Optical cross-connect using wavelength selective switches

This application is based upon and claims the benefits of priority from the prior Japanese Patent Application No. 2006-330120 filed Dec. 7, 2006, the entire contents of which are incorporated herein by reference.

1. Field of the Invention

The present invention relates to an optical cross-connect, and more particularly to an optical cross-connect that switches optical signals of N routes (N=1, 2, . . . ).

2. Description of the Related Art

Many optical network systems use the wavelength-division multiplexing (WDM) technology to interconnect network nodes such as cities and countries. WDM transmission systems involve various control techniques to, for example, deliver optical signals with different wavelengths to intended destinations, or to add or drop a desired wavelength to/from a desired route of optical signals. The latter is known as “optical add and drop multiplexing” (OADM). Also required is the feature of automatic protection switching, which enables a failed channel to be switched instantly to an alternate channel with a then-available wavelength in the event of a failure between network nodes. Optical cross-connect is a device that offers such optical switching functions, and there is an increased interest in wavelength selective switches (WSS) as a key component to realize optical cross-connects.

FIG. 33 shows the basic structure of a 1-input 3-output (1×3) WSS. The illustrated WSS 40 is formed from a spectroscopic system 41 , a lens 42 , and a mirror 43 . The incoming light through the input port IN is reflected to one of the three output ports OUT 1 to OUT 3 . A desired output port can be selected by varying the tilt angle of the mirror 43 .

FIG. 34 shows a structure of a conventional optical cross-connect. The illustrated optical cross-connect 50 is a wavelength cross-connect (WXC) designed to switch four routes each carrying 40 WDM channels, with fixed add wavelengths. The optical cross-connect 50 includes 1×4 WSSs 51 - 1 to 51 - 4 , 4×1 WSSs 52 - 1 to 52 - 4 , combiners 53 - 1 to 53 - 4 , splitters 54 - 1 to 54 - 4 , WDM amplifiers 55 - 1 to 55 - 4 and 56 - 1 to 56 - 4 , transmitters Tx, and receivers Rx.

Input routes Rin# 1 to Rin# 4 reach the input ports of corresponding 1×4 WSSs 51 - 1 to 51 - 4 through WDM amplifiers 55 - 1 to 55 - 4 . The outputs of 4×1 WSSs 52 - 1 to 52 - 4 are directed to output routes Rout# 1 to Rout# 4 through WDM amplifiers 56 - 1 to 56 - 4 . Signals on each add route A# 1 to A# 4 are entered to input ports of a corresponding combiner 53 - 1 to 53 - 4 through 40 transmitters Tx. Each drop route D# 1 to D# 4 has 40 receivers Rx coupled to the outputs of a corresponding splitter 54 - 1 to 54 - 4 .

Input and output ports of WSSs are named as shown in the legend of FIG. 34 , the order being determined by the structure of a WSS. Such ports of WSSs are connected as follows: Output port OUT 1 of the first 1×4 WSS 51 - 1 is connected to input port IN 1 of the second 4×1 WSS 52 - 2 . Output port OUT 2 of the WSS 51 - 1 is connected to input port IN 1 of the third 4×1 WSS 52 - 3 . Output port OUT 3 of the WSS 51 - 1 is connected to input port IN 1 of the fourth 4×1 WSS 52 - 4 . Output port OUT 4 of the WSS 51 - 1 is connected to input port of the first splitter 54 - 1 .

Output port OUT 1 of the second 1×4 WSS 51 - 2 is connected to input port IN 1 of the first 4×1 WSS 52 - 1 . Output port OUT 2 of the WSS 51 - 2 is connected to input port IN 2 of the third 4×1 WSS 52 - 3 . Output port OUT 3 of the WSS 51 - 2 is connected to input port IN 2 of the fourth 4×1 WSS 52 - 4 . Output port OUT 4 of the WSS 51 - 2 is connected to input port of the second splitter 54 - 2 .

Output port OUT 1 of the third 1×4 WSS 51 - 3 is connected to input port IN 2 of the first 4×1 WSS 52 - 1 . Output port OUT 2 of the WSS 51 - 3 is connected to input port IN 2 of the second 4×1 WSS 52 - 2 . Output port OUT 3 of the WSS 51 - 3 is connected to input port IN 3 of the fourth 4×1 WSS 52 - 4 . Output port OUT 4 of the WSS 51 - 3 is connected to input port of the third splitter 54 - 3 .

Output port OUT 1 of the fourth 1×4 WSS 51 - 4 is connected to input port IN 3 of the first 4×1 WSS 52 - 1 . Output port OUT 2 of the WSS 51 - 4 is connected to input port IN 3 of the second 4×1 WSS 52 - 2 . Output port OUT 3 of the WSS 51 - 4 is connected to input port IN 3 of the third 4×1 WSS 52 - 3 . Output port OUT 4 of the WSS 51 - 4 is connected to input port of the fourth splitter 54 - 4 .

The output of the first combiner 53 - 1 is connected to input port IN 4 of the first 4×1 WSS 52 - 1 . The output port of the second combiner 53 - 2 is connected to input port IN 4 of the second 4×1 WSS 52 - 2 . Likewise, the output of the combiner 53 - 3 is connected to input port IN 4 of the third 4×1 WSS 52 - 3 , and the output of the combiner 53 - 4 is connected to input port IN 4 of the fourth 4×1 WSS 52 - 4 .

This conventional optical cross-connect 50 is unable to change the wavelengths of individual add signals since it uses fixed-wavelength transmitters Tx. The optical cross-connect 50 also has to use expensive WDM amplifiers 55 - 1 to 55 - 4 and 56 - 1 to 56 - 4 each having a dispersion compensation fiber (DCF) to compensate for the large optical loss of combiners 53 - 1 to 53 - 4 and splitters 54 - 1 to 54 - 4 .

FIG. 35 shows the structure of another conventional optical cross-connect, and FIG. 36 shows four types of WSSs and their input and output port numbers as a legend for WSSs shown in FIG. 35 . This optical cross-connect 60 can add optical signals with desired wavelengths, unlike the optical cross-connect 50 of FIG. 34 .

The optical cross-connect 60 has four 1×7 WSSs 61 - 1 to 61 - 4 on the input side to receive incoming signals from input routes Rin# 1 to Rin# 4 . Likewise, it has four 1×4 WSSs 66 - 1 to 66 - 4 to receive add signals from add routes A# 1 to A# 4 . On the output side, four 7×1 WSSs 62 - 1 to 62 - 4 are employed to send signals to output routes Rout# 1 to Rout# 4 . For drop routes D# 1 to D# 4 , four 4×1 WSSs 67 - 1 to 67 - 4 are placed.

Those WSSs are connected with each other in the following way: Output port OUT 1 of the first 1×7 WSS 61 - 1 is connected to input port IN 1 of the second 7×1 WSS 62 - 2 . Output port OUT 2 of the WSS 61 - 1 is connected to input port IN 1 of the third 7×1 WSS 62 - 3 . Output port OUT 3 of the WSS 61 - 1 is connected to input port IN 1 of the fourth 7×1 WSS 62 - 4 . Output port OUT 4 of the WSS 61 - 1 is connected to input port IN 1 of the first 4×1 WSS 67 - 1 . Output port OUT 5 of the WSS 61 - 1 is connected to input port IN 1 of the second 4×1 WSS 67 - 2 . Output port OUT 6 of the WSS 61 - 1 is connected to input port IN 1 of the third 4×1 WSS 67 - 3 . Output port OUT 7 of the WSS 61 - 1 is connected to input port IN 1 of the fourth 4×1 WSS 67 - 4 .

Output port OUT 1 of the second 1×7 WSS 61 - 2 is connected to input port IN 1 of the first 7×1 WSS 62 - 1 . Output port OUT 2 of the WSS 61 - 2 is connected to input port IN 2 of the third 7×1 WSS 62 - 3 . Output port OUT 3 of the WSS 61 - 2 is connected to input port IN 2 of the fourth 7×1 WSS 62 - 4 . Output port OUT 4 of the WSS 61 - 2 is connected to input port IN 2 of the first 4×1 WSS 67 - 1 . Output port OUT 5 of the WSS 61 - 2 is connected to input port IN 2 of the second 4×1 WSS 67 - 2 . Output port OUT 6 of the WSS 61 - 2 is connected to input port IN 2 of the third 4×1 WSS 67 - 3 . Output port OUT 7 of the WSS 61 - 2 is connected to input port IN 2 of the fourth 4×1 WSS 67 - 4 .

Output port OUT 1 of the third 1×7 WSS 61 - 3 is connected to input port IN 2 of the first 7×1 WSS 62 - 1 . Output port OUT 2 of the WSS 61 - 3 is connected to input port IN 2 of the third 7×1 WSS 62 - 3 . Output port OUT 3 of the WSS 61 - 3 is connected to input port IN 3 of the fourth 7×1 WSS 62 - 4 . Output port OUT 4 of the WSS 61 - 3 is connected to input port IN 3 of the first 4×1 WSS 67 - 1 . Output port OUT 5 of the WSS 61 - 3 is connected to input port IN 3 of the second 4×1 WSS 67 - 2 . Output port OUT 6 of the WSS 61 - 3 is connected to input port IN 3 of the third 4×1 WSS 67 - 3 . Output port OUT 7 of the WSS 61 - 3 is connected to input port IN 3 of the fourth 4×1 WSS 67 - 4 .

Output port OUT 1 of the fourth 1×7 WSS 61 - 4 is connected to input port IN 3 of the first 7×1 WSS 62 - 1 . Output port OUT 2 of the WSS 61 - 4 is connected to input port IN 3 of the second 7×1 WSS 62 - 2 . Output port OUT 3 of the WSS 61 - 4 is connected to input port IN 3 of the third 7×1 WSS 62 - 3 . Output port OUT 4 of the WSS 61 - 4 is connected to input port IN 4 of the first 4×1 WSS 67 - 1 . Output port OUT 5 of the WSS 61 - 4 is connected to input port IN 4 of the second 4×1 WSS 67 - 2 . Output port OUT 6 of the WSS 61 - 4 is connected to input port IN 4 of the third 4×1 WSS 67 - 3 . Output port OUT 7 of the WSS 61 - 4 is connected to input port IN 4 of the fourth 4×1 WSS 67 - 4 .

Output port OUT 1 of the first 1×4 WSS 66 - 1 is connected to input port IN 4 of the first 7×1 WSS 62 - 1 . Output port OUT 2 of the WSS 66 - 1 is connected to input port IN 4 of the second 7×1 WSS 62 - 2 . Output port OUT 3 of the WSS 66 - 1 is connected to input port IN 4 of the third 7×1 WSS 62 - 3 . Output port OUT 4 of the WSS 66 - 1 is connected to input port IN 4 of the fourth 7×1 WSS 62 - 4 .

Output port OUT 1 of the second 1×4 WSS 66 - 2 is connected to input port IN 5 of the first 7×1 WSS 62 - 1 . Output port OUT 2 of the WSS 66 - 2 is connected to input port IN 5 of the second 7×1 WSS 62 - 2 . Output port OUT 3 of the WSS 66 - 2 is connected to input port IN 5 of the third 7×1 WSS 62 - 3 . Output port OUT 4 of the WSS 66 - 2 is connected to input port IN 5 of the fourth 7×1 WSS 62 - 4 .

Output port OUT 1 of the third 1×4 WSS 66 - 3 is connected to input port IN 6 of the first 7×1 WSS 62 - 1 . Output port OUT 2 of the WSS 66 - 3 is connected to input port IN 6 of the second 7×1 WSS 62 - 2 . Output port OUT 3 of the WSS 66 - 3 is connected to input port IN 6 of the third 7×1 WSS 62 - 3 . Output port OUT 4 of the WSS 66 - 3 is connected to input port IN 6 of the fourth 7×1 WSS 62 - 4 .

Output port OUT 1 of the fourth 1×4 WSS 66 - 4 is connected to input port IN 7 of the first 7×1 WSS 62 - 1 . Output port OUT 2 of the WSS 66 - 4 is connected to input port IN 7 of the second 7×1 WSS 62 - 2 . Output port OUT 3 of the WSS 66 - 4 is connected to input port IN 7 of the third 7×1 WSS 62 - 3 . Output port OUT 4 of the WSS 66 - 4 is connected to input port IN 7 of the fourth 7×1 WSS 62 - 4 .

For the first add route A# 1 , the optical cross-connect 60 has forty wavelength-tunable transmitters Txv- 1 to Txv- 40 , five 8×1 WSSs 63 a - 1 to 63 a - 5 , a 5×1 WSS 64 - 1 , and a WDM amplifier 65 - 1 . The forty transmitters Txv- 1 to Txv- 40 transmit forty optical signals, respectively. Each WSS 63 a - 1 to 63 a - 5 receives eight out of forty optical signals from the transmitters Txv- 1 to Txv- 40 and selectively outputs one optical signal to the 5×1 WSS 64 - 1 . The 5×1 WSS 64 - 1 then chooses one optical signal from among five optical signals received from the five 8×1 WSSs 63 a - 1 to 63 a - 5 . The WDM amplifier 65 - 1 amplifies the output of this WSS 64 - 1 before feeding it as an add signal to the first 1×4 WSS 66 - 1 . Other three add routes A# 2 to A# 4 have the same hardware structure as the first add route A# 1 described above.

For the first drop route D# 1 , the optical cross-connect 60 has a WDM amplifier 68 - 1 , a 1×5 WSS 69 - 1 , and five 1×8 WSS 70 a - 1 to 70 a - 5 , and forty receivers Rx- 1 to Rx- 40 . The WDM amplifier 68 - 1 amplifies a drop signal produced by the first 4×1 WSS 67 - 1 . Five outputs of the WSS 69 - 1 are connected to five input ports of different 1×8 WSSs 70 a - 1 to 70 a - 5 to route the drop signal to any one of their forty output ports as specified. Eight outputs of the first 1×8 WSS 70 a - 1 are directed to a group of receivers Rx- 1 to Rx- 8 , as are the outputs of the second 1×8 WSS 70 a - 2 to a subsequent group of receivers Rx- 9 to Rx- 16 . Likewise, other three 1×8 WSSs 70 a - 3 to 70 a - 5 provide their outputs to corresponding receiver groups Rx- 17 to Rx- 24 , Rx- 25 to Rx- 32 , and Rx- 33 to Rx- 40 . Note that only one of eight signals of each group carries active drop signal information. Each receiver Rx- 1 to Rx- 40 receives an optical drop signal (if present). Other three drop routes D# 2 to D# 4 have the same hardware structure as the first drop route D# 1 .

The number of WSS ports on the input and output routes may be increased or decreased depending on how many routes are to be switched. In the case where the network system needs more wavelengths in WDM transmission, the WSSs on add and drop routes have to be replaced to provide more ports.

The above conventional optical cross-connect 60 realizes the selective add-wavelength WXC functions by combining 1×7/7×1 WSSs, 1×4/4×1 WSSs, 1×8/8×1 WSSs, and 1×5/5×1 WSSs, but without using optical combiners or splitters. Like the one shown in FIG. 34 , the optical cross-connect 60 of FIG. 35 can switch cross-connections between four routes each having 40 WDM channels. For example, the WSSs 63 a - 1 to 63 a - 5 and 64 - 1 on the first add route A# 1 perform fast switching to add 40-ch WDM signals to a desired output route through the WDM amplifier 65 - 1 and 1×4 WSS 66 - 1 .

As an example of such conventional optical cross-connecting techniques, Japanese Patent Application Publication No. 2004-343231, paragraphs 0017 to 0019, FIG. 1, proposes an optical cross-connect having input and output ports for individual channels to enable interface with existing network facilities.

The above-described architectures of conventional optical cross-connects may not be optimal in terms of how they use WSSs. Specifically, the optical cross-connect 60 of FIG. 35 includes many WSSs in its add/drop sections. This design is advantageous over the optical cross-connect 50 of FIG. 34 in its smaller signal loss since no optical combiners or splitters are required. It also eliminates the need for costly WDM amplifiers with DCF. On the other hand, the increased use of WSSs also means increased product cost, as well as an increased number of interconnections between them, thus making a cross-connect device more complex and larger.

The conventional optical cross-connect 60 provides various switching patterns for optical cross-connection, and many of them can work simultaneously. In reality, however, the simultaneity of cross-connections is not always required in real-world applications. FIGS. 37 and 38 show tables T 61 and T 62 representing the simultaneous feasibility of optical cross-connections that the optical cross-connect 60 of FIG. 35 can or cannot provide. Specifically, table T 61 summarizes simultaneous feasibility of optical cross-connections between input route Rin# 1 and add route A# 1 . Table T 62 , on the other hand, summarizes the same for cross-connections between input route Rin# 1 and drop route D# 1 .

Table entries having a value of “YES” mean that the corresponding switching paths can work simultaneously. The top-left entry of table T 61 , for example, is marked “YES,” indicating that an add switching path from add route A# 1 to output route Rout# 1 can work together with a through switching path from input route Rin# 1 to output route Rout# 2 .

Table entries having a value of “NO” indicate that the corresponding combination of switching paths cannot work simultaneously. Reversely stated, one of those paths can work selectively. See the entry at column 2, row 1 of table T 61 , for example. This entry is marked “NO,” meaning that an add switching path from add route A# 1 to output route Rout# 2 cannot work together with a through switching path from input route Rin# 1 to output route Rout# 2 . This is because these two paths collide at an output port of the second 7×1 WSS 62 - 2 . The other table T 62 is supposed to be interpreted in the same way.

FIG. 39 shows yet another table representing simultaneous feasibility of optical cross-connections. This table T 63 is a logical product of table T 61 and table T 62 , which is produced by performing a logical operation on each corresponding pair of entries of those two tables. Specifically, let x and y represent respectively the values of table T 61 and table T 62 at a particular column and row. The logical product of two table entries (x, y) takes a value of “YES” if (x,y)=(YES,YES), or a value of “NO” if (x,y)=(YES,NO) or (NO,YES) or (NO,NO). The resulting value is then set to a corresponding entry of table T 63 .

Referring to table T 63 , the entries having a value of “YES” indicate that the corresponding through switching path and add/drop switching paths can work simultaneously. See the entry at column 3, row 1, for example. This entry has a value of “YES,” meaning that a through switching path from input route Rin# 1 to output route Rout# 2 , an add switching path from add route A# 1 to output route Rout# 3 , and a drop switching path from input route Rin# 3 to drop route D# 1 can all work at the same time.

Other entries of table T 63 have a value of “NO” to indicate that either the corresponding through switching path or the corresponding add/drop switching paths can work selectively. See the top-left entry of table T 63 , for example. This table entry indicates that the optical cross-connect 60 does not allow simultaneous use of a through switching path from input route Rin# 1 to output route Rout# 2 , an add switching path from add route A# 1 to output route Rout# 1 , and a drop switching path from input route Rin# 1 to drop route D# 1 . That table entry, however, implies that either the through switching path alone or the add/drop switching paths can be implemented.

In reality, however, it may not be necessary to make available every combination of a through path and add/drop paths for simultaneous use. Rather, most cases require either the former or the latter alone. Table T 63 , however, shows many switching patterns marked “YES.” This means that the conventional optical cross-connect 60 is not optimized as to the simultaneity of its switching paths. The entire circuit of an optical cross-connect will be smaller in size if it is possible to reduce some simultaneous switching patterns of through and add/drop paths (which means reducing “YES” and increasing “NO” in table T 63 ).

In view of the foregoing, it is an object of the present invention to provide an optimally-sized optical cross-connect that switches optical signal paths with a minimum number of wavelength selective switches and their interconnections.

To accomplish the above object, the present invention provides an optical cross-connect switching optical signals of N routes (N=1, 2, . . . ). This optical cross-connect has 2-input N-output (2×N) wavelength selective switches coupled respectively to N input routes, and N-input 2-output (N×2) wavelength selective switches coupled respectively to N output routes. One input port of each 2×N wavelength selective switch receives an optical signal from a corresponding input route, while the other input port receives an optical add signal from a corresponding add route. One output port of each N×2 wavelength selective switch outputs an optical signal to a corresponding output route, while the other output port outputs an optical drop signal to a corresponding drop route.

The above and other objects, features and advantages of the present invention will become apparent from the following description when taken in conjunction with the accompanying drawings which illustrate preferred embodiments of the present invention by way of example.

FIG. 1 is a conceptual view of an optical cross-connect according to the present invention.

FIG. 2 shows a basic structure of 2×N WSS.

FIG. 3 shows a basic structure of N×2 WSS.

FIGS. 4 to 8 show various cases where 2×N WSS cannot switch signals properly.

FIG. 9 shows a structure of an optical cross-connect according to an embodiment of the present invention.

FIGS. 10 to 17 show various switching paths that an optical cross-connect can produce.

FIGS. 18 to 20 show tables representing simultaneous feasibility of optical cross-connections.

FIG. 21 shows a variant of the foregoing optical cross-connect according to the present invention.

FIGS. 22 to 29 show various switching paths that an optical cross-connect can produce.

FIGS. 30 to 32 show tables representing simultaneous feasibility of optical cross-connections.

FIG. 33 shows a basic structure of 1×N WSS.

FIG. 34 shows a structure of a conventional optical cross-connect.

FIG. 35 shows a structure of another conventional optical cross-connect.

FIG. 36 shows four types of WSSs and their input and output port numbers as a legend for WSSs shown in FIG. 35 .

FIGS. 37 to 39 show tables representing simultaneous feasibility of optical cross-connections.

Preferred embodiments of the present invention will be described below with reference to the accompanying drawings, wherein like reference numerals refer to like elements throughout.

FIG. 1 is a conceptual view of an optical cross-connect according to the present invention. This optical cross-connect 1 switches optical signals between N routes, which include input routes, N output routes, N add routes, and N drop routes (N=1, 2, . . . ). To perform this function, the optical cross-connect 1 includes (a) 2-input N-output (2×N) wavelength selective switches (WSS) 11 - 1 to 11 -N coupled respectively to N input routes, and (b) N-input 2-output (N×2) wavelength selective switches 12 - 1 to 12 -N coupled respectively to N output routes.

The nth 2×N wavelength selective switch (shown as 2×N WSS#n in FIG. 1 ) (n=1, 2, . . . N) coupled to the nth input route (input route #n) receives an optical signal from the nth input route at its first input port and an optical add signal from the nth add route at its second input port.

The nth N×2 wavelength selective switch (shown as N×2 WSS#n in FIG. 1 ) coupled to the nth output route (output route #n) outputs an optical signal from its first output port to the nth output route and an optical drop signal from its second output port to the nth drop route.

Wavelength Selective Switch

Referring now to FIGS. 2 and 3 , this section will describe the basic structure of 2×N and N×2 WSSs used in the proposed optical cross-connect 1 . Specifically, FIG. 2 shows a 2×N WSS 11 - 1 , and FIG. 3 shows an N×2 WSS 12 - 1 . Both WSSs 11 - 1 and 12 - 1 have a spectroscopic system 11 a , a lens 11 b , and a mirror 11 c . They select specified input-output pairs by varying the angle of the mirror 11 c.

Optical signals entered to two adjacent input ports of a WSS come out of two adjacent output ports at symmetric positions to the respective input ports. This is an inherent nature of WSS structure. Accordingly, if the mirror 11 c is adjusted so as to direct one input signal to a particular output port, then it automatically determines at which output port an adjacent input signal will appear. The 2×N WSS 11 - 1 , for example, has two input ports IN 1 and IN 2 . Suppose that the 2×N WSS 11 - 1 directs a signal from the outermost input port IN 1 to the outermost output port OUT 1 . This automatically means that a signal entered to the next input port IN 2 goes to an inner output port OUT 2 . (Other output ports OUT 3 and OUT 4 never output this signal.) For another example, suppose that the 2×N WSS 11 - 1 is adjusted to direct a signal from the outermost input port IN 1 to the second output port OUT 2 . In this case, a signal entered to the other input port IN 2 goes to the third output port OUT 3 . (Other output ports OUT 1 and OUT 4 never output this signal.)

The N×2 WSS 12 - 1 of FIG. 3 operates in a similar way. Suppose, for example, that it is configured to direct a signal from the outermost input port IN 1 to the outermost output port OUT 1 . In this case, the next output port OUT 2 should be dedicated to no other than the second input port IN 2 (i.e., other input ports IN 3 and IN 4 cannot be assigned to the output port OUT 2 ). Likewise, if the N×2 WSS 12 - 1 is configured to direct a signal from the second input port IN 2 to the outermost output port OUT 1 , then the second output port OUT 2 should be dedicated to no other than the third input port IN 3 (i.e., other input ports IN 1 and IN 4 cannot be assigned to the output port OUT 2 ).

As can be seen from the above examples, a WSS reflects optical signals from two adjacent input ports to two adjacent output ports in a particular way. That is, one input port and its corresponding output port are symmetric to each other with respect to a line defined by the tilt of the mirror 11 c , as are the other input port and its corresponding output port with respect to the same line of symmetry. Because of this nature of WSS, several patterns of optical switching cannot be put into practice by the use of a WSS. FIGS. 4 to 8 illustrate such infeasible switching patterns, assuming the use of a 2×N WSS.

Specifically, FIG. 4 shows an infeasible switching pattern where the output ports are not immediately adjacent with each other. FIG. 5 shows another infeasible switching pattern where input ports are mapped not to different output ports, but to a single output port. Because of the symmetry of input ports and corresponding output ports explained earlier, neither the switching pattern of FIG. 4 nor that of FIG. 5 can be realized.

Other infeasible switching patterns are when: two optical paths are twisted ( FIG. 6 ); no output port exists at a symmetric position to the input port ( FIG. 7 ); a single input is mapped to multiple outputs ( FIG. 8 ).

More specifically, FIG. 6 shows a case where one input signal entered to input port IN 1 appears at output port OUT 2 while the other input signal entered to input port IN 2 appears at output OUT 1 . This switching pattern is not feasible since the two rays can never be reflected by the same mirror 11 c in that way. (The switching would work fine if it was output port OUT 3 where the signal entering IN 2 comes out.) As this example shows, the WSS does not allow two internal optical paths to intersect with each other.

FIG. 7 shows a case where one signal entering the second input port IN 2 comes out of the outermost output port OUT 1 . In this case another signal entered to input port IN 1 goes to where no output ports are present, hence another infeasible switching pattern. Shown in FIG. 8 is another infeasible pattern, where the WSS cannot distribute one input to multiple output ports because of its inherent structure.

Optical Cross-Connect

This section describes a more specific embodiment of the optical cross-connect 1 according to the present invention. FIG. 9 shows an optical cross-connect 1 a according to an embodiment of the present invention. This optical cross-connect 1 a performs switching between four routes each carrying 40 WDM channels. Four 2×4 WSSs 11 - 1 to 11 - 4 are coupled to input routes Rin# 1 to Rin# 4 , and four 4×2 WSSs 12 - 1 to 12 - 4 are coupled to output routes Rout# 1 to Rout# 4 .

Output ports OUT 1 to OUT 3 of four 2×4 WSSs 11 - 1 to 11 - 4 are connected to input ports of four 4×2 WSSs 12 - 1 to 12 - 4 . Specifically, the connection between output and input ports are as follows:

Output port OUT 1 of the first 2×4 WSS 11 - 1 is connected to input port IN 1 of the second 4×2 WSS 12 - 2 . Output port OUT 2 of the WSS 11 - 1 is connected to input port IN 1 of the third 4×2 WSS 12 - 3 . Output port OUT 3 of the WSS 11 - 1 is connected to input port IN 1 of the fourth 4×2 WSS 12 - 4 .

Output port OUT 1 of the second 2×4 WSS 11 - 2 is connected to input port IN 1 of the first 4×2 WSS 12 - 1 . Output port OUT 2 of WSS 11 - 2 is connected to input port IN 2 of third 4×2 WSS 12 - 3 . Output port OUT 3 of the WSS 11 - 2 is connected to input port IN 2 of the fourth 4×2 WSS 12 - 4 .

Output port OUT 1 of the third 2×4 WSS 11 - 3 is connected to input port IN 2 of the first 4×2 WSS 12 - 1 . Output port OUT 2 of the WSS 11 - 3 is connected to input port IN 2 of the second 4×2 WSS 12 - 2 . Output port OUT 3 of the WSS 11 - 3 is connected to input port IN 3 of the fourth 4×2 WSS 12 - 4 .

Output port OUT 1 of the fourth 2×4 WSS 11 - 4 is connected to input port IN 3 of the first 4×2 WSS 12 - 1 . Output port OUT 2 of the WSS 11 - 4 is connected to input port IN 3 of the second 4×2 WSS 12 - 2 . Output port OUT 3 of the WSS 11 - 4 is connected to input port IN 3 of the third 4×2 WSS 12 - 3 .

In the add section shown in the lower-left quarter of FIG. 9 , the optical cross-connect 1 a has the following elements to select an add signal on the first add route A# 1 : forty wavelength-tunable transmitters Txv- 1 to Txv- 40 , five 8×1 WSSs 13 a - 1 to 13 a - 5 , a 5×1 WSS 14 - 1 , an optical amplifier (WDM amplifier) 15 - 1 , and a 1×2 coupler C 1 - 1 . The forty transmitters Txv- 1 to Txv- 40 transmit forty optical signals, respectively. Each 8×1 WSS 13 a - 1 to 13 a - 5 receives eight out of forty optical signals from the transmitters Txv- 1 to Txv- 40 and outputs one selected optical signal to the subsequent 5×1 WSS 14 - 1 . The 5×1 WSS 14 - 1 then chooses one optical signal from among five optical signals received from the five 8×1 WSSs 13 a - 1 to 13 a - 5 . The optical amplifier 15 - 1 amplifies the output of this WSS 14 - 1 when sending it as an add signal to the coupler C 1 - 1 . Other three add routes A# 2 to A# 4 have the same hardware structure as the first add route A# 1 described above.

The 1×2 couplers C 1 - 1 to C 1 - 4 on the add side split each given add signal into two branches to distribute the signal to two different elements. Specifically, the coupler C 1 - 1 for add route A# 1 distributes its add signal to input port IN 2 of the first 2×4 WSS 11 - 1 and input port IN 4 of the first 4×2 WSS 12 - 1 . Likewise, the coupler C 1 - 2 distributes an add signal of add route A# 2 to input port IN 2 of the second 2×4 WSS 11 - 2 and input port IN 4 of the second 4×2 WSS 12 - 2 . The coupler C 1 - 3 distributes an add signal of add route A# 3 to input port IN 2 of the third 2×4 WSS 11 - 3 and input port IN 4 of the third 4×2 WSS 12 - 3 . The coupler C 1 - 4 distributes an add signal of add route A# 4 to input port IN 2 of the fourth 2×4 WSS 11 - 4 and input port IN 4 of the fourth 4×2 WSS 12 - 4 .

The 2×1 couplers C 2 - 1 to C 2 - 4 on the drop side combine two optical paths into one to produce a drop signal. Specifically, the coupler C 2 - 1 for drop route D# 1 receives signals from output port OUT 4 of the first 2×4 WSS 11 - 1 and output port OUT 2 of the first 4×2 WSS 12 - 1 . The resulting drop signal contains one of these two signals that is active. Likewise, the coupler C 2 - 2 combines signals from output port OUT 4 of the second 2×4 WSS 11 - 2 and output port OUT 2 of the second 4×2 WSS 12 - 2 , thereby producing a drop signal for drop route D# 2 . The coupler C 2 - 3 combines signals from output port OUT 4 of the third 2×4 WSS 11 - 3 and output port OUT 2 of the third 4×2 WSS 12 - 3 , thereby producing a drop signal for drop route D# 3 . The coupler C 2 - 4 combines signals from output port OUT 4 of the fourth 2×4 WSS 11 - 4 and output port OUT 2 of the fourth 4×2 WSS 12 - 4 , thereby producing a drop signal for drop route D# 4 .

In the drop section shown in the lower-right quarter of FIG. 9 , the optical cross-connect 1 a includes the following elements for drop route D# 1 : a 2×1 coupler C 2 - 1 , an optical amplifier (WDM amplifier) 16 - 1 , a 1×5 WSS 17 - 1 , five 1×8 WSSs 18 a - 1 to 18 a - 5 , and forty receivers Rx- 1 to Rx- 40 . The optical amplifier 16 - 1 amplifies a drop signal supplied from the corresponding coupler C 2 - 1 . The 1×5 WSS 17 - 1 has five output ports connected respectively to the input ports of different 1×8 WSSs 18 a - 1 to 18 a - 5 . These WSSs 18 a - 1 to 18 a - 5 bring the drop signal to any one of their forty output ports as specified. Eight outputs of the first 1×8 WSS 18 a - 1 are directed to a group of receivers Rx- 1 to Rx- 8 , as are the outputs of the second 1×8 WSS 18 a - 2 to a subsequent group of receivers Rx- 9 to Rx- 16 . Likewise, other three 1×8 WSSs 18 a - 3 to 18 a - 5 provide their outputs to corresponding receiver groups Rx- 17 to Rx- 24 , Rx- 25 to Rx- 32 , and Rx- 33 to Rx- 40 . Each receiver Rx- 1 to Rx- 40 receives an optical drop signal (if supplied). Other three drop routes D# 2 to D# 4 have the same hardware structure as the first drop route D# 1 .

Switching Between Input Route and Add Route ( 1 )

Referring now to FIGS. 10 to 13 , this section will describe how the optical cross-connect 1 a switches optical signals of input route Rin# 1 and add route A# 1 . FIG. 10 shows an example of switching paths that the optical cross-connect 1 a can produce. The bold dotted line indicates an add path AD 1 from add route A# 1 to output route Rout# 1 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 .

Referring back to FIG. 9 , the optical cross-connect 1 a provides two paths for delivering an add signal to a specific output route. One path goes from a coupler C 1 - 1 to C 1 - 4 to input port IN 2 of a 2×4 WSS 11 - 1 to 11 - 4 , and then through a selected 4×2 WSS 12 - 1 to 12 - 4 . This path is chosen in the case where the add route number is different from the output route number of interest. The other path goes from a coupler C 1 - 1 to C 1 - 4 to input port IN 4 of a corresponding 4×2 WSS 12 - 1 to 12 - 4 . This path is chosen in the case where the add route number is the same as the output route number of interest.

Referring again to the case of FIG. 10 , an add signal received from add route A# 1 goes to output route Rout# 1 . This add signal is directed to input port IN 4 of the first 4×2 WSS 12 - 1 since the add route number # 1 equals the output route number # 1 . The selected add path AD 1 reaches the intended output route Rout# 1 without interfering with the through path TH 2 , TH 3 , or TH 4 . Thus, in the case shown in FIG. 10 , the add path AD 1 can work simultaneously with any one of the through paths TH 2 , TH 3 , and TH 4 .

FIG. 11 shows another example of switching paths. The bold dotted line indicates an add path AD 2 from add route A# 1 to output route Rout# 2 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 .

In the case shown in FIG. 11 , the add path AD 2 cannot work simultaneously with the through path TH 2 . This is because the WSS 11 - 1 is inherently unable to route optical signals from two input ports IN 1 and IN 2 to a single output port OUT 1 (see FIG. 5 for the infeasible pattern of two paths colliding at an output port). The add path AD 2 cannot work simultaneously with the through path TH 3 either. This is because the WSS 11 - 1 is inherently unable to route one optical signal from input port IN 1 to output port OUT 2 when another optical signal goes from input port IN 2 to output port OUT 1 (see FIG. 6 for the case where two paths being twisted). For the same reason, the add path AD 2 cannot work simultaneously with the through path TH 4 .

FIG. 12 shows yet another example of switching paths. The bold dotted line indicates an add path AD 3 from add route A# 1 to output route Rout# 3 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 .

In the case shown in FIG. 12 , the add path AD 3 can work simultaneously with the through path TH 2 . However, the same add path AD 3 cannot work simultaneously with the next through path TH 3 because the two paths collide at output port OUT 2 of the WSS 11 - 1 . The add path AD 3 is unable to work simultaneously with the through path TH 4 either since these two paths are twisted within the WSS 11 - 1 .

FIG. 13 shows still another example of switching paths. The bold dotted line indicates an add path AD 4 from add route A# 1 to output route Rout# 4 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 .

In the case shown in FIG. 13 , the add path AD 4 cannot work simultaneously with the through path TH 2 since, at the WSS 11 - 1 , their output ports OUT 1 and OUT 3 are not immediately adjacent to each other (see FIG. 4 ). On the other hand, the add path AD 4 can work simultaneously with the next through path TH 3 . The add path AD 4 cannot work simultaneously with the through path TH 4 since the two paths collide at output port OUT 3 of the WSS 11 - 1 .

Switching Between Input Route and Drop Route ( 1 )

Referring now to FIGS. 14 to 17 , this section will describe how the optical cross-connect 1 a switches optical signals of, for example, input route Rin# 1 and drop route D# 1 . FIG. 14 shows several switching paths that the optical cross-connect 1 a can produce. The bold dotted line indicates a drop path DR 1 from input route Rin# 1 to drop route D# 1 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 .

In the case shown in FIG. 14 , the drop path DR 1 cannot work simultaneously with the through path TH 2 . This is because the WSS 11 - 1 is inherently unable to bring an optical signal from input port IN 1 to two output ports OUT 1 and OUT 4 at the same time (see FIG. 8 ). For the same reason, the drop path DR 1 cannot work simultaneously with the other through paths TH 3 and TH 4 .

FIG. 15 shows another example of switching paths. The bold dotted line indicates a drop path DR 2 from input route Rin# 2 to drop route D# 1 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 . In the case shown in FIG. 15 , the drop path DR 2 can work simultaneously with any one of the three through paths TH 2 , TH 3 , and TH 4 .

FIG. 16 shows yet another example of switching paths. The bold dotted line indicates a drop path DR 3 from input route Rin# 3 to drop route D# 1 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 . In the case shown in FIG. 16 , the drop path DR 3 can work simultaneously with any one of the three through paths TH 2 , TH 3 , and TH 4 .

FIG. 17 shows still another example of switching paths. The bold dotted line indicates a drop path DR 4 from input route Rin# 4 to drop route D# 1 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 . In the case shown in FIG. 17 , the drop path DR 4 can work simultaneously with any one of the three through paths TH 2 , TH 3 , and TH 4 .

Cross-Connection Feasibility Table ( 1 )

This section describes tables that represent simultaneous feasibility of optical cross-connections. FIGS. 18 and 19 show examples of such tables T 11 and T 12 , respectively. Specifically, table T 11 summarizes simultaneous feasibility of cross-connections between input route Rin# 1 and add route A# 1 , which has been discussed in an earlier section with reference to FIGS. 10 to 13 . Table T 12 , on the other hand, summarizes simultaneous feasibility of cross-connections between input route Rin# 1 and drop route D# 1 , which has been discussed in the preceding section with reference to FIGS. 14 to 17 .

The table entries having a value of “YES” indicate that the corresponding combination of switching paths can work simultaneously with each other. The top-left entry of table T 11 , for example, is marked “YES,” meaning that an add switching path from add route A# 1 to output route Rout# 1 can work together with a through switching path from input route Rin# 1 to output route Rout# 2 .

Table entries having a value of “NO” indicate that the corresponding combination of switching paths cannot work simultaneously (in other words, one of those paths can work selectively). The letter in parentheses gives the reason for that non-simultaneity. Specifically, “NO (a)” means that two switching paths collide at an output port of WSS. “NO (b)” means that two switching paths are twisted within a WSS. “NO (c)” means that two output ports are not immediately adjacent. “NO (d)” means that one input signal is directed to multiple output ports. See the entry at column 2, row 1 of table T 11 , for example. This entry has a value of “NO (a),” indicating that the add switching path from add route A# 1 to output route Rout# 2 cannot work together with a through switching path from input route Rin# 1 to output route Rout# 2 because these two paths collide at an output port of a WSS.

FIG. 20 shows yet another table representing simultaneous feasibility of optical cross-connections. This table T 13 is a logical product of table T 11 and table T 12 , which is created by performing a logical operation on each corresponding pair of entries of those two tables. Specifically, let x and y represent respective values of table T 11 and table T 12 at a particular column and row. The logical product of two table entries (x, y) takes a value of “YES” if (x,y)=(YES,YES) or a value of “NO” if (x,y)=(YES,NO) or (NO,YES) or (NO,NO). The resulting value is then set to a corresponding entry of table T 13 .

In table T 13 , the entries having a value of “YES” indicate that the corresponding through switching path and add/drop switching paths can work simultaneously. See the entry at column 3 , row 1 , for example. This entry has a value of “YES,” meaning that a through switching path from input route Rin# 1 to output route Rout# 2 , an add switching path from add route A# 1 to output route Rout# 3 , and a drop switching path from input route Rin# 3 to drop route D# 1 can all work at the same time.

The other entries of table T 13 have a value of “NO” to indicate that either the corresponding through switching path or the corresponding add/drop switching paths can work selectively. See the top-left entry of table T 13 , for example. This table entry indicates that the optical cross-connect 1 a does not allow the simultaneous use of a through switching path from input route Rin# 1 to output route Rout# 2 , an add switching path from add route A# 1 to output route Rout# 1 , and a drop switching path from input route Rin# 1 to drop route D# 1 . Reversely stated, the table entry suggests that either the through switching path alone or the add/drop switching paths can be implemented.

In reality, it may not be necessary to make available every combination of a through path and add/drop paths for simultaneous use. Rather, most cases require either the former or the latter alone. Under this assumption, table T 13 is more optimal, or less redundant, than table T 63 discussed earlier in FIG. 39 since table T 13 contains a smaller number of entries marked “YES” than table T 63 . The present embodiment reduces the number of simultaneously workable switching patterns.

More specifically, compare the switching patterns P 1 to P 4 shown in table T 13 of FIG. 20 with the corresponding entries of table T 63 of FIG. 39 . The conventional optical cross-connect 60 ( FIG. 35 ) allows the simultaneous use of those switching patterns P 1 to P 4 . In contrast, the optical cross-connect 1 a according to the present embodiment removes the simultaneity of such patterns P 1 to P 4 . In the case of pattern P 1 , the optical cross-connect 1 a requires to choose either a through switching path from Rin# 1 to Rout# 3 or add/drop switching paths from A# 1 to Rout# 2 and from Rin# 2 to D# 1 . In the case of pattern P 2 , it requires to choose a through switching path from Rin# 1 to Rout# 4 or add/drop switching paths from A# 1 to Rout# 2 and from Rin# 2 to D# 1 . Further, in the case of pattern P 3 , it requires to choose a through switching path from Rin# 1 to Rout# 4 or add/drop switching paths from A# 1 to Rout# 3 and from Rin# 3 to D# 1 . In the case of pattern P 4 , it requires to choose a through switching path from Rin# 1 to Rout# 2 or add/drop switching paths from A# 1 to Rout# 4 and from Rin# 4 to D# 1 .

In summary of this section, the proposed optical cross-connect 1 a eliminates many of the unnecessary combinations of through switching paths and add/drop switching paths, thus optimizing the number of WSSs used and interconnections between them. The present invention thus contributes to the realization of an optimally-sized optical cross-connect.

Optical Cross-Connect ( 2 )

This section describes a modified version of the foregoing optical cross-connect 1 a according to the present invention. FIG. 21 shows an optical cross-connect 2 a designed to perform switching between N routes each having 40 WDM channels (N=4 in this example). This optical cross-connect 2 a has a 2×(N+1) WSS for each input route and an (N+1)×2 WSS for each output route. Here, the prefix “2×(N+1)” stands for two inputs and (N+1) outputs, and “(N+1)×2” stands for (N+1) inputs and two outputs. More specifically, four 2×5 WSSs 21 - 1 to 21 - 4 are coupled to input routes Rin# 1 to Rin# 4 , and four 5×2 WSSs 22 - 1 to 22 - 4 are coupled to output routes Rout# 1 to Rout# 4 .

The 2×5 WSSs 21 - 1 to 21 - 4 are connected to the 5×2 WSSs 22 - 1 to 22 - 4 in the following way: Output port OUT 1 (first output port) of the first 2×5 WSS 21 - 1 is connected to input port IN 2 (second input port) of the first 5×2 WSS 22 - 1 . Output port OUT 2 (second output port) of the WSS 21 - 1 is connected to input port IN 1 (first input port) of the first 5×2 WSS 22 - 1 . Output port OUT 3 of the WSS 21 - 1 is connected to input port IN 3 of the second 5×2 WSS 22 - 2 . Output port OUT 4 of the WSS 21 - 1 is connected to input port IN 3 of the third 5×2 WSS 22 - 3 . Output port OUT 5 of the WSS 21 - 1 is connected to input port IN 3 of the fourth 5×2 WSS 22 - 4 .

Output port OUT 1 of the second 2×5 WSS 21 - 2 is connected to input port IN 2 of the 5×2 WSS 22 - 2 . Output port OUT 2 of the WSS 21 - 2 is connected to input port IN 1 of the second 5×2 WSS 22 - 2 . Output port OUT 3 of the WSS 21 - 2 is connected to input port IN 3 of the first 5×2 WSS 22 - 1 . Output port OUT 4 of the WSS 21 - 2 is connected to input port IN 4 of the third 5×2 WSS 22 - 3 . Output port OUT 5 of the WSS 21 - 2 is connected to input port IN 4 of the fourth 5×2 WSS 22 - 4 .

Output port OUT 1 of the third 2×5 WSS 21 - 3 is connected to input port IN 2 of the third 5×2 WSS 22 - 3 . Output port OUT 2 of the WSS 21 - 3 is connected to input port IN 1 of the third 5×2 WSS 22 - 3 . Output port OUT 3 of the WSS 21 - 3 is connected to input port IN 4 of the first 5×2 WSS 22 - 1 . Output port OUT 4 of the WSS 21 - 3 is connected to input port IN 4 of the second 5×2 WSS 22 - 2 . Output port OUT 5 of the WSS 21 - 3 is connected to input port IN 5 of the fourth 5×2 WSS 22 - 4 .

Output port OUT 1 of the fourth 2×5 WSS 21 - 4 is connected to input port IN 2 of the fourth 5×2 WSS 22 - 4 . Output port OUT 2 of the WSS 21 - 4 is connected to input port IN 1 of the fourth 5×2 WSS 22 - 4 . Output port OUT 3 of the WSS 21 - 4 is connected to input port IN 5 of the first 5×2 WSS 22 - 1 . Output port OUT 4 of the WSS 21 - 4 is connected to input port IN 5 of the second 5×2 WSS 22 - 2 . Output port OUT 5 of the WSS 21 - 4 is connected to input port IN 5 of the third 5×2 WSS 22 - 3 .

In the add section shown in the lower-left quarter of FIG. 21 , the optical cross-connect 2 a has the following elements to select an add signal on the first add route A# 1 : forty wavelength-tunable transmitters Txv- 1 to Txv- 40 , five 8×1 WSSs 23 a - 1 to 23 a - 5 , a 5×1 WSS 24 - 1 , and an optical amplifier (WDM amplifier) 25 - 1 . The forty transmitters Txv- 1 to Txv- 40 transmit forty optical signals, respectively. Each 8×1 WSS 23 a - 1 to 23 a - 5 receives eight out of forty optical signals from the transmitters Txv- 1 to Txv- 40 and outputs one selected optical signal to the subsequent 5×1 WSS 24 - 1 . The 5×1 WSS 24 - 1 then chooses one optical signal out of five optical signals received from the 8×1 WSS 23 a - 1 to 23 a - 5 . The optical amplifier 25 - 1 amplifies the output of this WSS 24 - 1 when sending it as an add signal to the first 2×5 WSS 21 - 1 . Other three add routes A# 2 to A# 4 have the same hardware structure as the first add route A# 1 described above.

The first optical amplifier 25 - 1 for add route A# 1 sends its output signal (i.e., selected add signal) to input port IN 2 of the first 2×5 WSS 21 - 1 . Likewise, the second optical amplifier 25 - 2 for add route A# 2 sends its output signal to input port IN 2 of the second 2×5 WSS 21 - 2 . The third optical amplifier 25 - 3 for add route A# 3 sends its output signal to input port IN 2 of the third 2×5 WSS 21 - 3 . The fourth optical amplifier 25 - 4 for add route A# 4 sends its output signal to input port IN 2 of the fourth 2×5 WSS 21 - 4 .

In the drop section shown in the lower-right quarter of FIG. 21 , four optical amplifiers (WDM amplifiers) 26 - 1 to 26 - 4 serve as the front end for switched drop signals. Specifically, the first optical amplifier 26 - 1 receives a signal for drop route D# 1 from output port OUT 2 of the first 5×2 WSS 22 - 1 . Likewise, the second optical amplifier 26 - 2 receives a signal for drop route D# 2 from output port OUT 2 of the second 5×2 WSS 22 - 2 . The third optical amplifier 26 - 3 receives a signal for drop route D# 3 from output port OUT 2 of the third 5×2 WSS 22 - 3 . The fourth optical amplifier 26 - 4 receives a signal for drop route D# 4 from output port OUT 2 of the fourth 5×2 WSS 22 - 4 .

In addition to the optical amplifier 26 - 1 mentioned above, the first drop route D# 1 involves a 1×5 WSS 27 - 1 , five 1×8 WSSs 28 a - 1 to 28 a - 5 , and forty receivers Rx- 1 to Rx- 40 . The optical amplifier 26 - 1 amplifies a drop signal received from the first 5×2 WSS 22 - 1 . The 1×5 WSS 27 - 1 has five outputs connected respectively to five input ports of different 1×8 WSSs 28 a - 1 to 28 a - 5 . With those 1×8 WSSs 28 a - 1 to 28 a - 5 , the drop signal is directed to one of their forty output ports as specified. Specifically, eight outputs of the first 1×8 WSS 28 a - 1 are directed to a group of receivers Rx- 1 to Rx- 8 , as are the outputs of the second 1×8 WSS 28 a - 2 to a subsequent group of receivers Rx- 9 to Rx- 16 . Likewise, other three 1×8 WSSs 28 a - 3 to 28 a - 5 provide their outputs to corresponding receiver groups Rx- 17 to Rx- 24 , Rx- 25 to Rx- 32 , and Rx- 33 to Rx- 40 . Each receiver Rx- 1 to Rx- 40 receives an optical drop signal (if supplied). Other three drop routes D# 2 to D# 4 have the same hardware structure as the first drop route D# 1 described above.

Switching Between Input Route and Add Route ( 2 )

Referring now to FIGS. 22 to 25 , this section will describe how the optical cross-connect 2 a switches optical signals of, for example, input route Rin# 1 and add route A# 1 . FIG. 22 shows an example of switching paths that the optical cross-connect 2 a can produce. The bold dotted line indicates an add path AD 1 from add route A# 1 to output route Rout# 1 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 .

In the case shown in FIG. 22 , the add path AD 1 cannot work simultaneously with the through path TH 2 since these two paths are twisted in the 2×5 WSS 21 - 1 . For the same reason, the add path AD 1 cannot work simultaneously with the other two through paths TH 3 and TH 4 .

FIG. 23 shows another example of switching paths. The bold dotted line indicates an add path AD 2 from add route A# 1 to output route Rout# 2 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 .

In the case shown in FIG. 23 , the add path AD 2 from add route A# 1 to output route Rout# 2 cannot work simultaneously with the through path TH 2 since these two paths collide at output port OUT 3 of the 2×5 WSS 21 - 1 . The add path AD 2 is unable to work simultaneously with the through path TH 3 or TH 4 since the paths are twisted within the 2×5 WSS 21 - 1 .

FIG. 24 shows yet another example of switching paths. The bold dotted line indicates an add path AD 3 from add route A# 1 to output route Rout# 3 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 .

In the case shown in FIG. 24 , the add path AD 3 can work simultaneously with the through path TH 2 . The add path AD 2 , however, cannot work simultaneously with the through path TH 3 since these two paths collide at output port OUT 4 of the 2×5 WSS 21 - 1 . The add path AD 2 is also unable to work simultaneously with the other through path TH 4 since these paths are twisted within the 2×5 WSS 21 - 1 .

FIG. 25 shows still another example of switching paths. The bold dotted line indicates an add path AD 4 from add route A# 1 to output route Rout# 4 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 .

In the case shown in FIG. 25 , the add path AD 4 cannot work simultaneously with the through path TH 2 since, at the 2×5 WSS 21 - 1 , their output ports OUT 3 and OUT 5 are not immediately adjacent to each other (see FIG. 4 ). The add path AD 4 , on the other hand, can work simultaneously with the through path TH 3 . The add path AD 4 cannot work simultaneously with the through path TH 4 since the two paths collide at output port OUT 5 of the 2×5 WSS 21 - 1 .

Switching Between Input Route and Drop Route ( 2 )

Referring now to FIGS. 26 to 29 , this section describes how the optical cross-connect 2 a switches optical signals of, for example, input route Rin# 1 and drop route D# 1 . FIG. 26 shows an example of switching paths that the optical cross-connect 2 a can produce. The bold dotted line indicates a drop path DR 1 from input route Rin# 1 to drop route D# 1 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 .

In the case shown in FIG. 26 , the drop path DR 1 cannot work simultaneously with the through path TH 2 since it is impossible for a WSS to route a single input signal to two output ports at the same time. For the same reason, the drop path DR 1 cannot work simultaneously with the other through paths TH 3 and TH 4 .

FIG. 27 shows another example of switching paths. The bold dotted line indicates a drop path DR 2 from input route Rin# 2 to drop route D# 1 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 . In the case shown in FIG. 27 , the drop path DR 4 can work simultaneously with any one of the three through paths TH 2 , TH 3 , and TH 4 .

FIG. 28 shows yet another example of switching paths. The bold dotted line indicates a drop path DR 3 from input route Rin# 3 to drop route D# 1 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 . In the case shown in FIG. 28 , the drop path DR 3 can work simultaneously with any one of the three through paths TH 2 , TH 3 , and TH 4 .

FIG. 29 shows still another example of switching paths. The bold dotted line indicates a drop path DR 4 from input route Rin# 4 to drop route D# 1 . The bold solid lines represent the following three through paths: TH 2 from input route Rin# 1 to output route Rout# 2 ; TH 3 from input route Rin# 1 to output route Rout# 3 ; and TH 4 from input route Rin# 1 to output route Rout# 4 . In the case shown in FIG. 29 , the drop path DR 4 can work simultaneously with any one of the three through paths TH 2 , TH 3 , and TH 4 .

Cross-Connection Feasibility Table ( 2 )

This section describes tables that represent simultaneous feasibility of optical cross-connections. FIGS. 30 and 31 show examples of such tables T 21 and T 22 , respectively. Specifically, table T 21 summarizes simultaneous feasibility of optical cross-connections between input route Rin# 1 and add route A# 1 , which has been discussed in an earlier section with reference to FIGS. 22 to 25 . Table T 22 , on the other hand, summarizes simultaneous feasibility of optical cross-connections between input route Rin# 1 and drop route D# 1 , which has been discussed in the preceding section with reference to FIGS. 26 to 29 . For the notation of those tables, see the preceding section entitled “Cross-Connection Feasibility Table ( 1 )” describing FIGS. 18 and 19 .

FIG. 32 shows another table representing simultaneous feasibility of optical cross-connections. This table T 23 is a logical product of table T 21 and table T 22 , which is created by performing a logical operation on each corresponding pair of entries of those two tables. Specifically, let x and y represent respective values of table T 21 and table T 22 at a particular column and row. The logical product of two table entries (x, y) takes a value of “YES” if (x,y)=(YES,YES) or a value of “NO” if (x,y)=(YES,NO) or (NO,YES) or (NO,NO). The resulting value is then set to a corresponding entry of table T 23 .

In table T 23 , the entries having a value of “YES” indicate that the corresponding through switching path and add/drop switching paths can work simultaneously. See, for example, the entry at column 3 , row 1 . This entry has a value of “YES,” meaning that a through switching path from input route Rin# 1 to output route Rout# 2 , an add switching path from add route A# 1 to output route Rout# 3 , and a drop switching path from input route Rin# 3 to drop route D# 1 can all work at the same time.

The other entries of table T 23 have a value of “NO” to indicate that either the corresponding through switching path or the corresponding add/drop switching paths can work selectively. See the top-left entry of table T 23 , for example. This table entry indicates that the optical cross-connect 2 a does not allow simultaneous use of a through switching path from input route Rin# 1 to output route Rout# 2 , an add switching path from add route A# 1 to output route Rout# 1 , and a drop switching path from input route Rin# 1 to drop route D# 1 . Reversely stated, the table entry suggests that either the through switching path alone or the add/drop switching paths can be implemented.

In reality, it may not be necessary to make available every combination of a through path and add/drop paths for simultaneous use. Rather, most cases require either the former or the latter alone. Under this assumption, table T 23 is more optimal, or less redundant, than table T 63 discussed earlier in FIG. 39 since table T 23 contains a smaller number of entries marked “YES” than table T 63 . The present embodiment reduces the number of simultaneously workable switching patterns.

More specifically, compare the switching patterns P 1 to P 4 shown in table T 23 of FIG. 32 with the corresponding entries of table T 63 of FIG. 39 . The conventional optical cross-connect 60 ( FIG. 35 ) allows the simultaneous use of those switching patterns P 1 to P 4 . In contrast, the optical cross-connect 2 a according to the present embodiment removes the simultaneity of such patterns P 1 to P 4 . In the case of P 1 , the optical cross-connect 2 a requires to choose either a through switching path from Rin# 1 to Rout# 3 or add/drop switching paths from A# 1 to Rout# 2 and from Rin# 2 to D# 1 . In the case of pattern P 2 , it requires to choose a through switching path from Rin# 1 to Rout# 4 or add/drop switching paths from A# 1 to Rout# 2 and from Rin# 2 to D# 1 . Further, in the case of pattern P 3 , it requires to choose a through switching path from Rin# 1 to Rout# 4 or add/drop switching paths from A# 1 to Rout# 3 and from Rin# 3 to D# 1 . In the case of pattern P 4 , it requires to choose a through switching path from Rin# 1 to Rout# 2 or add/drop switching paths from A# 1 to Rout# 4 and from Rin# 4 to D# 1 .

In summary of this section, the proposed optical cross-connect 2 a eliminates many of the unnecessary combinations of through switching paths and add/drop switching paths, thus optimizing the number of WSSs used and interconnections between them. The present invention thus contributes to the realization of an optimally-sized optical cross-connect.

To summarize the above discussion, the present invention offers an optical cross-connect including a 2×N (2-input, N-output) wavelength selective switch for each of N input routes and a N×2 (N-input, 2-output) wavelength selective switch for each of N output routes. One input port of each 2×N wavelength selective switch receives an optical signal from a corresponding input route, while the other input port of the same receives an optical add signal from a corresponding add route. One output port of each N×2 wavelength selective switch outputs an optical signal to a corresponding output route, while the other output port outputs an optical drop signal to a corresponding drop route. This structure eliminates unnecessary combinations of through switching paths and add/drop switching paths, thus greatly reducing the number of WSSs used and interconnections between them.

The foregoing is considered as illustrative only of the principles of the present invention. Further, since numerous modifications and changes will readily occur to those skilled in the art, it is not desired to limit the invention to the exact construction and applications shown and described, and accordingly, all suitable modifications and equivalents may be regarded as falling within the scope of the invention in the appended claims and their equivalents.