Optical signal processing device

Abstract

To provide an optical signal processing device which reduce the number of installation of an element for switching a node of an optical signal in each wavelength, and has an advantage in a prevention of an increase of a large scale of the element in a low cost.SOLUTION: A TPA 1 having: N ports P1; and M cross-connect ports 17, comprises: a port P3 connected to each port P1; N optical switches 12 having M+Nf ports P4 connected to each port P1; a port P5 connected to any one of first to Mth ports from the M+Nf ports P4; M optical variable confluence directional couplers 13 having a port P6 which is not connected to each port P4; Nf ports P9 connected to any one from M+1th to M+Nf ports of each port P4; a wavelength selection directional coupler 14 having a port P10 which is not connected to each port P4; and an optical confluence directional coupler 16 connected to the port P6, each cross-connect port 17, and a port P10. Each Nf is set to a value obtained by rounding down a residual number by dividing N with K+1 when the number of optical signal which can be connected to one cross-connect port 17 is set to K.SELECTED DRAWING: Figure 6

Description

Technical Field

[0001] The present invention relates to an optical signal processing apparatus.

Background Art

[0002] Due to the spread of data communication networks such as the Internet, further increase in the capacity of optical communication networks is desired. To meet such an expansion of network demand, wavelength division multiplexing (WDM) communication has been put into practical use. In recent years, the demand for wavelength selective switches (WSS) that switch optical signal nodes for each wavelength without converting optical signals into electrical signals has also been increasing. The configuration of a node using a wavelength selective switch is called a reconfigurable optical add / drop multiplexing (ROADM) system. Non-Patent Document 1 discloses a node having a configuration called a multi-degree ROADM that processes optical signals from a plurality of nodes. Non-Patent Document 2 discloses a configuration using a multicast switch (MCS) that functions as a transponder aggregator (TPA) that connects optical signals from an arbitrary node to an arbitrary transceiver (transponder) in such a system.

[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 section (WXC: Wavelength Cross Connect) 101 composed of a multicast switch 102 that combines an SPL (splitter) and a WSS, and a WSS 103. WDM signals are input to the optical node from M input fibers 111 to 11M. Among the input signals, the signals to be added / dropped at the optical node are sent to TPA 104a which is a drop TPA, and then transmitted from TPA 104a to any one of N transponders 131 to 13N. Similarly, the optical signals generated in transponders 131 to 13N are output to any one of output fibers 121 to 12M via TPA 104b which is an add TPA and the WSS 103 of the cross-connect section 101. In this specification, with reference to the element of interest, the node where an optical signal is input (received) or output (transmitted) is hereinafter also referred to as the "forward path".

[0004] FIG. 2 is a functional block diagram for explaining TPA 104a which is a drop TPA among the TPAs shown in FIG. 1. However, since TPA 104b which is an add TPA has the same configuration as TPA 104a, in this specification, the description of TPA 104a is used instead of the description of TPA 104b. 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 each of the 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 input and filter the output of the M×1 switch 211. The optical filters 221 to 22N output the filtered signals to the transponders 131 to 131N shown in FIG. 1 respectively.

[0005] However, since the MCS involves splitting of the input optical signal, it inherently involves loss of the optical signal. That is, in the MCS, since the optical signal is split into N by the 1×N splitters 201 to 20M, the intensity of the optical signal decreases by 1 / N. For example, a theoretical loss of 9 dB occurs when 8 transponders are connected, and a theoretical loss of 12 dB occurs when 16 transponders are connected. Generally, since a transponder requires an input of optical power equal to or greater than the minimum reception sensitivity (the minimum optical power capable of receiving a signal), there is an upper limit to the number M that can be split. To relax the regulation by this upper limit, an optical amplifier such as an EDFA (Erbium-doped Fiber Amplifier) may be inserted between the multicast switch 102 and the TPA shown in FIG. 1.

[0006] However, while the EDFA increases the intensity of the optical signal, it has the drawback of degrading the OSNR (Optical Signal-to-Noise Ratio) because it generates ASE (Amplified Spontaneous Emission) noise. To solve this, as shown in FIG. 2, it is conceivable to insert optical filters 221 to 22N between the transponder and the MCS to remove ASE. However, the insertion of optical filters has drawbacks such as increasing the number of components and leading to an increase in signal loss that should have been compensated for by the EDFA.

[0007] Patent Document 1 describes combining a plurality of wavelength selective switches (WSSs) and non-wavelength selective switches to realize a TPA function (M×N WSS) without theoretical loss in order to eliminate the loss of the optical signal. FIG. 3 is a diagram for explaining such a configuration of Patent Document 1, and shows an M×N WSS by an optical system called a SPOC (Spatial and Planar Optical Circuit) that combines an optical waveguide and a spatial optical system. The configuration shown in FIG. 3 can be said to be a replacement of the M 1×N splitters in the MCS shown in FIG. 2 with M 1×N WSSs for selecting transponders connected by a spatial optical system.

[0008] In the optical path shown 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 section 101 shown in FIG. 1. The WDM signal is converted into collimated light through an SBT (Space Beam Transformer) circuit having a lens function integrated in an optical waveguide and output to a free-space optical system. At this time, the input waveguides a, b, and c are arranged so that the emission directions of the light waves are different. In the free-space optical system, the WDM signal is wavelength-dispersed in the direction perpendicular to the plane of the paper by the diffraction grating Gr, and then passes through a plurality of lenses L SP , L DP and is incident on the liquid crystal element LC, which is a switching element, at different positions in the y-axis direction for each wavelength. At this time, depending on the input waveguides a, b, and c, the liquid crystal element LC is incident at different positions A, B, and C in the x-axis direction. The liquid crystal element LC is composed of a plurality of pixel elements that can individually modulate the optical phase two-dimensionally. In the liquid crystal element LC, the light is deflected and reflected in the vertical direction (x-axis direction) of the plane of the paper for each wavelength. That is, the optical signals wavelength-dispersed in the direction perpendicular to the plane of the paper are deflected and reflected in different directions and independently for each of the input waveguides a, b, and c. The reflected optical signals are coupled to any one of SBT11_1 to 11_4 and input to the optical waveguide optical system. The optical signals input to the optical waveguide optical system propagate through any one of the connection waveguides 12_1-a to 12_1-c, 12_2-a to 12_3-c,..., 12_4-a to 12_4-c according to the input waveguides a, b, and c, pass through N M×1 switches 13_1 to 13_3, and are output from any one of the output waveguides 1 to 4 and input to the transponder.

[0009] The M×N WSS described above does not cause loss due to branching. In addition, since the WSS function composed of the free-space optical system provides a wavelength filtering function, noise components of wavelengths other than the signal wavelength are removed. Thus, the M×N WSS configuration has excellent functions and performance compared to MCS in that there is no theoretical loss and the wavelength selection function is automatically included.

Prior Art Documents

Patent Documents

[0010]

Patent Document 1

Non-Patent Document

[0011]

Non-Patent Document 1

Non-Patent Document 2

Non-Patent Document 3

Non-Patent Document 4

Summary of the Invention

Problems to be Solved by the Invention

[0012] However, in the M×N WSS, it is necessary to install as many expensive WSSs as the number of paths to which the WSSs are connected, and this is a problem from the perspective of scalability. For example, if it is an 8-path system, 8 WSSs are required, and if it is a 16-path system, 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 and there is a relationship of.

[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. In Non-Patent Document 3 and Non-Patent Document 4, when M = 8 and N = 24, that is, an example of a performance index of about M×N = 192 is shown. This is due to the limitation of the liquid crystal element LC, and it is difficult to expand to a larger scale. When subject to such limitations, for example, when trying to realize a 16-path TPA with M = 16, although 16 WSSs for 16 paths need to be prepared, the number of connectable outputs N is limited to 12.

[0014] Furthermore, since the WSS of the M×N TPA needs to integrally form a free-space optical system and a waveguide optical system, it is difficult to manufacture. In particular, the M×1 switch using the thermo-optical effect described in Patent Document 1 has a problem from the perspective of reliability 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 an optical signal processing device that is advantageous in reducing the number of installed elements for switching optical signal nodes for each wavelength, reducing costs, and preventing large-scale elements.

Means for Solving the Problems

[0016] To achieve the above object, an optical signal processing apparatus according to one embodiment of the present invention is an optical signal processing apparatus having N first ports and M second ports, the optical signal processing apparatus including: N optical switches each having a third port connected to the first port and M + Nf fourth ports not connected to the first port; M optical variable combiners and splitters each having a fifth port connected to any one of the first to M-th ports among the M + Nf fourth ports and a sixth port not connected to the fourth port; a matrix switch having Nf first extension ports connected to any one of the (M + 1)-th to (M + Nf)-th ports among the M + Nf fourth ports and a second extension port not connected to the fourth port; and an optical combiner and splitter connected to the sixth port, the second port, and the tenth port, where Nf is a value obtained by dividing N by K + 1 and rounding down the remainder, where K is the number of optical signals connectable to one of the second ports.

Advantages of the Invention

[0017] According to the above embodiment, it is possible to provide an optical signal processing apparatus that reduces the number of installation bases of elements for switching optical signal nodes for each wavelength, is advantageous for reducing costs, and preventing the scale-up of elements.

Brief Description of the Drawings

[0018]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Embodiments for Carrying Out the Invention

[0019] Prior to the following specific description of the embodiments of the invention, the concept of the present invention will be described. The function of TPA includes a function of dropping optical signals transmitted from other nodes other than a predetermined node of interest (hereinafter also referred to as "own node") at the own node, and an add function of configuring a connection for transmitting an optical signal to be transmitted from the own node to other nodes. Hereinafter, an example of the add function of transmitting an optical signal from the own node to other nodes will be used for explanation. Also, the phrase "connecting an optical signal" in this specification refers to making it possible to transmit or receive an optical signal to / from a connection target.

[0020] Here, let the number of ports to other nodes connected to the TPA be M, and the number of ports to the transponder be N. Generally, N is often a multiple of 2, and N = 4, 6, 8, 12, 16, 24, etc. are often set. FIGS. 4(a), 4(b), and 4(c) are diagrams showing the combinations of the number of possible paths and the number of optical signals in each path when the transponder is set to 8 (FIG. 4(a)), 7 (FIG. 4(b)), and 9 (FIG. 4(c)), respectively. In the case where N is set to 8 as shown in FIG. 4(a), the maximum number of optical signals input to or output from the TPA is 8. The example with the serial number 1 in FIG. 4(a) shows that all 8 optical signals are connected to one path. Also, the example with the serial number 7 shows that 4 optical signals are connected to each of the 2 paths. The example with the serial number 21 shows that one optical signal is connected to each of the 8 paths. Here, the wavelengths of the optical signals passing through the same path are all different.

[0021] Also, in FIG. 4(a), the relationship between the number of paths to be connected and the maximum value of the number of signals routed to the path with the minimum number of connected signals is as follows. The following notation indicates {the number of paths to be connected, the maximum number of optical signals}. {1, 8}, {2, 4}, {3, 2}, {4, 2}, {5, 1}, {6, 1}, {7, 1}, {8, 1}

[0022] Among the above, for example, {1, 8} corresponds to the serial number 1 in FIG. 4(a), indicating that there is 1 path to be connected, and 8 optical signals with different wavelengths are assigned to this path. Also, for example, {5, 1} corresponds to the serial number 11 in FIG. 4(a), indicating that there are 5 paths to be connected, and 1 optical signal is assigned to 2 of the 5 paths.

[0023] Similarly, in the case of N = 16, {Number of paths to be connected, maximum number of optical signals} = {1, 16}, {2, 8}, {3, 5}, {4, 4}, {5, 3}, {6, 2}, {7, 2}, {8, 2}, {9, 1}, {10, 1}, {11, 1}, {12, 1}, {13, 1}, {14, 1}, {15, 1}, {16, 1}. From this point of view, for the number of paths Mmax to be connected and the maximum number Nmax of optical signals routed to the path with the minimum number of connected optical signals, Nmax = Int(N / Mmax) ··· Equation (1) it can be seen that there is such a relationship. In Equation (1), Int() represents truncation. The above {number of paths to be connected, maximum number of optical signals} corresponds to {Mmax, Nmax} in Equation (1). Therefore, for example, under the condition of {1, 8}, 8 = Int(8 / 1) holds. Also, under the condition of {5, 1}, 1 = Int(8 / 5) holds.

[0024] From the above content, in an M × N WSS-type TPA, it can be seen that from the perspective of noise cut, the number of wavelength selective multiplexers (WSSs) that should be originally installed is not necessary for the number of paths. For example, when N = 8, the maximum number of paths that can be connected is 8. When the number of paths to be connected is 5 or more, only one optical signal is routed, or there are 5 or more paths where the signal is not routed, and these paths do not require a wavelength selection function. The maximum number of paths to which a plurality of optical signals that require a wavelength selection function are connected is 3 (serial number 18 in Fig. 4(a)). Similarly, when N = 8 and the number of paths to be connected is 4, the maximum number of paths to which a plurality of optical signals are connected is 4 (serial number 17 in Fig. 4(a)). Therefore, the wavelength filter to be installed for noise cut, that is, the wavelength selective multiplexer, only needs to be Nf = 4 at most. When the number of paths to be connected is 3, wavelength selective multiplexers may be provided for all three paths. However, since 3 is smaller than 4, in the TPA when N = 8, it can be said that the maximum number of wavelength selective multiplexers to be installed is 4.

[0025] More generally, the maximum number Nf of wavelength selective multiplexers / demultiplexers only needs to be N / 2. Therefore, when N = 16, the maximum number of wavelength selective multiplexers / demultiplexers that can be installed is 8. In the above, examples where N is a multiple of 2 such as 8 and 16 are shown (as will be described later, this 2 means that one wavelength is routed in the same path). However, when N is not divisible by 2, the maximum number of wavelength selective multiplexers / demultiplexers to be installed, Nf = Int(N / 2). Such examples are shown in the case of N = 7 in Fig. 4(b) and the case of N = 9 in Fig. 4(c).

[0026] Furthermore, consider the case where there are two or more optical signals that can be routed without a filter in one path. That is, the above consideration is based on the premise that wavelength selective multiplexers / demultiplexers are provided for all paths to which a plurality of optical signals are connected. However, if there is an allowable range in the number of optical signals connected to one path, the maximum number of wavelength selective multiplexers / demultiplexers to be installed will be even less. For this reason, consider the case where K (K > 2) optical signals can be routed without a filter in one path. Figs. 5(a), 5(b) and 5(c) all show cases where N is 8 and the number K of optical signals that can be routed without a filter is 2, 3, and 4, respectively, in tabular form. In Fig. 5, the number of paths that require a filter is shown shaded. That is, the table in Fig. 5(a) indicates that the optical signals generated by a certain transponder are not affected by or can be ignored the ASE from other transponders connected to the same path. In other words, an optical signal of a predetermined wavelength is not affected by the ASE of that optical signal if there is one (a total of two wavelengths) of optical signals of other wavelengths routed in the same path, but is affected if there are two (a total of three wavelengths) of optical signals of other wavelengths routed in the same path. From this perspective, in the table of Fig. 5(a), filters are required for serial numbers 1 to 16, and among these, filters need to be provided for two paths at serial numbers 4, 7, 8, 12, and 13. Similarly, in the table of Fig. 5(b), filters need to be provided for two paths at serial number 7, and in the table of Fig. 5(c), one filter needs to be provided for serial numbers 1 to 6.

[0027] Considering the above inductively, the maximum number of channels for which a wavelength selective multiplexer / demultiplexer is required, i.e., the maximum number Nf of wavelength selective multiplexers / demultiplexers to be installed, is Nf = Int(N / (K + 1)) ··· Equation (2) It can be seen that this is the case. Therefore, the number of wavelength selective multiplexers / demultiplexers to be installed is determined by K specified by the optical signal-to-noise ratio (OSNR: Optical Signal-to-Noise Ratio) allowed by the transmission design of the network system, and it can be seen that the number of wavelength selective multiplexers / demultiplexers to be installed may not be required for M channels connected to the node. An example that does not satisfy this is the case where M < Nf = Int(N / (K + 1)), and examples of this include the examples described in Non-Patent Document 3 and Non-Patent Document 4. In these examples, since M = 8 and N = 24, when K = 1, M = 8 < Int(N / (K + 1)) = 12

[0028] Therefore, when N = 24 and the optical signal is most evenly distributed to each channel, a known M×N WSS configuration is preferable. However, even when N = 24, when the number of channels exceeds 12, the configuration of the embodiment of the present invention described above is advantageous. Also, the maximum number Nf of wavelength selective multiplexers / demultiplexers to be installed when allowing up to 3 optical signals to be connected to one channel is as follows. Nf = Int(24 / (3 + 1)) = 6 From the above, it is clear that this embodiment is advantageous in reducing the number of installed wavelength selective multiplexers / demultiplexers compared to known configurations and suppressing an increase in the element area.

[0029] [First Embodiment] Based on the above considerations, a first embodiment of an optical processing apparatus of the present invention that reduces the number of WSSs to be installed will be described. FIG. 6 is a circuit diagram for explaining TPA1, which is an optical signal processing apparatus of the first embodiment. In this description, the side of the transponder is taken as the input, and the side of cross-connect port 17 is taken as the output. However, TPA1 can also operate with the side of transponder 11 as the output and the side of cross-connect port 17 as the input.

[0030] The TPA1 of the first embodiment is an optical signal processing device having N first ports and M second ports. In the first embodiment, the first ports are ports P1 connected to transponders 111 to 11N, and the second ports are cross-connect ports 171 to 17M connected to a cross-connect section (not shown). Also, in the first embodiment, transponders 111 to 11N are described as the input side, and cross-connect ports 171 to 17N are described as the output side. However, it goes without saying that the TPA1 can operate in the opposite manner, with cross-connect ports 171 to 17M as the input side and transponders 111 to 11N as the output side. Also, in the first embodiment, both the numbers of N and M are set to 8.

[0031] TPA1 includes N optical switches 121 to 12N having ports P3 connected to port P1 and M + Nf ports P4 not connected to port P1, M optical variable combiners / splitters 131 to 13M having ports P5 connected to any of the first to M-th ports among ports P4 and port P6 not connected to port P4, and Nf first extended ports which are ports P9 connected to any of the (M + 1)-th to (M + Nf)-th ports among M + Nf ports P4 and a second extended port which is port P 10 having a matrix switch 15, port P6, a cross-connect port 17, and an optical combiner 16 not connected to port P10. Ports P4 and port P9 of matrix switch 15 may be directly connected or indirectly connected via other elements. In the first embodiment, TPA1 includes wavelength selective multiplexers / demultiplexers 141, 14Nf, and port P9 is indirectly connected to port P4 via wavelength selective multiplexers / demultiplexers 141, 14Nf. The ports of wavelength selective multiplexers / demultiplexers 141, 14Nf connected to port P4 are port P7, and the ports not connected to port P4 are port P8. And in the first embodiment, in such a configuration, the number of optical signals connectable to one cross-connect port 17 is K, and Nf is the value obtained by dividing N by K + 1 and rounding down the remainder, that is, Nf = Int(N / (K + 1)) (where Int() represents rounding down).

[0032] In the above, "a port P5 that is connected to any one of the first to Mth ports among the ports P4 of a certain optical switch 12" means that any one of the first to Mth ports of the port P4 is connected to one optical variable combiner / splitter. Also, "a port P7 that is connected to the (M + 1)th to (M + Nf)th ports among the ports P4 of a certain optical switch 12" means that any one of the ports of the (M + 1)th to (M + Nf)th ports of the port P4 is connected to one wavelength selective combiner / demultiplexer. For example, when M is 8 and Nf is 2, the optical variable combiner / splitter 131 is connected to the first port P4 of each optical switch 12. Also, the optical variable combiner / splitter 13M is connected to the Mth port P4 of each optical switch 12. Similarly, the wavelength selective combiner / demultiplexer 141 is connected to the 9th port P4 of each optical switch 12. 14Nf is connected to the 10th port P4 of each optical switch 12.

[0033] Among the above configurations, in this specification, when it is not necessary to distinguish each of the transponders 111 to 11N, it is simply denoted as the transponder 11. Also, in this specification, when it is not necessary to distinguish each of the optical switches 121 to 12N, it is simply denoted as the optical switch 12, when it is not necessary to distinguish each of the optical variable combiner / splitters 131 to 131M, it is simply denoted as the optical variable combiner / splitter 13, when it is not necessary to distinguish the wavelength selective combiner / demultiplexers 141, 14Nf, it is simply denoted as the wavelength selective combiner / demultiplexer 14, and when it is not necessary to distinguish each of the optical combiner / splitters 161 to 16M, it is simply denoted as the optical combiner / splitter 16. The optical switch 12 is a (1×(N + Nf)) switch having one port P3 and N + Nf ports P4. The optical variable combiner / splitter 13 is an (N×1) optical variable combiner / splitter having N ports P5 and one port P6. The wavelength selective combiner / demultiplexer 14 is an (N×1) wavelength selective combiner / demultiplexer having N ports P7 and one port P8. The matrix switch 15 has Nf ports P9 and M ports P 10 and.

[0034] The above configuration may be formed on a common optical waveguide substrate 6. Further, the above configuration may be divided into a plurality of units, each of which may function in cooperation. Also, the connection between the optical switch 12 and the wavelength selective multiplexer / demultiplexer 14 may be made via an expansion port group.

[0035] In the TPA1 shown in FIG. 6, the optical signals of N transponders 11 are connected to the cross-connect port 17. First, the optical signal output from the transponder 11 is input to any one of the N optical switches 12. The optical switch 12 routes the input optical signal of the transponder 11 to any one of the ports P4. Among the outputs of the optical switch 12, M outputs are connected to any one of the optical variable combiner / splitters 13, and the remaining Nf outputs are input to the wavelength selective multiplexer / demultiplexer 14. The optical variable combiner / splitter 13 is a combiner / splitter having a function of selectively combining K signals out of the N optical switches 12 without wavelength dependence. The optical variable combiner / splitter 13 has a principle loss in the first embodiment, and the signal loss is 10×log(1 / K) dB. The configuration of such an optical variable combiner / splitter 13 will be described in detail later.

[0036] The wavelength selective multiplexer / demultiplexer 14 may be a general wavelength selective switch and has N ports P7. Each of the ports P7 is connected to any one of the ports P4 of the optical switches 121 to 12N. Further, the port P8 of the wavelength selective multiplexer / demultiplexer 14 is connected to the port P9 of the matrix switch 15. The matrix switch 15 has a function of connecting the number of wavelength selective multiplexer / demultiplexers 14, that is, Nf input signals to M outputs without internal blocking. The optical signal output from the port P 10 is combined with the optical signal output from the port P6 by the optical combiner 16 and output to the cross-connect port 17.

[0037] Next, the operation of TPA1 will be described using the tables in FIGS. 5(a) to 5(c). As described above, the tables in FIGS. 5(a) to 5(c) all have an input number M = 8 and an output number N = 8, and the allowable amount K of ASE noise superimposed on one path is for the cases where K is 2, 3, and 4 respectively. In the case of the table in FIG. 5(a), since K is 2, Nf = Int(8 / (2 + 1)) = 2 holds, and the number of wavelength selective multiplexers / demultiplexers 14 is 2. In the first embodiment, for the cross - connect ports 17 corresponding to the paths where more than two optical signals are routed, the optical signals that have passed through the wavelength selective multiplexers / demultiplexers 14 among the optical signals output from any of the transponders 11 are connected. On the other hand, for the cross - connect ports 17 corresponding to the paths where two or fewer optical signals are routed, the optical signals output from any of the transponders 11 are connected after passing through any of the optical variable combiners / splitters 13.

[0038] For example, in the example of serial number 1 in the table of FIG. 5(a), since the signals from all the transponders 11 are connected to one path, this signal passes through either of the wavelength selective multiplexers / demultiplexers 141, 14Nf and is connected to the cross - connect port 17. In the example of serial number 13, for the two cross - connect ports 17 that transmit three optical signals, the optical signals passing through the paths of the wavelength selective multiplexers / demultiplexers 141 or 14Nf are connected, and for the two cross - connect ports 17 that transmit only one optical signal, the optical signals passing through any of the optical variable combiners / splitters 13 are connected.

[0039] Next, the case of K = 3 will be considered. In the case of the table in FIG. 5(b), since K is 3, Nf = Int(8 / (3 + 1)) = 2 holds, and the number of wavelength selective multiplexers / demultiplexers 14 is 2. As shown in FIG. 5(b), except for serial number 7, there is only one path to which more than three optical signals are connected, and in the case of serial number 7, more than three optical signals are connected to two paths. Therefore, according to TPA1 of the first embodiment, it can be seen that even when K = 3, the number of wavelength selective multiplexers / demultiplexers 14 may be at most 2.

[0040] Next, consider the case where K = 4. In the case of the table in Fig. 5(c), since K is 4, Nf = Int(8 / (4 + 1)) = 1 holds, and the number of wavelength selective multiplexers / demultiplexers 14 is 1. As shown in Fig. 5(c), in any case from through number 1 to 6, there is one path through which more than 4 signals are routed. Therefore, according to the TPA1 of the first embodiment, it can be seen that when K = 4, the number of wavelength selective multiplexers / demultiplexers 14 may be at most 1.

[0041] As described above, in the first embodiment, when routing of a plurality of optical signals is allowed in one path, the number of wavelength selective multiplexers / demultiplexers provided in the TPA1 can be further reduced. However, even when no ASE noise from other transponders is allowed in one path, it is possible to reduce the number of wavelength selective multiplexers / demultiplexers compared to a known configuration. That is, in such a case, in the first embodiment, Nf = Int(8 / (1 + 1)) = 4 holds. In a known M×N WSS, it is necessary to provide a WSS for each cross-connect port, so Nf = M = 8. From this, it can be seen that the first embodiment requires half the number of wavelength selective multiplexers / demultiplexers of a known configuration.

[0042] Furthermore, as is clear from Equation (2), the first embodiment is limited only by the number of transponders connected to the TPA1 and does not depend on the number of paths for outputting optical signals to the outside. In other words, for the TPA of the first embodiment, even if the number of output-side paths is increased, the number of WSSs remains unchanged, which is suitable for a network that requires a large number of output-side paths.

[0043] Next, each configuration shown in Fig. 6 will be described. (Optical variable combiner / splitter) FIG. 7 is a diagram for explaining the details of the optical variable combiner / splitter 13 shown in FIG. 6. The optical variable combiner / splitter 13 has N input ports 341 to 34N for inputting optical signals from the optical switch 12. The input ports 341 to 34N are respectively connected to Mach-Zehnder interferometers 351 to 35N which are optical variable combiner / splitter elements. Hereinafter, when it is not necessary to distinguish each of the Mach-Zehnder interferometers 351 to 35N, it is simply referred to as "Mach-Zehnder interferometer 35".

[0044] The Mach-Zehnder interferometer 35 has two arm waveguides 37 sandwiched between two 3dB directional couplers 31 and 32, and a phase shifter 33 loaded on at least one of them. One of the outputs of the Mach-Zehnder interferometer 35 is connected to the input on the side where the input waveguides 34N to 341 of the next-stage Mach-Zehnder interferometer 35 are not connected. Therefore, the optical signals from the input waveguides 341 to 34N propagate to the output waveguide 36 according to the phase state of the Mach-Zehnder interferometer 35.

[0045] Next, the operation of the Mach-Zehnder interferometer 35 as the optical variable combiner / splitter 13 will be described. The operation requirement of the optical variable combiner / splitter 13 is to select and combine K of the optical signals to be input to the input waveguides 341 to 34N and output them to the output waveguide 36. Since K is expected to be small, it is preferable to give an optical path length difference of about half the wavelength of the signal wavelength to the two arm waveguides 37 so that each Mach-Zehnder interferometer 35 is connected to the through port of the Mach-Zehnder interferometer in the non-driven state.

[0046] FIG. 8 is a diagram showing the output optical intensity to the cross-connect port when the phase of the phase shifter 33 is changed in the Mach-Zehnder interferometer 35 shown in FIG. 7. The horizontal axis in FIG. 8 represents the combination or branching ratio in the optical variable combiner / splitter 13, and the vertical axis represents the phase (rad) of the phase shifter 33. For example, when K = 2 and signals from the input waveguides 341 to 34N are combined, when combining the signals from the input port 342 and the input port 345, the Mach-Zehnder interferometer 352 may be connected in a 100% cross configuration, and the Mach-Zehnder interferometer 355 may be connected in a 50% cross configuration, and the respective phases may be set accordingly. Such a setting can be achieved by setting the phase shifter 33 to the phases indicated by r2 and r3 in FIG. 8. In this case, the signal loss from the output waveguides 36 from the input waveguides 342 and 345 is 3 dB.

[0047] As another example, when K = 3 and signals from the input waveguides 342, 345, and 346 are combined, when combining the signals from the input ports 342, 345, and 346, the Mach-Zehnder interferometer 352 is connected in a 100% cross configuration, the Mach-Zehnder interferometer 355 is connected in a 50% cross configuration, and the Mach-Zehnder interferometer 356 is connected in a 33% cross configuration. The setting of the phase shifter at this time is r1, r2, and r3 shown in FIG. 8. In this case, the signal loss from the output waveguides 36 from the input waveguides 342 and 345 is 4.77 dB.

[0048] (Wavelength Selective Combiner / Splitter) FIG. 9 is a diagram for explaining the details of the wavelength selective combiner / splitter 14 shown in FIG. 6, and is a functional diagram of a general wavelength selective combiner / splitter. The wavelength selective combiner / splitter 14 has N ports P7 (ports on the wavelength separation side), and in FIG. 9, the optical signals of each port P7 are indicated by optical signals 211 to 21N. The optical signal input from the port P7 experiences an arbitrarily settable filtering spectrum and is output to P8. Therefore, the wavelength selective combiner / splitter 14 suppresses components other than the wavelength of the optical signal as shown in the spectrum 23, and multiplexes the wavelengths and outputs them to the port P8 (common port).

[0049] (Matrix Switch) FIG. 10 is a diagram for explaining the details of the matrix switch 15 shown in FIG. 6. The matrix switch 15 has Nf ports P9 and M ports P 10 and. In FIG. 10, the path to which the WDM signal input from the port P9 is connected is determined by the 1×M switches 511 and 51Nf connected to the port P9. The 1×M switch 511 includes M element switches 511-1 to 511-M, and the 1×M switch 51Nf includes M element switches 51Nf-1 to 51Nf-M. In order to reduce the leakage of the signal to other paths, as shown in the legend L in FIG. 10, it is preferable that the element switches 511-1 etc. are configured such that two-stage Mach-Zehnder interferometers are connected in series.

[0050] The Nf×1 switches 521 to 52M select the optical signals output from the 1×M switch 511 or the 1×M switch 51Nf. In the example of FIG. 10, since Nf is 2, the Nf×1 switches 521 etc. are Mach-Zehnder interferometers. However, when Nf>2, the Mach-Zehnder interferometer may be in a tree shape or in a tree shape similar to the 1×M switches 511-1 to 511-Nf. Also, the Nf×1 switches 521 to 52M do not need to connect the Mach-Zehnder interferometers in multiple stages like the element switches 511-1 to 511-M. This is because the Nf×1 switches 521 to 52M do not require a high extinction ratio, and the extinction is performed by the 1×M switches 511 to 51Nf.

[0051] (Combining and branching ratio of the optical combiner / splitter) TPA1 shown in Fig. 6 has the port P6 of the optical variable combiner / splitter 13 connected to the optical combiner / splitter 16. At this time, the combination / splitting ratio in the optical combiner / splitter 16 is preferably set according to the losses of the path via the optical variable combiner / splitter 13 and the path via the wavelength selective multiplexer / demultiplexer 14. Such a setting will be described. The signal loss IL1 from the transponder 11 to the optical combiner / splitter 16 in the path passing through the optical variable combiner / splitter 13 is represented by the following formula (3). Also, the signal loss IL2 from the transponder 11 to the optical combiner / splitter 16 in the path passing through the wavelength selective multiplexer / demultiplexer 14 is represented by the following formula (4). In formula (3), α is the excess loss associated with the optical variable combiner / splitter 13 and propagation, β is the excess loss associated with the matrix switch and propagation, or the connection loss for optical propagation to the wavelength selective multiplexer / demultiplexer 14, etc. IL1 = -10×log(1 / K) + α [dB] ··· Formula (3) IL2 = IL WSS + β [dB] ··· Formula (4)

[0052] In TPA1, it is preferable that the signal loss IL1 and the signal loss IL2 be equal. Assuming the combination / splitting ratio in the optical combiner / splitter 16 is γ:1 - γ, the loss of each path expressed in decibels is IL1 = -10×log(1 / K) + α - 10×log(γ) IL2 = -IL WSS + β - 10×log(1 - γ) Therefore, -10×log(1 / K) + α - 10×log(γ) = -IL WSS + β - 10×log(1 - γ), and the γ of the combination / splitting ratio of the optical combiner / splitter 16 may be set so as to satisfy this.

[0053] In the first embodiment, the optical combiner / splitter 16 may be provided with a function to change the above γ, making it possible to change the combination / splitting ratio of the optical signal in the optical combiner / splitter 16. However, the first embodiment is not limited to such a configuration, and a preset fixed combination / splitting ratio may be set for the optical combiner / splitter 16 in Fig. 6.

[0054] Specific numerical examples are shown below. In the case of M = 8, N = 8, and K = 2 shown in the table of Fig. 5(a), as described above, the theoretical loss occurring in the optical variable combiner / splitter 13 is 3 dB. Therefore, if the loss of the optical switch 12 is ignored, the loss from the transponder 11 to the optical combiner / splitter 16 is 3 dB. On the other hand, since the typical loss of the wavelength selective multiplexer / demultiplexer 14 is about 6 dB, the loss from the transponder 11 to the optical combiner / splitter 16 is also 6 dB. Therefore, the branching ratio of the optical combiner / splitter 16 is set to 33.3:66.7 so as to compensate for this loss difference, and 66.7% of the signal components from the wavelength selective multiplexer / demultiplexer 14 may be set to converge to the cross-connect port 17. In this case, the signal loss from the transponder 11 to the cross-connect port 17 is 7.8 dB via any path. Here, it should be noted that the theoretical loss of the 8-way MCS is 9 dB, while the TPA of the same scale in the first embodiment can achieve a reduction in signal loss. That is, it is clear that the configuration of the first embodiment also has a remarkable effect in terms of signal loss.

[0055] [Second Embodiment] In the first embodiment, the optical combiner / splitter 16 and the optical variable combiner / splitter 13 are separately configured. The second embodiment provides a configuration (shown as the optical variable combiner / splitter 29 in Fig. 11) in which the optical combiner / splitter 16 and the optical variable combiner / splitter 13 are integrated and have their functions combined. Fig. 11 is a diagram for explaining the optical variable combiner / splitters 29-1 to 29-M (simply referred to as "optical variable combiner / splitter 29" when there is no need to distinguish each of them) of the second embodiment. The optical variable combiner / splitter 29 does not have the configuration of a known optical splitter, but has a configuration similar to a Mach-Zehnder interferometer and operates as an optical switch. The optical variable combiner / splitter 29 connects one of the input ports of the optical combiner / splitter 16 to the output waveguide 36 of the optical variable combiner / splitter 13 shown in Fig. 7, and the other to the port P of the matrix switch 10An optical signal heading for a certain direction passes through either a path passing through the variable optical multiplexer / splitter 13 or a path passing through the wavelength selective multiplexer / demultiplexer 14, so the variable optical multiplexer / splitter 29 can uniquely select either path. As in the first embodiment, the second embodiment can reduce the total signal loss from the transponder 11 to the cross connect port 17 to be smaller than that of the known MCS, regardless of whether the path passes through the variable optical multiplexer / splitter 13 or the wavelength selective multiplexer / demultiplexer 14.

[0056] [Third embodiment] The first embodiment shows an example in which the optical switch 12, the variable optical multiplexer / splitter 13, the wavelength selective multiplexer / splitter 14, the matrix switch 15 and the optical multiplexer / splitter 16 are all formed on one optical waveguide substrate. The TPA 2 of the third embodiment differs from the first embodiment in that the optical switch 12, the variable optical multiplexer / splitter 29 of the second embodiment in which the variable optical multiplexer / splitter 13 and the optical multiplexer / splitter 16 are integrated, and the matrix switch 105 are formed on the optical waveguide substrate 6, and the wavelength selective multiplexer / splitter 14 is configured as a separate module. FIG. 12 is a diagram for explaining such a TPA 2 of the third embodiment. In the TPA 2, the ports and waveguides on the optical waveguide substrate 6 are connected to the ports of the wavelength selective multiplexer / splitter 14 by optical fibers 61.

[0057] In the third embodiment, the variable optical multiplexer / splitter 29 outputs all input optical signals at the same power to the output port P 20 When considering the case where the optical waveguide substrate 6 is used for routing to the above-mentioned range, it is understood that the optical waveguide substrate 6 has the same function as a normal multicast switch. That is, the optical waveguide substrate 6 can be regarded as an MCS configured with an expansion port to which a wavelength selective multiplexer / demultiplexer 14 can be connected to a normal MCS. Therefore, in an optical network with strict OSNR requirements, it is possible to connect a wavelength selective multiplexer / demultiplexer 14 to the optical waveguide substrate 6 of the third embodiment to form a TPA 2 with a built-in filter function. Also, in an optical network with a margin in the OSNR requirements, the third embodiment provides the freedom to use the optical waveguide substrate 6 as a normal MCS.

[0058] [Fourth Embodiment] Furthermore, in one embodiment of the present invention, an optical amplifier may be inserted into the path of the wavelength selective multiplexer / demultiplexer 14 to compensate for signal loss. The fourth embodiment is different from the first and third embodiments in that it includes an optical amplifier. In the fourth embodiment, TPA3 is formed on one optical waveguide substrate. FIG. 13 is a diagram for explaining TPA3 of the fourth embodiment. The TPA3 shown in FIG. 13 is configured by providing optical amplifiers 711 and 71Nf to the TPA2 of the third embodiment. As the optical amplifiers 71 and 7Nf, an EDFA is suitable, but a semiconductor optical amplifier (SOA) may also be used. Further, when there is a margin in the OSNR requirement, the wavelength selective multiplexer / demultiplexer 14 may be removed and only an optical amplifier may be installed.

[0059] The optical variable combiner / splitter 29 sets the combination / splitting ratio of the combiner / splitter composed of a Mach-Zehnder interferometer so that most of the optical power of the included optical variable combiner / splitter is routed to the cross-connect port 17. In this case, here, the signal loss of the path on the side of the wavelength selective multiplexer / demultiplexer 14 where most is not routed to the cross-connect port 17 increases, but the loss can be compensated from 711 to 71Nf by an optical amplifier. Next, a specific numerical example will be used for explanation. For example, when M = 8, N = 8, and K = 2, if the combination / splitting ratio of the combiner / splitter is 9:1 and 90% of the optical power is routed from the optical variable combiner / splitter 29 to the cross-connect port 17. In this case, the theoretical loss IL0 of the path passing through the optical variable combiner / splitter 13 from the transponder 11 to the cross-connect port 17 is as follows. IL0=-10×log(1 / 2)-10×log(9 / 10)=3.4dB

[0060] On the other hand, the signal loss IL2 of the path passing through the wavelength selective multiplexer / demultiplexer 14 is expressed as follows, assuming the loss IL WSS of the WSS is 6dB. IL2=IL WSS -10×log(1 / 10)=16dB Therefore, according to the fourth embodiment, by inserting the optical amplifiers 711 and 71Nf with a gain of 12.6 dB, it becomes possible to reduce the signal loss to 3.4 dB regardless of the path. Since the signal loss of the MCS with M = 8 and N = 8 of the same scale is 9 dB, it can be said that the fourth embodiment has a great effect on reducing the signal loss. Furthermore, as described above, the number of optical filters to be installed was required to be the number of paths in the MCS with 8 outputs, but in the first to fourth embodiments, the maximum number of optical filters is only 4, which greatly contributes to cost reduction.

[0061] As described above, the first to fourth embodiments of the present invention solve the problems of known transponder aggregation devices. That is, compared with the MCS type transponder aggregation device, the first to fourth embodiments provide an inexpensive filtering function and reduce the principle loss. Also, compared with the M×N WSS type transponder aggregation device, it has the effect of reducing the number of wavelength selective multiplexers to be installed.

Explanation of Signs

[0062] 1, 2, 3 TPA 6 Optical waveguide substrate 11 Transponder 12 Optical switch 13, 29 Optical variable combiner / splitter 14 Wavelength selective multiplexer 15 Matrix switch 16 Optical combiner / splitter 17 Cross-connect port 31, 32 Directional coupler 33 Phase shifter 37 Arm waveguide 34 Input waveguide 35 Mach-Zehnder interferometer 36 Output waveguide 61 Optical fiber 711, 71Nf Optical amplifier

Claims

1. An optical signal processing apparatus having N first ports and M second ports, comprising: N optical switches each having a third port connected to the first port and M + Nf fourth ports not connected to the first port; M optical variable combiners / splitters each having a fifth port connected to any one of the first to M-th of the M + Nf fourth ports and a sixth port not connected to the fourth port; A matrix switch having Nf first extension ports connected to any one of the (M + 1)-th to (M + Nf)-th of the M + Nf fourth ports and a second extension port not connected to the fourth port; A optical combiner / splitter connected to the sixth port, the second port, and the second extension port; The optical signal processing apparatus, wherein Nf is a value obtained by dividing N by K + 1 and rounding down the remainder, where K is the number of optical signals connectable to one of the second ports.

2. The optical signal processing apparatus according to claim 1, wherein the optical variable combiner / splitter selects K signals from the signals connected to the fifth port and connects them to the sixth port.

3. The optical signal processing apparatus according to claim 1 or 2, wherein part of the first port, the optical switch, the optical variable combiner / splitter, the optical combiner / splitter, the wavelength selective multiplexer / demultiplexer, and the matrix switch are provided on a common waveguide substrate, and the other part is configured as a separate module.

4. The optical signal processing apparatus according to any one of claims 1 to 3, wherein the first extension port of the matrix switch is connected to the fourth port via at least one of a wavelength selective multiplexer / demultiplexer and an optical amplifier.

5. The optical signal processing apparatus according to any one of claims 1 to 4, wherein the optical combiner / splitter has a function of changing at least one of the branching ratio and the combining ratio of the optical signals.

Images