comprising: K first optical switches connected to K of the N first ports; L second optical switches connected to (NK) of the N first ports; M first optical multiplexers / splitters connected in a mesh configuration with each of the K first optical switches; L second optical multiplexers / splitters connected in a mesh configuration with each of the L second optical switches; a matrix switch connected to the second optical multiplexer / splitter; and M third optical multiplexers / splitters connected to the first optical multiplexer / splitter, the matrix switch, and the second ports. [Effects of the Invention]

[0017] According to the above embodiment, it is possible to provide an optical signal processing device that has a low signal loss in the path from input to output and that provides a low-cost transponder aggregation function. [Brief explanation of the drawings]

[0018] [Figure 1] FIG. 10 is a schematic functional block diagram for explaining the MCS disclosed in Non-Patent Document 2. [Figure 2] FIG. 2 is a functional block diagram for explaining a drop TPA among the TPAs ​​shown in FIG. 1. [Figure 3] FIG. 1 is a diagram for explaining a configuration in which a plurality of wavelength selective switches and non-wavelength selective switches are combined, as described in Patent Document 1. [Figure 4] 1 is a diagram for explaining an optical signal processing device according to an embodiment of the present invention; [Figure 5] FIG. 5 is a table showing the results of an example in which the optical signal processing device of the embodiment shown in FIG. 4 is used. DETAILED DESCRIPTION OF THE INVENTION

[0019] [TPA Configuration] 4 is a diagram illustrating an outline of a TPA4, which is an optical signal processing device according to one embodiment of the present invention. The TPA4 has a first port and a second port for outputting or inputting an optical signal to or from the outside. In this embodiment, port P, which is connected to N transponders 411 to 41N, corresponds to the first port, and cross-connect ports 481 to 48M, which are connected to an external cross-connect unit (not shown), correspond to the second port. The TPA4 also includes optical switches 421 to 42K, which are K first optical switches connected to K of the N transponders 411 to 41K, and optical switches 52(K+1) to 52N, which are L second optical switches connected to (NK) of the N transponders 41(K+1) to 41N. The TPA 4 also includes optical multiplexers and branchers (optical splitters) 431 to 43M, which are M first optical multiplexers and branchers connected in a mesh with the optical switches 421 to 42K, and L optical multiplexers and branchers 441 to 44L, which are connected in a mesh with the L optical switches 52(K+1) to 52N. The TPA 4 also includes a matrix switch 46 connected to the optical multiplexers and branchers 441 to 44L, and M optical multiplexers and branchers 471 to 47M connected to the optical multiplexers and branchers (optical splitters) 431 to 43M, the matrix switch 46, and the optical multiplexers and branchers 441 to 44L. The optical multiplexers / splitters 471 to 47M branch the optical signals passing between the optical multiplexers / splitters 431 to 43M and the cross-connect port 48 and output the branched optical signals to the matrix switch 46, or merge the optical signals output from the matrix switch 46 into the optical signals passing between the optical multiplexers / splitters 431 to 43M and the cross-connect port 48.

[0020] In this specification, in the above configuration, when there is no need to distinguish between the transponders 411 to 41N, they will simply be referred to as "transponder 41." Furthermore, when there is no need to distinguish between the cross-connect ports 481 to 48M, they will simply be referred to as "cross-connect port 48." Furthermore, when there is no need to distinguish between the optical switches 421 to 42K, they will simply be referred to as "optical switch 42." When there is no need to distinguish between the optical switches 52(K+1) to 52N, they will simply be referred to as "optical switch 52." When there is no need to distinguish between the optical multiplexer / branchers 431 to 43M, they will simply be referred to as "optical multiplexer / brancher 43." When there is no need to distinguish between the optical multiplexer / branchers 441 to 44L, they will simply be referred to as "optical multiplexer / brancher 44." When there is no need to distinguish between the optical multiplexer / branchers 471 to 47M, they will simply be referred to as "optical multiplexer / brancher 47."

[0021] In this embodiment, the transponder 41 side is the input and the cross-connect port 48 side is the output. As shown in FIG. 4, the number of inputs N is set to "8" and the number of outputs M is set to "8." The N input optical signals are input to the optical switch 42 and the optical switch 52. If the number of optical signals input to the optical switch 42 out of the N is K, the number L of optical signals input to the optical switch 42 is equal to N minus K. The optical switch 42 is a 1-input, M-output (1×8) optical switch, the optical switch 52 is a 1-input, L-output (1×4) optical switch, the optical multiplexer / splitter 43 is K inputs, 1 output (4×1), and the optical multiplexer / splitter 44 is L inputs, 1 output (4×1). Therefore, in this embodiment, the 1×8 optical switches 421 to 424 input the optical signals from the transponders 411 to 414. Furthermore, 1x4 optical switches 525 to 528 input optical signals from transponders 415 to 418. However, it should be understood that this embodiment does not limit the numerical values ​​of N, M, K, and L to the above example, and K may be any integer smaller than N.

[0022] The TPA 4 can also operate with the transponder 41 side as the output and the cross-connect port 48 side as the input. In other words, the TPA 4 operates as the DEMUX drop side. In such a case, the input-output relationships of the optical switches 42 and 52 and the optical multiplexers / splitters 43, 44, and 47 are naturally reversed.

[0023] Furthermore, in this embodiment, K optical filters 511 to 51K and 51(K+1) to 51N are connected between the transponder 41 and the optical switch 42, and between the transponder 41 and the optical switch 52, respectively (when there is no need to distinguish between them, they will be simply referred to as "optical filters 51"). However, this embodiment is not limited to a configuration including the optical filter 51, and when there is no need to filter the optical signal, the optical filter 51 may be removed from the TPA 4 to reduce the number of parts.

[0024] The eight output signals of each optical switch 42 are connected in a mesh pattern to each of the eight optical multiplexers and branchers 43. Here, "mesh pattern" means that the eight output signals output from one optical switch 42 are input to all eight optical multiplexers and branchers 43, as shown in FIG. 4. According to the mesh connection, a total of four optical signals are input to each of the optical multiplexers and branchers 43. Furthermore, the four output signals of each optical switch 52 are connected in a mesh pattern to each of the four optical multiplexers and branchers 44. Therefore, a total of four optical signals are input to each of the optical multiplexers and branchers 44. Such optical multiplexers and branchers 43 and optical switches 42 form a 4 × 8 multicast. Furthermore, the optical multiplexers and branchers 44 and optical switches 52 form a 4 × 4 multicast switch.

[0025] The optical multiplexer / splitter 47 branches the optical signal between the optical multiplexer / splitter 43 and the cross-connect port 48, and connects the branched optical signal to the matrix switch 46. The output signal of the optical multiplexer / splitter 44 is also connected to the matrix switch 46. Optical amplifiers 451 to 45L are connected between the matrix switch 46 and the optical multiplexer / splitter 44 (hereinafter, when there is no need to distinguish between them, they will be simply referred to as "optical amplifier 45") and amplify the optical signal traveling from the optical multiplexer / splitter 44 to the matrix switch 46. A plurality of nodes connecting the optical multiplexers / splitters 441 to 44L and the optical amplifiers 451 to 45L will be referred to as an expansion port group 491, and a plurality of nodes connecting the optical amplifiers 451 to 45L and the matrix switch 46 will be referred to as an expansion port group 492.

[0026] In the above configuration, optical amplifier 45 is provided on extended nodes 491 and 492. However, the optical amplifier need only amplify the optical signals input from transponders 41(K+1) through 41N, and is not limited to being provided on extended nodes 491 and 492. Optical amplifier 45 is preferably placed between optical multiplexer / splitter 47 and optical multiplexer / splitter 44.

[0027] In this embodiment, the above configuration may be formed and integrated on an optical waveguide substrate 470. Furthermore, the above configuration may be a combination of multiple elements that exhibit the above functions.

[0028] Next, optical signal transmission in the above configuration will be described. Optical signals input to the TPA 4 by transponders 411 to 41K pass through optical filters 51 and enter four optical switches 42. The optical switch 42 selects one of the optical multiplexers / splitters 43 and outputs an output signal. The optical multiplexer / splitter 43 selects and multiplexes up to four of the input optical signals, and outputs the multiplexed signals to the optical multiplexer / splitter 47. Meanwhile, optical signals input to the TPA 4 by transponders 41(K+1) to 41N pass through optical filters 51 and enter the optical switch 52. The optical switch 52 selects one of the optical multiplexers / splitters 44 and outputs an output signal. The output signal from the optical multiplexer / splitter 44 is amplified by an optical amplifier 45 and connected to a matrix switch 46. The matrix switch 46 outputs an output signal to the optical multiplexer / splitter 47, where it is multiplexed with the output signal from the optical multiplexer / splitter 43.

[0029] [Action and effect] The TPA 4 of this embodiment has the same function as an M×N multicast switch, but is advantageous in terms of loss. The following describes the operation and effects of the TPA 4, which is an optical signal processing device of this embodiment. This description will be given taking as an example the operation of the TPA 4 as a DeMux (DeMultiplexer) as viewed from the cross-connect port 48 side. That is, in the example of M=8 and N=8 shown in FIG. 4 , an optical signal input from the cross-connect port 48 is branched by eight optical multiplexers / branchers 47, with one portion input to the optical multiplexer / brancher 43 and the other portion propagating to the matrix switch 46. Here, we consider a case where the input optical signal is branched into two, with a branching ratio of γ:1−γ.

[0030] Here, the effect of splitting an optical signal by the optical multiplexer / splitter 47 will be explained. It is known that the degree of signal loss in a TPA varies depending on the number of transponders connected. For example, in the case of 8 inputs (or 8 outputs) or 4 outputs (or 4 outputs), the signal loss from input to output is small enough to be tolerable. The inventors of the present invention focused on this point and split the optical signal in the 8x8 TPA 4, dividing it into an 8-input, 4-output multicast switch and a 4-input, 4-output multicast switch with low signal loss for processing. That is, the cross-connect ports 481 to 48M and the transponders 411 to 41K constitute one multicast switch, and the cross-connect ports 481 to 48M and the transponders 41(K+1) to 41N constitute one transponder aggregation function. In this way, this embodiment minimizes signal loss between the transponders 411 to 41K and the cross-connect ports 481 to 48M.

[0031] In the above configuration, the preferred value of the ratio (mux / split ratio) γ at which the optical multiplexer / splitter 47 combines or distributes optical signals from the optical multiplexer / splitter 43 and the matrix switch 46 varies depending on the application, number of input / output ports, and specifications of the TPA 4. To meet these conditions, the optical multiplexer / splitter 47 of this embodiment may have a function to change the mux / split ratio. Furthermore, in this embodiment, as shown in FIG. 4, an optical amplifier 45 may be provided between the matrix switch 46 and the optical multiplexer / splitter 44 to compensate for signal loss of the signal distributed to the matrix switch 46.

[0032] Next, the optical signal input from the cross-connect port 48 to the optical multiplexer / splitter 43 has its power split into four by the optical multiplexer / splitter 43, and is then input to the transponders 411 to 41K via the optical switch 42. The above optical signal path is referred to as "path 1" here. The propagation loss I from the cross-connect port 48 to the transponder 41 on path 1 is L1 is expressed by the following equation (1). I L1=-10×log(γ)-10×log(1 / K)...Equation (1)

[0033] On the other hand, of the eight optical signals branched by the optical multiplexer / brancher 47 and propagated to the matrix switch 46, four are selected by the matrix switch 46 and output to the expansion port group 492. Then, the optical signals pass through the expansion port group 491 via the optical amplifier 45, and are branched into four by the optical multiplexer / branchers 441 to 44L. Each of the branched optical signals is input to the optical switches 52(K+1) to 52N, passes through the optical filter 51, and reaches the transponders 41(K+1) to 41N. In this embodiment, this path is referred to as path 2. The propagation loss I from the cross-connect port 48 of path 2 to the transponder 41 is L2 is expressed by the following formula (2). In formula (2), G amp is the gain of the optical amplifier 45. I L2 =-10×log(1-γ)-10×log(1 / (NK))-G amp ...Equation (2)

[0034] This embodiment, which includes the optical amplifier 45, can compensate for the loss of optical signals passing through path 2, while optical signals passing through path 1 suffer losses of −10×log(γ) and −10×log(1 / K) due to branching. However, this embodiment can sufficiently reduce signal loss on path 1 by adjusting the γ of the optical multiplexer / splitter 47 and the number of transponders (i.e., K) connected to the optical switch 52 among the transponders 41. Then, after setting optimal conditions for reducing signal loss on path 1, the optical amplifier 45 can compensate for signal loss occurring on path 2. This configuration is advantageous in reducing the number of optical amplifiers and preventing an increase in the number of components and size of the device. Furthermore, as described above, in this embodiment, the number of optical switches 52 and optical multiplexer / splitters 44 is set to L, which is the value obtained by subtracting the number of transponders K connected to path 1 from the number of inputs N. This is because it is only necessary to transmit signals to path 2 equal to the number of transponders connected to path 2.

[0035] As described above, the optical signal processing device of this embodiment does not use a WSS, but branches optical signals from the transponder 41 to the cross-connect port 48 to form an 8-input, 4-output multicast switch and a 4-input, 4-output multicast switch with low signal loss, thereby reducing signal loss in the optical signal processing device. This configuration reduces the number of components and suppresses increases in device area compared to known configurations in which a WSS is provided for each path. Furthermore, even when optical amplifiers 45 are provided for the branched paths in this embodiment, the number of optical amplifiers 45 is only L, which is the number of paths N minus K. Therefore, the number of optical amplifiers required is clearly smaller than in known configurations. Furthermore, the optical signal processing device of this embodiment has the advantage that, even when the number of paths M is increased to 12, 16, etc., the number of required optical amplifiers does not increase, or can be suppressed, by appropriately setting the number K. This embodiment thus provides an optical signal processing device that can extract high-quality signals while suppressing the cost of the transponder aggregation function. [Example]

[0036] Next, an example of the embodiment described above will be explained using specific numerical values. In this embodiment, the conditions set are that the number of transponders N is 8, the number of cross-connect ports is M, and the joining / branching ratio γ is 0.9. In this embodiment, the cross-connect ports are the input side, and the transponders are the output side. Under these conditions, 90% of the optical signals input from the M ports are branched to path 1 described in the embodiment, and 10% are branched to path 2. At this time, the inventors changed K to adjust the gain G of the optical amplifier 45 (FIG. 4) so ​​that the signal loss in each path 1 and the signal loss in path 2 are equal. amp (including the signal loss of the optical filter). The results are shown in the table in Figure 5. The theoretical loss of a known N=8 multicast switch is 9 dB.

[0037] In the table of FIG. 5, γ is the combining / branching ratio of the optical combining / branching device 47 of FIG. 4, K is the number of optical switches 42 connected to the transponders 411 to 41K, G amp indicates the gain of the optical amplifier 45, path 1 indicates the path between the transponder 411 to 41K and the cross-connect port 48, path 2 indicates the path between the transponder 41(K+1) to 41N and the cross-connect port 48, and L indicates NK, i.e., the number of optical switches 52 connected to the transponders 41(K+1) to 41N. As shown in FIG. 4, this L matches the number of optical amplifiers 45 in FIG.

[0038] As shown in the table in Figure 5, when K = 2 and the optical amplifier gain Gamp is set to 14.3 dB, the signal loss on both paths 1 and 2 is 3.5 dB. When K = 4 and the optical amplifier gain Gamp is set to 9.5 dB, the signal loss on both paths 1 and 2 is 6.5 dB. When K = 6 and the optical amplifier gain Gamp is set to 4.8 dB, the signal loss on both paths 1 and 2 is 8.2 dB. In both cases, this is smaller than that of a known multicast switch. Furthermore, the combining / branching ratio γ of the optical combining / branching device 47 can be tunable using a Mach-Zehnder interferometer or the like formed on the optical waveguide. This can relax the gain performance requirements for the optical amplifier 45. As shown in the table in Figure 5, when γ = 0.8 and K = 2, an optical amplifier with a gain approximately 3.5 dB lower than that when γ = 0.9 is sufficient. Even in such a case, the signal loss as a TPA is sufficiently reduced to 4 dB compared to the theoretical loss of 9 dB. [Explanation of symbols]

[0039] 41 Transponder 42,52 Optical switch 43, 44, 47 Optical multiplexer / splitter 45 Optical Amplifier 46 Matrix Switch 48 cross-connect ports 51 Optical Filter 470 Optical waveguide substrate 491,492 expansion ports

Claims

1. An optical signal processing device having N first ports and M second ports, K first optical switches connected to K of the N first ports; L second optical switches connected to (N−K) of the N first ports; M first optical multiplexers / splitters connected in a mesh pattern to the K first optical switches, respectively; L second optical multiplexers / splitters connected in a mesh pattern to the L second optical switches, respectively; a matrix switch connected to the second optical multiplexer / splitter; M third optical multiplexers / splitters connected to the first optical multiplexer / splitter, the matrix switch, and the second port; Equipped with The optical signal processing device, wherein the L is equal to a number obtained by subtracting the K from the N.

2. 2. The optical signal processing device according to claim 1, wherein the first optical switch is a 1×M optical switch, the second optical switch is a 1×L optical switch, the first optical multiplexer / splitter is a K×1 optical multiplexer / splitter, the second optical multiplexer / splitter is an (N-K)×1 optical multiplexer / splitter, and the matrix switch is an L×M matrix switch.

3. 2. The optical signal processing device according to claim 1, wherein the first optical switch is an M×1 optical switch, the second optical switch is an L×1 optical switch, the first optical multiplexer / splitter is a 1×K optical multiplexer / splitter, the second optical multiplexer / splitter is a 1×(N-K) optical multiplexer / splitter, and the matrix switch is an M×L matrix switch.

4. It also has an expansion port, The optical signal processing device according to claim 1 , wherein the matrix switch and the second optical multiplexer / splitter are connected via the expansion port.

5. 5. The optical signal processing device according to claim 4, wherein at least a portion of the first port, the second port, the first optical switch, the second optical switch, the first optical multiplexer / splitter, the second optical multiplexer / splitter, the third optical multiplexer / splitter, the matrix switch, and the expansion port are formed on a common optical waveguide substrate.

6. The optical signal processing device according to claim 1 , further comprising an optical amplifier between the third optical multiplexer / splitter and the second optical multiplexer / splitter.

7. The optical signal processing device according to claim 1 , wherein the third optical multiplexer / splitter has a function of changing a multiplexing / splitter ratio.

Images