SYSTEM OF LARGE- SCALE ROBOTIC FIBER CROSS-CONNECTS USING MULTI-FIBER TRUNK RESERVATION

Abstract

A large scale, non-blocking fiber optic cross-connect system consists of multiple stages, including a central multifiber per connection system. The number of ports of this cross-connect system scales to over 10,000, in an incremental, modular, field expandable approach. Two separate arrays of “edge” cross-connect systems using KBS methodology are positioned on opposite sides of a central core cross-connect system, wherein the core system is comprised of switchable blocks of multi-fiber trunk lines, each terminated in a single connector that is reconfigurable by robotic means. The trunk lines between edge cross-connects are controlled by a trunk line management system to provision/deprovision blocks of multiple connections at a time in a “core” cross-connect circuit block between edge cross-connects. The core system is configured to controllably interconnect the physically separate edge cross-connect systems which concurrently direct data along selected paths to and from the central core circuit block.

Claims

1-10. (canceled)

11. A system for controllably switching optical signal carrying physical links among a variable plurality of optical fibers, the system comprising: a first plurality of physical link sorting modules, each including optical fibers that carry different ones of a plurality of separate input optical signals to different optical fiber outputs, each sorting module being constructed and adapted to respond individually to first command signals to interweave different physical links therein to selected ones of the optical fiber outputs; a plurality of signal transferring core trunk lines receiving the optical signals from the first plurality of physical link sorting modules at different ones of a plurality of multi-fiber trunk line groupings, extending to different ones of a plurality of spaced-apart outputs via variable trunk line interconnections controlled in response to a second set of command signals; a second plurality of physical link sorting modules constructed and adapted to receive the optical signals transported from the multi-fiber trunk lines, and also constructed and adapted to respond to third command signals to deliver optical signals received from the trunk lines to selected output terminals from each of the second plurality of physical link sorting modules; and a control system coupled to transmit command signals to control the physical link sorting modules and configure the signal transferring core trunk lines in accordance with requirements then existing for optical signal transmission along different selected physical links.

12. The system of claim 11, wherein the first plurality of physical link sorting modules and the second plurality of physical link sorting modules are variable in number, and each physical link sorting module has like pluralities of separate inputs and outputs.

13. The system of claim 11, wherein each link sorting module is constructed and adapted to interweave the physical links therein between other fibers in the module without entanglement by transporting the physical links through the system in accordance with an algorithm computing physical links as mathematical strands whose spatial relationships are ordered by the mathematics of knots, braids and strands to ensure entanglement of strands is prevented.

14. The system of claim 11, wherein the first plurality of physical link sorting modules and the second plurality of physical link sorting modules comprise network topology managers having like numbers of optical fibers in variably determinable paths between inputs and outputs.

15. The system of claim 14, wherein the plurality of signal transferring core trunk lines define a trunk line manager comprising a multiplicity of multi-fiber trunk line sets, each set comprising a fixed like number of optical fibers which is a fraction of a total number of core trunk lines in a corresponding network topology manager.

16. The system of claim 15, wherein the trunk line manager variably interconnects the output terminals of a first network topology manager with input terminals of a second network topology manager in accordance with commands from the control system.

17. The system of claim 15, wherein the network topology managers each have about 1,000 or more input fiber ports and 1,000 or more output fiber ports, and the trunk line manager has about the same number of multi-fiber inputs and outputs.

18. The system of claim 17, wherein each input and output is able to be interconnected by multi-fiber trunk lines.

19. The system of claim 14, wherein the first and second pluralities of physical link sorting modules each comprise like numbers of network topology manager blocks, each block having a like number of inputs switchable under command signals to selected individual outputs for a given individual block, and wherein the core trunk lines are arranged in like fractional groupings of a limited number of fiber sets equaling in total the like numbers of fibers in the network topology manager blocks.

20. The system of claim 11, wherein the first plurality of physical link sorting modules and the second plurality of physical link sorting modules and the core trunk lines are bi-directional.

21. The system of claim 11, wherein the number of first and second plurality of original physical link sorting modules can be increased and/or decreased during operation.

22. A switching system for controlling a transfer of optical signals from a plurality of optical input lines to selected optical output lines as determined by control signals, comprising: a plurality of optical input lines arranged in a two-dimensional array in a first plane having orthogonal input axes and extending out of the first plane to a parallel second plane wherein the lines are distributed about a first central axis, the lines between the array of the first plane and the second plane being constructed and adapted in a three-dimensional first switching mechanism to be positioned by first activators in response to first control signals in accordance with a knots, braids and strands principle and responsively moved to selectable first switching positions in the first plane; a second signal controlled multi-line and intermediate switching system coupled to the first switching positions and including multi-lines extending therefrom to a third plane, the intermediate switching system arranged in a two-dimensional array in the third plane having orthogonal input axes and extending out of the third plane to a parallel fourth plane wherein the multi-lines are distributed about a second central axis, the multi-lines between the array of the third plane and the fourth plane being constructed and adapted in a three-dimensional second switching mechanism to be positioned by second activators in response to second control signals in accordance with a knots, braids and strands principle and responsively moved to selectable second switching positions in the third plane; a third plurality of optical lines operatively coupled to the multi-lines in the intermediate switching system and receiving signals therefrom, the plurality of optical output lines arranged in a two-dimensional array in a fifth plane having orthogonal input axes and extending out of the fifth plane to a parallel sixth plane wherein the lines are distributed about a third central axis, the lines between the array of the fifth plane and the sixth plane being constructed and adapted in a three-dimensional third switching mechanism to be positioned by third activators in response to third control signals in accordance with a knots, braids and strands principle and responsively moved to selectable third switching positions in the fifth plane; and a central system generating control signals for: (i) the first switching mechanisms, (ii) the second switching mechanisms, and (iii) the third switching mechanisms, said control signals to control a selective transfer of optical signals through the system from input lines to output lines.

23. The switching system of claim 22, wherein rows of the first plane, third plane, and the fifth plane are independently translatable by the activators.

24. The switching system of claim 22, wherein the ordering of lines at the second plane, at the multi-lines at fourth plane, and at the fifth plane is fixed.

25. The switching system of claim 22, wherein the multi-lines number 8, 12, 16, 24 or 32 fibers in a common bundle.

26. The switching system of claim 25, wherein the multi-lines are single mode optical fiber within a common protective jacket.

27. The switching system of claim 22, wherein the control signals cause activator motors to move.

28. The switching system of claim 22, wherein the line and multi-line switching mechanisms each comprise a robotic arm and a connector gripper.

29. The switching system of claim 28, wherein the connector gripper can repeatedly engage single fiber connectors or multiple fiber connectors.

30. The switching system of claim 22, wherein the lines extending from initial coordinates in the first plane merge into a fixed one-dimensional array of locations at the second plane.

31. The switching system of claim 30, wherein the lines extending from the initial coordinates follow substantially straight paths.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0121] A better understanding of the invention may be had by reference to the accompanying drawings, in which the reference numbers refer to like parts, and in which:

[0122] FIG. 1 is a generalized block diagram of a system in accordance with exemplary embodiments hereof, comprising a TLM core unit and oppositely coupled input and output array edge units with generalized system controls;

[0123] FIG. 2 is an enlarged, fragmentary perspective view of a portion of the TLM's internal multi-fiber trunk lines used in the system of FIGS. 1 and 3;

[0124] FIG. 3 is a combined but generalized block diagram of a more detailed example of the large-scale cross-connect system in accordance with exemplary embodiments hereof;

[0125] FIG. 4 is a combined but generalized block diagram of the system with a further factor of two increase in port capacity;

[0126] FIG. 5 is an elevation view of a robotic gripper device as positioned in relation to multiple stacked rows of MPO (Mechanical Pull Out) connector receptacles;

[0127] FIG. 6 is a second elevation view of the gripper device shown in FIG. 5 showing further features thereof;

[0128] FIG. 7 is a block diagram of a portion of an aisle of adjacent NTMs wherein all NTMs are joined by a common robotic platform with one or more robotic modules traveling upon a common track connecting and spanning multiple cross-connect bays;

[0129] FIG. 8 is a perspective view of one of a stacked multiplicity of transversely activated rows, each row having multiple connector receptacles and corresponding linear tracks joining and spanning multiple cross-connect bays, and multiple 12-fiber bundles converging at a one-dimensional backbone;

[0130] FIG. 9 is a fragmentary perspective view of a portion of the MPO cross-connect system, showing details as to the cable ports, the MPO connector receptacles, and the translating connector receptacles;

[0131] FIG. 10 illustrates aspects of system in accordance with exemplary embodiments;

[0132] FIG. 11 illustrates a system in accordance with exemplary embodiments hereof in which a redundant cross-connect provides 96 to 1,056 links, non-blocking, without a core NTM;

[0133] FIG. 12 depicts an aspects of a core NTM according to exemplary embodiments hereof;

[0134] FIG. 13 illustrates an extension of the cross-connect of FIG. 11 beyond 1,056 links, for example to 1,248 links, by incorporation of a core NTM;

[0135] FIG. 14 depicts the extension of the cross-connect of FIGS. 11 to 2,112 links; and

[0136] FIG. 15 illustrates in block diagram form the further extension of the example of FIG. 11 to 12,672 links.

GLOSSARY AND ABBREVIATIONS

[0137] As used herein, including in the claims, unless used or described otherwise, the following terms or abbreviations have the following meanings:

[0138] LC means Lucent connector (or Little Connector or Local Connector);

[0139] MT means mechanical transfer;

[0140] MPO means multi-fiber push on; and

[0141] MTP is used by US Conec to describe their MPO connector. MTP is a registered trademark of US Conec.

DETAILED DESCRIPTION

[0142] This application relates to and extends from the previously disclosed patents and other publications on the Knots, Braids and Strands (KBS) methodology for controllably managing elements of a dense array of optical fibers without physical entanglement or unacceptable signal loss. A number of novel features and relationships devised to further advance the state of the KBS are disclosed and discussed in this introductory section of this specification to provide context and enable appreciation of more detailed features presented thereafter.

[0143] Refer now to FIG. 1, which includes other background features but principally depicts three generalized system blocks 101, 102, 103, greatly and differentially consolidated, representing important features and arranged as a large scale, multi-stage cross-connect system 110 in accordance with exemplary embodiments hereof. Each system block 101, 103 is functionally a physical link sorting module, enabling links to be arranged in any arbitrary ordered state by the controller 100 and subsequently reordered without limitation as needed. The system 110 is depicted for simplicity as unidirectional, with inputs 108 on the left and outputs 109 on the right, but it will be evident that a bidirectional capability is also evident but not described at this point. Here the system 110 blocks begin at the left with a first entry system 101-1 . . . 101-12 representing multiple KBS units, here presumptively positioned in parallel in a horizontal axis, with multiple spaced apart parallel inputs 108 on the left. The inputs feed the internal and separate but nonetheless system 100 controlled KBS repositionings in the central system 102. Outputs from the entry system 101 are repositioned to other coordinates as directed by the command system 100, but this repositioning, in accordance with exemplary embodiments hereof, is only preliminary and partial. At this point the partially redistributed fibers are not then coupled to the output system 103 of KBS units, except in the configuration, noted below, but to an interposed Trunk Line Manager 102 (TLM) which is positioned before the final bank of KBS units 103-1 . . . 103-12 in the output system 103. The TLM 102 effects another line redistribution before the final outputs 109 are first configured and then delivered, from the output system 103.

[0144] The inclusion and operation of the TLM 102 in concert with the multi-line entry system 101 and multi-line output system 103 present a number of novel departures in the technology of optical line switching. The Trunk Line Manager 102 does not switch individual fiber optic lines but sets of optical fibers (here twelve (12) each (although the number can be varied), encapsulated in an individual sleeve 34, as seen in the fragmentary view of FIG. 2. These fiber groupings or sets extend from the input side of the trunk line manager to the output side. Variable length buffers 120 in a compact arrangement compensate for length variations between the individual trunk line groupings. The design of such fiber buffers has been described previously in the aforementioned patents to Kewitsch et al.

[0145] Referring again to FIG. 1, it is to be understood that programmed independent and simultaneous movement of selected individual fibers extending from their vertical locations takes place through different appropriate active non-blocking KBS units 101 and 103. This step continues until their intermediate target locations on the entry side are reached so that interconnection can be made at the TLM 102. The referenced KBS patents can be accessed for their teachings as to three-dimensional interweaving of fibers as they are transported from one location to another, as to lateral incremental shifting of fiber rows to augment interweaving.

[0146] It should be understood that, although the optical fibers are manipulated in the Trunk Line Manager 102 in twelve line sets within their common housings, they are nonetheless interwoven as sets using the Knots, Braids and Strands methodology. Each fiber bundle is interwoven through the three-dimensional distribution of fiber bundles in the Trunk Line Manager using the KBS teaching until an input fiber 108 is connected to an output fiber 109 by an internal cross-connect.

[0147] At the output array 103, the individual inputs from the Trunk Line Manager 102 are again routed, in KBS mode, to selected individual outputs 109 under control of the routing command System 100. Consequently, it can be seen that there is not only capability for routing thousands of links selectively, but that a vast lateral switching capability is available on demand through the available capacity in the multiple and integrated lines in the Trunk Line Manager.

[0148] However, the present system is further uniquely different, as is evident from FIG. 3, and merits some detailed discussion here, of the intermediate section of the light transit paths between the input and output terminals. These paths changeably traverse their successive and different lengths of optical fiber, which are each separately positioned. Specifically, the input light paths are controllably routed into the entry system 101 on a single fiber KBS route, then go along a chosen multi-fiber KBS route 106 through the Trunk Line Manager 103 and then to the output system 103 by another controllable multi-fiber KBS route 107 to the chosen array. This provides a massive number of controllably variable light communication paths distributed transversely and extending longitudinally between the spaced apart input and output terminal arrays. Note that the longitudinal division of the KBS units is subdivided into initial 101, intermediate 102 and final 103, each unique individually as well as in combination. The tens of thousands of input lines 108 from external sources to the large-scale interconnect system 110 are received by the different KBS units 101-1 . . . 101-12 (see FIG. 3, etc.) within the system 110, partially redistributed under system command, and delivered individually to chosen inputs of the Trunk Line Manager unit 102. The TLM 102 is configured to very significantly increase the usable number of conveniently available signal paths between the input terminal array formed in combination by 101-1 . . . 101-12 and the output terminal array formed in combination by 103-1 . . . 103-12, employing novel configuration of optical fiber groups, each comprised of multiple fibers (typically 12 as shown herein), and said groups (rather than individual optical fibers) each configured in the KBS format.

[0149] More specifically, the TLM 102 comprises changeable KBS matrices in which the grouped sets 203 of fibers each traverse changeable paths between inputs adjacent the entry system 101 and outputs adjacent the output system 103. The internal fiber sets 203 changeably connected to external fiber groups 208, 209 on the TLM 102 engage as a group to an individual coupler 208 on the outer surface of the TLM 102, such that like couplers 209 at the opposite side of the TLM 102 through connections 107 can be attached to the output system 103. The single connection arrangement of internal multi-fiber sets 203 affords hitherto unattainable operative flexibility, there spatial limits established by a common spring-loaded buffer 112 serving each group 203 of twelve lines. The circular buffers have like radii of curvature large enough to prevent optical signal loss from the buffer loop.

[0150] As noted previously, within the TLM 102, the sets of internal fibers (internal multi fiber trunk lines 111) have repositionable segments 203 that are translatable under command in the KBS format, with length variations introduced during movement being automatically compensated by the buffers 112 as the common connectorized end 204 of the multiple fibers 203 moves to a new location, where it is coupled to a fiber on the exit system 103, which can be commanded through system 100 to move to a selected output 109.

[0151] Thus, the TLM 102 manages multi-fiber trunk lines 111 terminated in switchable connectors 204 and forms programmable interconnection fabric between the network topology managers 101, 103. To this end, trunk lines spanning the network topology managers are reserved by the topology management system 100 to provision contiguous blocks of twelve connections at a time. In a system 110 such as seen in FIG. 1, including a TLM 102 with twelve fibers per port, strict non-blocking scalability is achievable for greater than about 10,000×10,000 optical links. It will be shown that this capacity can be varied to be either smaller or larger, either fractionally or in multiples.

[0152] The system uses a novel configuration of multiply grouped optic fibers comprising a single optical fiber bundle 203 or “strand.” Referring now to FIG. 2, it can be seen that the multiple fiber groupings (here twelve in number although other groupings can be used, are diametrically compact with an encompassing cover or housing or sleeve 34 of suitable pliant material with connector end 204. FIG. 2 illustrates in somewhat idealized form (for clarity) a portion of the relative configuration of several internal multi-cables 203-1, 203-2, 203-3 as they are positioned in the TLM 102. Each multi-cable, as previously described, has a dozen separate but closely proximate individual optical fiber strands 35 comprising multi-fiber bundle 203 extending from a connector 204 on its changeable end for coupling to a signal from the input array 101 and is the repositionable segment of the multi-segment but contiguous trunk line 111 within the TLM. The paths of all multi-cables in the TLM 102 are changeable on command, as previously described, and they are selectively transferred. They each terminate, as shown in FIG. 2, in a multi-fiber block length compensating buffer 112 before they are transferred on to the output array 103. This enables the use of substantial path length variations in the TLM. This configuration permits fiber groupings to be drawn to different longitudinal lengths for connection purposes to different coupling points. In a particular example, individual optical fibers with a cladding diameter of 80 microns and a coating outer diameter of 165 microns are preferred because it allows 12 fibers to be closely packed within a flexible, low friction plastic tube having an outer diameter of less than 2 mm, preferably less than 1 mm Suitable materials for the extruded plastic sleeve include PVC, PVDF, PTFE, PFA, ETFE, etc.

[0153] Refer again to FIG. 3 for a more complete explanation of the management and operation of a high capacity optical cross-connect 110 example in accordance with exemplary embodiments hereof. This example is also configured to provide any-to-any connectivity between 12,672 transmit and receive lines, and constitutes a substantial advance in the state of the art. Nonetheless, other examples are also given below of systems in accordance with exemplary embodiments hereof having both greater and lesser number of lines, as seen in the Tables that follow. In FIG. 3 the system is shown as 12,672 individual transmitting links (Tx) and 12,672 receiving links (Rx). System control capability of the individual Input Edge NTMs 101, Output Edge NTMs 103 as well as the Core Trunk Line Manager 102 is evidenced by the above referenced patent of the present assignee, so this aspect will not be described in detail. Instead, as seen in FIG. 3, overall control capability is encapsulated within the overall Trunk Line Reservation System 100 as previously described relative to FIG. 1.

[0154] FIG. 4 shows a further factor of two increase in cross-connect capacity, for a system comprised or two cross-connects 110-1, 110-2 that extend the capability to, for example, 11,220 transmit and received lines to equipment of type A and 11,220 transmit and receive lines to equipment of type B.

[0155] In FIG. 3, as in FIG. 1, the array of Input Edge Network Topology Managers comprises an array of twelve units 101-1, 101-2 . . . 101-12, as does the array of twelve units 103-1, 103-2 . . . 103-12 of Output Edge Network Topology Managers. The input edge array 101 of NTMs partially and internally redirect the individual inputs they receive onto selected ones of the plurality of trunk lines 106 as input to the core TLM 102. In the core TLM, plural line groupings 111 are established and output, now as redirected trunk line pluralities 107. The output edge NTMs 103 again redirect these received signals under unified command to different terminals at the final outputs 109. As illustrated in FIG. 11, it will be recognized that not all the NTM fiber modules 104 shown need be included, or employed, and that the system can be incrementally scaled to cross-connect from 96 to 12,652 links or more as shown in the sequence of FIGS. 10-13. It will also be recognized that external connections and details, such as input and output ports, LC connectors and MPO connectors have not been shown for simplicity in view of their obviousness in the state of the art.

[0156] It should be noted that the modularity and configuration of these systems facilitate versatility in system design and adaptation for both varying design and usage. Referring to FIG. 11, each NTM in comprised of say, eleven fiber modules 104 that each consist of 96 individual fibers, terminated at both ends, as with conventional single fiber LC connectors (not shown in detail). The network of NTMs 101, 103, each comprised of fiber modules 104, can be scaled upwardly, up to twenty-four NTMs can be individually and incrementally connected to the core (or central) TLM 102, twelve on the input side and twelve on the output side. NTMs may be distributed across the data center and their trunk lines may be run back to a central TLM.

[0157] In contrast in FIG. 12, a TLM fiber module 105 comprises 96 trunk lines, with each trunk line consisting of twelve (12) separate fibers as a set, for a total of 1,132 fibers for the set. The TLM modules each include 96 twelve-fiber bundles, each bundle terminated in a single 12 fiber MPO connector. Therefore, the cumulative total can rise to 12,672×12,672 fibers as shown.

[0158] In accordance with exemplary embodiments hereof, the TLM thus functions as a multiplier of cross-connect ports while maintaining the desired non-blocking connectivity. For a TLM with 1,056 multi-fiber parts, the overall duplex or simplex port count is multiplied by the number of fibers per trunk line. For a system based on 12-fiber MPO connectors within the TLM, the port multiplier is twelve (12), for a 24-fiber MPO the port multiplier is twenty-four (24). A significant novel capability of this inventive concept is that trunk lines in the multi-fiber sets may be unassigned and thus held available until they are reserved by control system 100 to link a particular input and output NTM with a multi-fiber trunk line 111. The maximum number of unassigned and therefore potentially available ports may be defined as “reservation port overhead.” For example, potentially up to eleven fibers of a 12-fiber trunk line 111 may initially be unassigned. The edge NTMs reserve TLM trunk lines to other edge NTMs in 12-fiber increments, even if only one (1) fiber is needed initially. Therefore, the maximum number of reserved but potentially unused fibers in the trunk lines due to the discrete nature of 12-fiber trunks, totals eleven unutilized fibers per trunk, multiplied by eleven underutilized trunks per system, and multiplied by eleven systems. This totals 1,452 potentially unused but reserved fibers out of the total 12,672. By limiting the number of edge links to the total number of links minus the reservation overhead (11,220 links in this particular example), the trunk link reservation approach ensures arbitrary, any-to-any non-blocking interconnections. In this case the trunk reservation blocking probability is nil.

[0159] FIG. 3 is only one example of applicant's massively scalable cross-connect, providing from 96 to 12,672 non-blocking links. The core TLM manages 12 fiber trunk lines between the edge NTMs and provides first non-blocking connections. Expandability is an advantageous further feature, moreover, because the 96 (23-fiber trunk)×96 (12 fiber trunk) modules are added incrementally to the core TLM as additional trunk lines and edge NTMs are deployed. This cross connect system is incrementally scalable.

[0160] The robotic cross-connect systems 110 disclosed in exemplary embodiments hereof provide low loss, software-defined fiber optic connections between an extremely large number of pairs of ports 108, 109. The system consists of a central or core cross-connect 102 in which trunk lines 111, including internal switchable segments 203, connecting opposing cross-connect switches 101-n, 103-n, are reserved in blocks of twelve fibers. The core NTM 102 is connected to a number of edge cross-connects 101, 103 that can arbitrarily reconfigure up to 1056 individual LC fiber ports. The number of ports of the scalable arrangement 110 ranges from 48×48 up to 12,672×12,672 and beyond. The core cross-connect 102 serves as a port multiplier to achieve ever increasing port counts in an incremental, redundant, non-blocking fashion. The port multiplication factor is directly related to the number of individual optical fibers 38 grouped into a trunk line 111. The core 102 and edge 101, 103 cross-connects operate by identical KBS principles. The core reconfigures trunk lines 111 with switchable segment 203 terminated in 12-fiber MPO connectors 204, while the edge 101, 103 reconfigures individual, single fibers 112 with moveable LC connectors 113 for full cross-connect flexibility without blocking. The process of internal connector 204 transport to a destination mating adapter 113 in both implementations includes a coordinated, sequential, multi-step movement of one or more robots 300 each transporting a gripper 205 within a high density of surrounding, suspended interconnects 203, and programmatic shuffling of each connector row 202 in accordance with the KBS algorithm as described in U.S. Pat. No. 8,463,091 referenced above. Before the robot plugs in the internal LC or MPO fiber connector 204 to its chosen final port 208, the polished fiber end face of fiber connector may be cleaned by an automated fiber end face cleaning module. The MPO 12-fiber connector 204 terminates the internal bundle 203 of twelve optical fibers, wherein each set of optical fibers 38 originate from an individual automatic, spring loaded take-up reel buffer assembly 120 residing on a tray comprised of multiple closely spaced take-up reels. The take-up reels ensure that all internal optical fiber bundles are maintained under slight tension in the fiber interconnect volume between the MPO connectors and the take-up reels, so that they follow substantially straight-line paths for all possible arrangements of connectors within ports.

[0161] This invention advances the state-of-the-art in non-blocking automated cross-connects from 1,000 links to greater than 10,000 software reconfigurable physical links as shown in block diagram form in FIG. 10. This incrementally scalable system of high performance automated optical links 110 consists of multiple NTMs 101, 103 interconnected by a single core TLM system 102 and orchestrated by the software/system 100. The TLM manages blocks of multi-fiber trunk lines 111,203 terminated in switchable MPO connectors 204 as shown in FIG. 2 and forms a programmable interconnection fabric connecting the intermediate links 106, 107 between NTMs. Trunk lines spanning NTMs are reserved by the TLM management system 102 to provision contiguous blocks of, say, twelve connections at a time, so that strict non-blocking scalability is achievable for 11,220×11,220 optical links in a system comprised overall of 12,672×16,672 optical fibers. FIG. 1 illustrates the block diagram of this system, including a TLM with over 1000 individual 12-fiber MPO ports.

[0162] In a further example specific to the twelve-fiber trunk line arrangement, twelve-fiber trunk lines 203 comprising the TLM provide the programmable, non-blocking fabric connecting inputs and outputs without being constrained by “blocking” or connectivity restrictions. While this system is inherently bi-directional, for typical duplex links using dual fibers, the transmit (Tx) fibers 108 are connected to the inputs and the receive (Rx) fibers 109 are connected to the outputs. The optical lines of the TLM 102 are reserved by the software in blocks of twelve-fibers. The TLM can connect two up to twenty-four NTMs (incrementally scalable) and each NTM can arbitrarily reconfigure up to 1,056 individual LC fiber ports.

[0163] In accordance with exemplary embodiments hereof, the TLM and NTMs operate by identical mathematical principles. The KBS algorithm is independent of whether the strand is physically a single or multi-fiber strand. The only requirement is that each strand follows a straight line in the interconnect volume where the cross-connections are actuated and managed. The TLM 102 at the core reconfigures trunk lines 111 terminated in multi-fiber (i.e., 12, 24, 36, 48) MPO connectors 204, for example, while the NTMs at the edge reconfigure individual, single fiber LC connectors 113. The internal connector gripper transport in both the TLM and NTM includes a coordinated, sequential, multi-step movement of the robot and programmatic shuffling of each connector row (as described in U.S. Pat. No. 8,463,091). However, the TLM and NTM are different with respect to their hardware implementation. In the TLM, an MPO multi-fiber connector 204 terminates an internal bundle of multiple optical fibers 203 and each set of optical fibers originates from a single spring-loaded take-up reel arranged on a flat tray comprised of multiple take-up reels. The take-up reel/fiber buffer 112 maintains tension of the fiber bundle between the MPO connectors and the take-up reels, ensuring the bundles follow straight-line paths for all possible configurations, which is necessary to ensure no physical entanglement of the bundles.

Link Reconfiguration Process with Multi-Fiber Trunk Reservation

[0164] The process of establishing a new physical connection in this system of NTMs/TLM consists of the following steps: [0165] 1. Specify input port and output port of new link [0166] 2. Determine corresponding Input Edge NTM 101-i and Output Edge NTM 103-j [0167] 3. Determine if an unused fiber is available on existing, reserved trunk line 111 between Input Edge NTM and Output Edge NTM. If none are available, first provision and reserve a new multi-fiber trunk line 111 between the Edge NTMs using the Core TLM 102. [0168] 4. Select a fiber within trunk line 111 whose connectivity at Edge NTMs can be switched most rapidly [0169] 5. Perform cross-connect between Input Edge NTM and input trunk line 106 [0170] 6. Perform cross-connect between Output Edge NTM and output trunk line 107 [0171] 7. Upon completion, a low insertion loss optical link between selected input and port ports is established

[0172] In a particular example shown in FIGS. 11 and 12 which highlights the modularity of this system, each NTM fiber module 104 consists of ninety-six individual fibers terminated at both ends with LC connectors. In contrast, each TLM fiber module 105 consists of ninety-six trunk lines 111, with each trunk line consisting of twelve separate fibers reconfigured as a set, for a total of 1,152 fibers. Either or both fiber modules 104, 105 can be added incrementally in a non-service affecting manner to an existing, partially filled TLM system 102. Each TLM fiber module includes ninety-six twelve-fiber ribbons or trunk lines, each ribbon terminated with twelve-fiber MPO connectors 204. As the network scales, up to twenty-four NTMs can be individually and incrementally connected to the central TLM, twelve on the input side (101-1 . . . 101-12) and twelve on the output side (103-1 . . . 103-12). NTMs may be distributed across the data center and their trunk lines may be run back to the central TLM.

[0173] The TLM increases the number of non-blocking cross-connect ports and forms a highly scalable switch matrix. For a TLM with 1,056 multi-fiber ports, the overall duplex or simplex port count for the system of NTMs is multiplied by the number of fibers per trunk line. For a system based on twelve-fiber MPO connectors within the TLM, the port multiplier is twelve, for twenty-fiber MPO the port multiplier is twenty-four. In exemplary embodiments hereof, the term “reservation port overhead” is defined. The overhead refers to the maximum number of potentially unavailable ports due to the reservation of partially filled trunk lines. For example, potentially one to eleven fibers of a twelve-fiber trunk line may be unassigned. The reservation port overhead for various configurations are presented in Table 1.

[0174] In a particular example in which trunk lines are in twelve fiber groupings, which is ideal in terms of keeping reservation overhead below 11.5%, the edge NTMs 101, 103 reserve TLM 102 trunk lines 111 to other edge NTMs in twelve fiber increments, even if only one fiber is needed initially. Therefore, the maximum number of reserved by potentially unused fibers in the trunk lines, due to the discrete nature of twelve-fiber trunks, totals eleven unutilized fibers per trunk, multiplied by eleven underutilized trunks per system, and multiplied by eleven systems. This totals 1,452 potentially unused but reserved fibers out of the total 12,672. By limiting the number of edge links to the total number of links minus the reservation overhead, the trunk link reservation approach ensures arbitrary, any-to-any non-blocking interconnections. In this case the trunk reservation blocking probability is null.

TABLE-US-00001 TABLE 1 Max No. of Links No. Links No. Links No. of (minus Trunk Minimum per Edge per Core No. Edge No. Fibers reserve- No. of reservation Reservation Link NTM NTM NTMs per Trunk ed links Links overhead) Overhead Availability 1,056 1,056 2 2 2 2,112 2,110 0.1% 99.9% 1,056 1,056 4 4 36 4,224 4,188 0.9% 99.1% 1,056 1,056 6 6 150 6,336 6,186 2.4% 97.6% 1,056 1,056 6 6 150 6,336 6,186 2.4% 97.6% 1,056 1,056 12 12 1,452 12,672 11,220 11.5% 88.5% 1,056 1,056 24 24 12,696 25,344 12,648 50.1% 49.9% 1,056 2,112 24 24 6,348 50,688 44,340 12.5% 87.5% 2,112 2,112 6 6 150 12,672 12,522 1.2% 98.8% 2,112 2,112 12 12 1,452 25,344 23,892 5.7% 94.3% 2,112 2,112 24 24 12,696 50,688 37,992 25.0% 75.0% 2,112 2,112 36 36 44,100 76,032 31,932 58.0% 42.0%

[0175] In a further example, Table 1 above presents scaling options based on the number of links per NTM as well as the number of fibers 38 per trunk line 111. The maximum number of reserved links relates to how many reserved but unused fibers within the trunk lines are possible under worst-case conditions. This reduces the number of ports available for cross-connection. The rightmost column entitled “Minimum Link Availability” describes the minimum fraction of links able to be arbitrarily cross-connected under various system configurations. Note that it may not be desirable in some cases to increase the number of fibers per trunk to 24, 36 or 48, because the reservation overhead (column 8 in Table 1) can increase significantly and negatively impact the system scalability. In general, it is preferred to minimize the reservation overhead and maximize the minimum link availability.

[0176] Incremental Scaling by 96 Ports

[0177] In a particular example shown in FIGS. 11 and 12, each fiber module 104 within the 1,056-port core cross-connect (101 or 103) is a 96×96 port device. Each module supports 96 ports of twelve fiber cable assemblies. The fiber modules can be added incrementally in a non-service affecting manner to an existing, partially filled cross-connect system Each module includes potentially eight ribbon fibers or fiber bundles terminated with twelve fiber MPO connectors on one side, and 96 individual LC ports on the opposite side.

[0178] In a further example, each fiber module 105 within the 1,056-port core cross-connect 102 consists of 96 trunk lines, each trunk line consisting of twelve bundled fibers. The fiber modules 105 can be added incrementally in a non-service affecting manner to an existing, partially filled core cross-connect system 102. Each module includes potentially 96 twelve fiber bundles, each bundle terminated with a 12-fiber MPO connectors on both ends.

[0179] Incremental Scaling by 1,056 Ports

[0180] In a further example, each edge cross-connect 101, 103 is comprised of 1056 links and each block of 1,056 ports can be added in redundant pairs by connecting each of the twelve fiber cables, terminated in MPO connectors, directly to the MPO core cross-connect. Edge cross-connect systems totaling up to 24 can be added incrementally in a non-service affecting manner.

[0181] In accordance with exemplary embodiments hereof, the core cross-connect 102 serves as a port multiplier by utilizing a multi-fiber trunk line reservation approach. For a core cross-connect with 1,056 multi-fiber ports, the overall duplex or simplex port count for the system of multiple 1,056 single fiber ports can be multiplied by the number of fibers per trunk line. For a system based on 12-fiber MPO connectors, the port multiplier is 12, for 24-fiber MPO the port multiplier is 24, for 36-fiber MPO the port multiplier is 36 and for 48 fiber MPO the port multiplier is 48, corresponding to 48,384 ports. Port overhead refers to the maximum number of potentially unavailable ports due to the reserving of partially filled trunk lines and the potential for 1 to 11 fibers of a trunk line to be unassigned.

[0182] 1:1 Redundancy

[0183] Each LC edge cross-connect 101, 103 with trunk lines 111 therebetween forms a pair of cross-connects with a potentially redundant pair of robots 300. This pair of devices can be reconfigured at both ends by any one of two edge cross-connects 101, 103, each edge cross-connect having its own robot 300.

[0184] Fast Execution of Cross-Connects

[0185] Multiple robots 300 in multiple edge cross-connects 101 can execute reconfigurations in parallel (FIG. 7). For the 12,672-port system, 12 robots can cross-connect 12 different connections simultaneously. Each cross-connect typically takes 1 minute to execute, so this enables the average cross-connect time to be reduced to 5 seconds. As shown in FIG. 7, the one or more robots operating in parallel can travel up and down an aisle on a common track 301 to reconfigure multiple NTMs 101 in parallel.

[0186] Robotic Gripper for MPO Connections

[0187] A gripper 205 attached to the end of robotic arm 201 (FIGS. 5-6) engages and disengages an internal female MPO connector 204 terminating the miniature fiber bundle 203 to plug it in or unplug it from an internal MPO receptacle 208 on the moveable row 202. The industry standard MPO connector uses a push-pull latching action and MPO stands for “Mechanical Pull Out” in the art. The gripper produces the plug/unplug actuation force by use of a miniature stepper motor 206. A compact solenoid 207 is used to lock the MPO connector 204 into the gripper 205. In a particular example, the MPO connector 204 further includes a permanent magnet with an attraction force of at least 5 N to an additional permanent magnet(s) attached to each connector track 210 for the purpose of retaining connector 204 within connector receptacle 208 and providing the fiber end face contact force necessary to maintain low insertion loss MPO connections.

[0188] Row of MPO Connections

[0189] MPO connectors 204 terminating multi-fiber bundles 203 plug into independently translatable rows 202 of connector receptacles 208 (FIG. 8). The alignment features of the gripper 205 latch onto and register with a particular connector track 210 to align with its corresponding connector receptacle 208. In a further example, a permanent magnet is affixed to distal end of track 210 to attract a corresponding magnet at the distal end of MPO connector body 204.

[0190] Reels for Ribbon Fiber

[0191] Examples of retractable fiber reels in which continuous lengths of fiber are packaged in a spring-loaded reel are disclosed in U.S. Pat. No. 7,315,681. In a particular example, the small diameter bundle is converted to a 12-fiber ribbon within the reel, which is about 3.0 mm tall and is interleaved along the spiral path defined by a coiled spring element. The coiled spring element is ideally 0.125 to 0.250 mm thick, 3.0 mm tall spring steel.

[0192] Backbone for 12-Fiber Bundle in Protective Strain Relief Tube with Dense 1 mm Spacing

[0193] Plastic tubing that is sufficiently flexible and low friction is applied to the fiber bundle as a protective sleeve 34. In a particular example, the tubing and optical fiber bundle therein can pass through a backbone array of flexible tubes to manage the bend radius of the fiber bundle within its interior. The minimum outer diameter of a close-packed 12-fiber bundle comprised of reduced cladding (RC) optical fiber is about 4×0.165 mm=0.66 mm and this can fit within a 1.5 mm outer diameter sleeve. If ribbon fiber is used, the individual fibers must first be singulated along the length to be bundled by using the Corning Inc. ribbon splitting tool RST-000, for example.

TABLE-US-00002 TABLE 2 Max No. of Links No. Links No. Links No. of (minus Trunk Minimum per Edge per Core No. Edge No. Fibers reserved No. of reservation Reservation Link NTM NTM NTMs per Trunk links Links overhead) Overhead Availability 1,056 1,056 1 12 0 1,056 1,056 0.0% 100.0% 1,056 1,056 2 12 12 2,112 2,100 0.6% 99.4% 1,056 1,056 3 12 48 3,168 3,120 1.5% 98.5% 1,056 1,056 4 12 108 4,224 4,116 2.6% 97.4% 1,056 1,056 5 12 192 5,280 5,088 3.6% 96.4% 1,056 1,056 6 12 300 6,336 6,036 4.7% 95.3% 1,056 1,056 7 12 432 7,392 6,960 5.8% 94.2% 1,056 1,056 8 12 588 8,448 7,860 7.0% 93.0% 1,056 1,056 9 12 768 9,504 8,736 8.1% 91.9% 1,056 1,056 10 12 972 10,560 9,588 9.2% 90.8% 1,056 1,056 11 12 1,200 11,616 10,416 10.3% 89.7% 1,056 1,056 12 12 1,452 12,572 11,120 11.5% 88.5%

[0194] Port Scaling via Multi-Fiber Trunk Line Reservation System Managed by a Central,

[0195] Multi-Fiber Core Cross-Connect

[0196] The maximum number of reserved but potentially unused fibers in the trunk lines 111, due to the discrete nature of twelve fiber trunks joining the edge cross-connects 101, 103 to the core cross-connect 102, totals eleven unutilized fibers per truck, eleven underutilized trunks per system, trunks to eleven other cross-connects=1,331 fibers out of the total 12,096. The edge NTMs reserve trunk lines of the core cross-connect in twelve fiber increments, even if only one fiber is needed initially. This potentially leaves eleven unused fibers per trunk, reserved for later provisioning.

[0197] Table 2 presents different scaling options based on the number of links per NTM 101 as well as the number of fibers per trunk link 111 and presents the trunk reservation overhead for 12 fibers per trunk line, as the number of fibers is scaled incrementally from 1,056 to 12,672. As trunk lines are reserved, potentially only one fiber within the multiple fiber trunk line is utilized. The maximum number of reserved links relates to how many reserved but unused fibers within the trunk lines are possible under worst-case conditions. This reduces the number of ports able to be cross-connected by an overhead factor. The rightmost column in Table 2, entitled “Minimum Link Availability”, describes the fraction of links 111 able to be arbitrarily cross-connected under various system configurations. The trunk link reservation approach does not degrade the non-blocking characteristic of the cross-connect system. The cross-connect system continues to support arbitrary, any-to-any non-blocking interconnections throughout the scaling process.

CONCLUSION

[0198] As used herein, including in the claims, the phrase “using” means “using at least,” and is not exclusive. Thus, e.g., the phrase “using Z” means “using at least Z.” Unless specifically stated by use of the word “only”, the phrase “using Z” does not mean “using only Z.”

[0199] In general, as used herein, including in the claims, unless the word “only” is specifically used in a phrase, it should not be read into that phrase.

[0200] It should be appreciated that the words “first” and “second” in the description and claims are used to distinguish or identify, and not to show a serial or numerical limitation. Similarly, the use of letter or numerical labels (such as “(a)”, “(b)”, and the like) are used to help distinguish and/or identify, and not to show any serial or numerical limitation or ordering.

[0201] As used herein, including in the claims, singular forms of terms are to be construed as also including the plural form and vice versa, unless the context indicates otherwise. Thus, it should be noted that as used herein, the singular forms “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise.

[0202] Throughout the description and claims, the terms “comprise”, “including”, “having”, and “contain” and their variations should be understood as meaning “including but not limited to”, and are not intended to exclude other components unless specifically so stated.

[0203] It will be appreciated that variations to the embodiments of the invention can be made while still falling within the scope of the invention. Alternative features serving the same, equivalent or similar purpose can replace features disclosed in the specification, unless stated otherwise. Thus, unless stated otherwise, each feature disclosed represents one example of a generic series of equivalent or similar features.

[0204] The present invention also covers the exact terms, features, values and ranges, etc. in case these terms, features, values and ranges etc. are used in conjunction with terms such as about, around, generally, substantially, essentially, at least etc. (i.e., “about 3” shall also cover exactly 3 or “substantially constant” shall also cover exactly constant).

[0205] As used herein, including in the claims, unless stated otherwise, ranges include their end values. Thus, e.g., a value in the range 2 to 12 includes the values 2 and 12. As another example, a number in the range 48 to 1,056 could be 48 or 1,056.

[0206] Use of exemplary language, such as “for instance”, “such as”, “for example” (“e.g.,”) and the like, is merely intended to better illustrate the invention and does not indicate a limitation on the scope of the invention unless specifically so claimed.

[0207] Those skilled in the art will readily observe that numerous modifications and alterations of the systems may be made while retaining the teachings of the invention. Accordingly, the above disclosure should be construed as limited only by the metes and bounds of the appended claims.

LEGEND

[0208] 34: Compliant sleeve [0209] 38: Individual fiber within internal trunk bundle [0210] 100: Trunk line reservation system [0211] 101: Edge cross-connects for single fiber link, input [0212] 102: Core cross-connects for multi-fiber links [0213] 103: Edge cross-connects for single fiber link, output [0214] 104: 96-port fiber module, simplex [0215] 105: 96-port fiber module, multiple fiber (e.g. 12 per connection) [0216] 106: Input edge to core fiber trunk line [0217] 107: Core to output edge fiber trunk line [0218] 108: Input fiber optic cables [0219] 109: Output fiber optic cables [0220] 110: Large scale cross-connect system [0221] 111: Multi-fiber trunk line [0222] 112: Internal fiber buffer [0223] 113: Input receptacle [0224] 114: Output receptacle [0225] 201: Base of robot telescopic arm [0226] 202: Row of MPO connector receptacles [0227] 203: Reduced diameter internal core fiber trunk line [0228] 204: MPO connector [0229] 205: Gripper [0230] 206: Stepper motor [0231] 207: Solenoid latch [0232] 208: Input MPO connector receptacle [0233] 209: Output MPO connector receptacle [0234] 210: Connector track [0235] 220: One dimensional internal backbone [0236] 300: Reconfiguration robot [0237] 301: Track for multiple reconfiguration robot configuration