High-throughput, low-latency next generation internet networks using optical tag switching

0001HIGH-THROUGHPUT, LOW-LATENCY NEXT GENERATION INTERNET NETWORKS USING OPTICAL TAG SWITCHING

BACKGROUND OF THE DISCLOSURE

00031. Field of the Invention

0004This invention relates to optical communication systems and, more

0005particularly, to an optical system which accommodates network traffic with high

0006throughput and low latency.

00072. Description of the Background Art

0008Recent research advances in optical Wavelength Division Multiplexing

0009(WDM) technology have fostered the development of networks that are orders of

0010magnitude higher in transmission bandwidth than existing commercial networks. While

0011such an increase in throughput is impressive on its own, a corresponding decrease in

0012network latency must also be achieved in order to realize the Next Generation Internet

0013(NGI) vision of providing the next generation of ultra high speed networks that can meet

0014the requirements for supporting new applications, including national initiatives. Towards

0015this end, current research efforts have focused on developing an ultra-low latency Internet

0016Protocol (IP) over WDM optical packet switching technology that promises to deliver the

0017two-fold goal of both high throughput with low latency. Such efforts, while promising,

0018have yet to fully realize this two-fold goal.

0019There are a number of challenging requirements in realizing such

0020IP/WDM networks. First, the NGI network must inter-operate with the existing Internet

0021and avoid protocol conflicts. Second, the NGI network must provide not only ultra low-

0022latency, but must take advantage of both packet-switched (that is, bursty) IP traffic and

0023circuit-switched WDM networks. Third, it is advantageous if the NGI network requires no synchronization between signaling and data payload. Finally, a desired objectiφ is

0024that the NGI network accommodates data traffic of various protocols and formats so that

0025it is possible to transmit and receive IP as well as non-IP signals without the need for

0026complicated synchronization or format conversion.

0027Comparison with other work

0028The Multi-Wavelength-Optical Network (MONET) system, as reported in

0029the article "MONET: Multi- Wavelength Optical Networking" by R. E. Wagner, et al. and

0030published in the Journal of Lightwave Technology, Vol. 14, No. 6, June 1996,

0031demonstrated a number of key milestones in optical network including transparent

0032transmission of multi-wavelength through more than 12 reconfigurable network elements

0033spread over the national scale fiber distance. The network, however, is circuit-switched

0034and suffers inefficiency in accommodating bursty traffic. The typical connection setup

0035time from request to switching is a few seconds, limited by capabilities of both Network

0036Control & Management (NC&M) and hardware. Recent efforts within the MONET

0037program to improve on the efficiency concentrated on the "Just-in-Time signaling"

0038scheme. This method utilizes embedded 1510 nm NC&M signaling which precedes the

0039data payload by an estimated delay time. This estimation must be accurately made for

0040each network configuration for every wavelength in order to synchronize the signaling

0041header and switching of the payload.

0042In accordance with the present invention, the optical packet header is

0043carried over the same wavelength as the packet payload data. This approach eliminates

0044the issue of header and payload synchronization. Furthermore, with a suitable use of

0045optical delay at each intermediate optical switch, it eliminates the need to estimate the

0046initial burst delay by incoφorating the optical delay directly at the switches. This makes a striking difference with Just-In-Time signaling in which the delay at each switctπalong

0047the path needs to be known ahead of time and must be entered in the calculation for the

0048total delay. Lastly, there is little time wasted in requesting a connection time and actually

0049achieving a connection. In comparison to a few second delays seen in MONET, the

0050present inventive subject matter reduces the delay to minimal, only limited by the actual

0051hardware switching delays at each switch. The current switching technology realizes

0052delays of only several microseconds, and shorter delays will be possible in the future.

0053Such a short delay can be incoφorated by using an optical fiber delay line at each

0054network element utilizing switches. The present inventive subject matter achieves the

0055lowest possible latency down to the fundamental limit of the hardware, and no lower

0056latency can be achieved by any other technique.

0057The Optical Networks Technology Consortium (ONTC) results were

0058reported in the article "Multiwavelength Reconfigurable WDM/ ATM/SONET Network

0059Testbed" by Chang et al. and published in the Journal of Lightwave Technology, Vol. 14,

0060No. 6, June 1996. Both Phase I (155 Mb/s, 4-wavelength) and Phase II (2.5 Gb/s, 8-

0061wavelength) of the ONTC program were configured on a Multihop ATM-based network.

0062While such an ATM based architecture added a large overhead and excluded the

0063possibility of a single-hop network, the packet/header signaling was made possible by

0064utilizing the isochronous ATM cell itself. This communication of NC&M information is

0065made through the same optical wavelength, potentially offering similar benefits as with

0066the technique of the present invention. However, the inventive technique offers a number

0067of significant advantages over the ATM-based signaling. First, the inventive technique

0068offers a single hop connection without the need to convert to electrical signals and buffer

0069the packets. Second, it offers far more efficient utilization of the bandwidth by eliminating excessive overheads. Third, it allows strictly transparent and ultra-low;

0070latency connections.

0071The ARPA sponsored All-Optical-Network (AON) Consortium results

0072were reported in an article entitled "A Wideband All-Optical WDM Network" , by I. P.

0073Kaminow et al. and published in the IEEE Journal on Selected Areas of Communication,

0074Vol. 14, No. 5, June, 1996. There were actually two parts of the AON program: WDM as

0075reported in the aforementioned article, and TDM reported in a companion paper in the

0076same issue. First the WDM part of the AON program is first discussed, followed by the

0077TDM part.

0078The AON architecture is a three-level hierarchy of subnetworks, and

0079resembles that of LANs, MANs, and WANs seen in computer networks. The AON

0080provides three basic services between Optical Terminals (OTs): A, B, and C services. A

0081is a transparent circuit-switched service, B is a transparent time-scheduled TDM/WDM

0082service, and C is a non-transparent datagram service used for signaling. The B service

0083uses a structure where a 250 msec frame is used with 128 slots per frame. Within a slot

0084or group of slots, a user is free to choose the modulation rate and format. The B-service

0085implemented on the AON architecture is closest to the IP over WDM which is the subject

0086matter of the present invention. However, the separation of NC&M signaling in the C-

0087service with the payload in the B-service requires careful synchronization between the

0088signaling header and the payload. This requirement becomes far more stringent as the

0089250 microsecond frame is used with 128 slots per frame with arbitrary bit rates. Not only

0090the synchronization has to occur at the bit level, but this synchronization has to be

0091achieve across the entire network. The scalability and interoperability are extremely

0092difficult since these do not go in steps with the network synchronization requirement. The present inventive subject matter requires no synchronization, inter-operates with

0093existing IP and non-IP traffic, and offers scalability.

0094TDM efforts are aimed at 100 Gb/s bit rates. In principle, such ultrafast

0095TDM networks have the potential to provide truly flexible bandwidth on demand at burst

0096rates of 100 Gb/s. However, there are significant technological challenges behind such

0097high bit rate systems mainly related to nonlinearities, dispersion, and polarization

0098degradations in the fiber. While the soliton technologies can alleviate some of the

0099difficulties, it still requires extremely accurate synchronization of the network — down to

0100a few picoseconds. In addition, the header and the payload must have the identical bit

0101rates, and as a consequence, bit-rate transparent services are difficult to provide. The

0102subject matter in accordance with the present invention requires no synchronization,

0103relies on no 100 Gb/s technologies, and offers transparent services.

0104The Cisco Coφoration recently announced a product based on Tag-

0105Switching and the general description of Cisco's Tag-Switching is available at the world-

0106wide-web site, (http://www.cisco.com/waφ/public/732/tag/). Cisco's (electronic) Tag

0107Switching assigns a label or "tag" to packets traversing a network of routers and switches.

0108In a conventional router network, each packet must be processed by each router to

0109determine the next hop of the packet toward its final destination. In an (electronic) Tag

0110Switching network, tags are assigned to destination networks or hosts. Packets then are

0111switched through the network with each node simply swaps tags rather than processing

0112each packet. An (electronic) Tag Switching network will consist of a core of (electronic)

0113tag switches (either conventional routers or switches), which connect to (electronic) tag

0114edge routers on the network's periphery. (Electronic) Tag edge routers and tag switches

0115use standard routing protocols to identify routes through the network. These systems then use the tables generated by the routing protocols to assign and distribute tag ^

0116information via a Tag Distribution Protocol. Tag switches and tag edge routers receive

0117the Tag Distribution Protocol information and build a forwarding database. The database

0118maps particular destinations to the tags associated with those destinations and the ports

0119through which they are reachable.

0120When a tag edge router receives a packet for forwarding across the tag

0121network, it analyzes the network-layer header and performs applicable network layer

0122services. It then selects a route for the packet from its routing tables, applies a tag and

0123forwards the packet to the next-hop tag switch.

0124The tag switch receives the tagged packet and switches the packet based

0125solely on the tag, without re-analyzing the network-layer header. The packet reaches the

0126tag edge router at the egress point of the network, where the tag is stripped off and the

0127packet delivered. After Cisco made its announcement about (Electronic) Tag Switching,

0128the IETF (Internet Engineering Task Force) has recommended a MPLS (Multi-protocol

0129Label Switching) to implement standardized, vendor-neutral (electronic) tag-switching

0130function in routers and switches, including ATM switches.

0131A number of features in the Cisco's (electronic) Tag Switching is similar

0132to the Optical Tag Switching which is the subject matter of the present invention, with

0133the features aimed at the similar goals of simplifying the processing required for packet

0134routing. The key differences are as follows. First, the optical tag switching is purely

0135optical in the sense that both tag and data payload are in an optical form. While each

0136plug-and-play module (a component of the present inventive system) senses the optical

0137tag, the actual packet does not undergo optical-to-electrical conversion until it comes out

0138of the network The Cisco's (electronic) Tag Switching will be all electrical, and applies electronic detection, processing, and retransmission to each packet at each router .ϋ

0139Secondly, the Optical Tag Switching of the present invention achieves lowest possible

0140latency and does not rely on utilizing buffers. Electronic tag switching will have far

0141greater latency due to electronic processing and electronic buffering. Thirdly, the Optical

0142Tag Switching of the present invention utilizes path deflection and/or wavelength

0143conversion to resolve blocking due to contention of the packets, whereas the Electronic

0144Tag Switching will only utilize electronic buffering as a means to achieve contention

0145resolution at the cost of increased latency, and the performance is strongly dependent on

0146packet size. The present invention covers packets of any length. Lastly, the Optical Tag

0147Switching of the present invention achieves a strictly transparent network in which data

0148of any format and protocol can be routed so long as it has a proper optical tag. Hence the

0149data can be digital of any bit rate, analog, or FSK (frequency-shifted-keying ) format.

0150The Electronic Tag Switching requires that data payload to have the given digital bit rate

0151identical to the electronic tag since the routers must buffer them electronically.

0152Another representative technology that serves as background to the

0153present invention is the so-called Session Deflection Virtual Circuit Protocol (SDVC),

0154which is based on deflection routing method. The paper entitled "The Manhattan Street

0155Network", by N. F. Maxemchuk" as published in the Proceedings on IEEE Globecom

0156'85, pp 255-261, December 1985, discusses that when two packets attempt to go to the

0157same destination, one packet can be randomly chosen for the preferred output link and

0158the other packet is "deflected" to the non-preferred link. This means that packets will

0159occasionally take paths that are not shortest paths. The deflection method utilized by the

0160present invention does not 'randomly' select the packet to go to the most preferred path;

0161rather, it attempts to look into the priorities of the packets, and send the higher priority packet to be routed to the preferred path. The packets will be deflected if they hav^ lower

0162priorities; however, both 'path deflection' and 'wavelength deflection' are utilized. The

0163path deflection is similar to conventional SDVC in that the optical packet will be simply

0164routed to the path of the next preference at the same wavelength. The wavelength

0165deflection allows the optical packet to be routed to the most preferred path but at a

0166different wavelength. This wavelength deflection is achieved by wavelength conversion

0167at the network elements. Partially limited wavelength conversion is utilized, meaning not

0168all wavelengths will be available as destination wavelengths for a given originating

0169wavelength. The wavelength deflection allows resolution of blocking due to wavelength

0170contentions without increasing the path delay. The combination of path and wavelength

0171deflections offers sufficiently large additional connectivities for resolving packet

0172contentions; however, the degree of partial wavelength conversion can be increased when

0173the blocking rate starts to rise. Such scalability and flexibility of the network is not

0174addressed by conventional SDVC.

SUMMARY OF THE INVENTION

0176The present invention utilizes a unique optical signaling header technique

0177applicable to optical networks. Packet routing information is embedded in the same

0178channel or wavelength as the data payload so that both the header and data information

0179propagate through the network with the same path and the associated delays. However,

0180the header routing information has sufficiently different characteristics from the data

0181payload so that the signaling header can be detected without being affected by the data

0182payload and that the signaling header can also be stripped off without affecting the data

0183payload. The inventive subject matter allows such a unique signal routing method to be overlaid onto the conventional network elements in a modular manner, by adding T o

0184types of 'Plug-and-Play' modules. The inventive subject matter overcomes the

0185shortcomings and limitation of other methods discussed in the Background section while

0186advantageously utilizing the full capabilities of optical networking.

0187In accordance with the broad method aspect of the present invention, a

0188method for propagating a data payload from an input network element to an output

0189network element in a wavelength division multiplexing system composed of a plurality of

0190network elements, given that the data payload has a given format and protocol, includes

0191the following steps: (a) generating and storing a local routing table in each of the network

0192elements, each local routing table determining a local route through the associated one of

0193the network elements; (b) adding an optical header to the data payload prior to inputting

0194the data payload to the input network element, the header having a format and protocol

0195and being indicative of the local route through each of the network elements for the data

0196payload and the header, the format and protocol of the data payload being independent of

0197the format and protocol of the header; (c) optically determining the header at each of the

0198network elements as the data payload and header propagate through the WDM network;

0199(d) selecting the local route for the data payload and the header through each of the

0200network elements as determined by looking up the header in the corresponding local

0201routing table; and (e) routing the data payload and the header through each of the network

0202elements in correspondence to the selected route.

0203In accordance with the broad system aspect of the present invention, the

0204system is arranged in combination with (a) an electrical layer; and (b) an optical layer

0205composed of a wavelength division multiplexing (WDM) network including a plurality

0206of network elements, for propagating a data payload generated by a source in the electrical layer and destined for a destination in the electrical layer, the data payloafl

0207having a given format and protocol. The system includes: (i) a first type of optical header

0208module, coupling the source in the optical layer and the WDM network, for adding an

0209optical header ahead of the data payload prior to inputting the data payload to the WDM

0210network, the header being indicative of a local route through the network elements for the

0211data payload and the header, the format and protocol of the data payload being

0212independent of those of the header; and (ii) a second type of optical header module,

0213appended to each of the network elements, for storing a local routing table in a

0214corresponding one of the network elements, each local routing table determining a

0215routing path through the corresponding one of the network elements, for optically

0216determining the header at the corresponding one of the network elements as the data

0217payload and header propagate over the WDM network, for selecting the local route for

0218the data payload and the header through the corresponding one of the network elements

0219as determined by looking up the header in the corresponding local routing table, and for

0220routing the data payload and the header through the corresponding one of the network

0221elements in correspondence to the selected route.

0222The present invention offers numerous features and benefits, including (1)

0223extremely low latency limited only by hardware delays; (2) high throughput and

0224bandwidth-on-demand offered by combining multi-wavelength networking and optical

0225tag switching; (3) priority based routing which allows higher throughput for higher

0226priority datagrams or packets; (4) scalable and modular upgrades of the network from the

0227conventional WDM to the inventive optical tag-switched WDM; (5) effective routing of

0228long datagrams, consecutive packets, and even non-consecutive packets; (6) cost-

0229effective utilization of optical components such as multiplexers and fibers; (7) interoperability in a multi-vendor environment; (8) graceful and step-by-step upgrades of

0230network elements; (9) transparent support of data of any format and any protocol; and

0231(10) high quality-of-service communications.

BRIEF DESCRIPTION OF THE DRAWINGS

0233The teachings of the present invention can be readily understood by

0234considering the following detailed description in conjunction with the accompanying

0235drawings, in which:

0236FIG. 1 is a pictorial representation of a general network illustrating the

0237coupling between the optical and electrical layers of the network as effected by one

0238aspect of the present invention;

0239FIG. 2 illustrates the optical layer of the network of FIG. 1 showing the

0240relationship between the optical signal header and data payload, and the use of the

0241header/payload in network setup;

0242FIG. 3 is a high-level block diagram of one Plug & Play module in

0243accordance with the present invention for header encoding and header removal;

0244FIG. 4 is a high-level block diagram of another Plug & Play module in

0245accordance with the present invention for routing a packet through a WDM network

0246element;

0247FIG. 5 is illustrative of a WDM circuit-switched backbone network;

0248FIG. 6 illustrates a network element of FIG. 1 with its embedded switch

0249and the use of local routing tables;

0250FIG. 7 depicts a block diagram of an illustrative embodiment of a header

0251encoder circuit for the Plug-&-Play module of FIG. 3; FIG. 8 depicts a block diagram of an illustrative embodiment of a hgader

0252remover circuit for the Plug-&-Play module of FIG. 3;

0253FIG. 9 depicts a block diagram of an illustrative embodiment of a header

0254detector circuit for the Plug-&-Play module of FIG. 4;

0255FIG. 10 depicts a block diagram for a more detailed embodiment of FIG. 4

0256wherein the tag-switch controller includes inteφosed demultiplexers, and header

0257detectors and fast memory; and

0258FIG. 11 is a flow diagram for the processing effected by each tag-switch

0259controller of FIG. 10.

0260To facilitate understanding, identical reference numerals have been used,

0261where possible, to designate identical elements that are common to the figures.

DETAILED DESCRIPTION

0263In order to gain an insight into the fundamental principles in accordance

0264with the present invention as well as to introduce terminology useful in the sequel, an

0265overview is first presented, followed by an elucidation of an illustration embodiment.

0266Overview

0267The present invention relates to a network for realizing low latency, high

0268throughput, and cost-effective bandwidth-on-demand for large blocks of data for NGI

0269applications. Cost-effective and interoperable upgrades to the network are realized by

0270inteφosing portable 'Plug-and-Play' modules on the existing WDM network elements to

0271effect so-called "WDM optical tag switching" or, synonymously, "optical tag switching".

0272The invention impacts both the hardware and software for the NGI network from all perspectives, including architecture, protocol, network management, network elemgnt

0273design, and enabling technologies.

0274As alluded to, the methodology carried out by the network and

0275concomitant circuitry for implementing the network are engendered by a technique called

0276WDM optical tag-switching ~ defined as the dynamic generation of a routing path for a

0277burst duration by an in-band optical signaling header. Data packets are routed through

0278the WDM network using an in-band WDM signaling header for each packet. At a

0279switching node, the signaling header is processed and the header and the data payload (1)

0280may be immediately forwarded through an already existing flow state connection, or (2) a

0281path can be setup for a burst duration to handle the header and the data payload. WDM

0282tag-switching enables highly efficient routing and throughput, and reduces the number of

0283IP-level hops required by keeping the packets routing at the optical level to one hop as

0284managed by the NC&M which creates and maintains routing information.

0285The depiction of FIG. 1 shows the inter-relation between optical layer 120

0286and electrical layer 110 of generic network 100 as provided by intermediate layer 130

0287coupling the optical layer and the electrical layer. Electrical layer 110 is shown, for

0288simplicity, as being composed of two conventional IP routers 111 and 112. Optical layer

0289120 is shown as being composed of network elements or nodes 121-125. Intermediate

0290layer 130 depicts conventional ATM/SONET system 131 coupling IP router 112 to

0291network element 122. Also shown as part of layer 130 is header network 132, which in

0292accordance with the present invention, couples IP router 111 to network element 121.

0293FIG. 1 pictorially illustrates the location of network 132 on a national-scale, transparent

0294WDM-based backbone network with full interoperability and reconfigurability. It is

0295important to emphasize at this point that the elements of FIG. 1 are illustrative of one embodiment in accordance with the present invention; thus, for example, elemental 1

0296may, in another embodiment, be an ATM router or even a switch.

0297Now with reference to FIG. 2, optical layer 120 of FIG. 1 is shown in

0298more detail including the basic technique, in accordance with the present invention, for

0299setting up a fast connection in optical network 201, composed of network elements 121-

0300125; the setup uses optical signaling header 210 for the accompanying data payload 211.

0301This technique combines the advantages of circuit-switched based WDM and packet-

0302switched based IP technologies. New signaling information is added in the form of an

0303optical signal header 210 which is carried in-band within each wavelength in the multi -

0304wavelength transport environment. Optical signaling header 210 is a tag containing

0305routing and control information such as the source, destination, priority, and the length of

0306the packet, propagates through optical network 201 preceding data payload 211. Each

0307WDM network element 121-125 senses optical signaling header 210, looks-up a

0308connection table (discussed later), and takes necessary steps such as cross-connections,

0309add, drop, or drop-and-continue. The connection table is constantly updated by

0310continuous communication between NC&M 220 and WDM network elements 121-125.

0311Data payload 211, which follows optical signaling header 210, is routed through a path in

0312each network element (discussed later) as established by the connection. With the

0313arrangement of FIG. 2, there is no need to manage the time delay between optical

0314signaling header 210 and data payload 211, shown by T in FIG. 2, because each network

0315element provides the optical delay needed for the short time required for connection set¬

0316up within each network element via delay on an inteφosed fiber. Moreover, the format

0317and protocol of the data payload is independent of that of the header, that is, for a given

0318network whereas the format and protocol of the header are pre-determined, the format and the protocol of the data payload can be the same as or different from those of the

0319header.

0320Each destination is associated with a preferred path which would

0321minimize 'the cost' - in FIG. 2, the overall path from source 123 to destination 122

0322includes paths 201 and 202 in cascade, both utilizing wavelength WP. This cost is

0323computed based on the total propagation distance, the number of hops, and the traffic

0324load. The preferred wavelength is defaulted to the original wavelength. For example, the

0325preferred wavelength on path 202 is WP. If this preferred path at the default wavelength

0326is already occupied by another packet, then network element 121 quickly decides if there

0327is an available alternate wavelength WA through the same preferred path. This alternate

0328wavelength must be one of the choices offered by the limited wavelength conversion in

0329network element 121. If there is no choice of wavelengths which allows transport of the

0330packet through the most preferred path, the next preferred path is selected (path

0331deflection). For example, in FIG. 2, paths 203 and 204 in cascade may represent the

0332alternative path. At this point, the preferred wavelength will default back to the original

0333wavelength WP. The identical process of looking for an alternate wavelength can

0334proceed if this default wavelength is again already occupied. In FIG. 2, path 203 is an

0335alternative path with the same wavelength WP, and path 204 is an alternate path using

0336alternate wavelength WA. In an unlikely case where there is no combination of path and

0337wavelength deflection can offer transport of the packet, network element 121 will decide

0338to drop the packet of lower priority. In other words, the new packet transport through the

0339preferred path at the originating wavelength takes place by dropping the other packet of

0340the lower priority which is already occupying the preferred path. Network elements 121-125 are augmented with two types of so-callSd

0341'Plug-and-Play' modules to efficiently handle bursty traffic by providing packet

0342switching capabilities to conventional circuit-switched WDM network elements 121-125

0343whereby signaling headers are encoded onto IP packets and are removed when necessary.

0344The first type of 'Plug-and-Play' module, represented by electro-optical

0345element 132 of FIG. 1, is now shown in block diagram form in FIG. 3. Whereas

0346conceptually module 132 is a stand-alone element, in practice, module 132 is integrated

0347with network element 121 as is shown in FIG. 3; module 132 is inteφosed between

0348compliant client interface (CCI) 310 of network element 121 and IP router 111 to encode

0349optical signaling header 210 onto the packets added into the network via header encoder

0350321, and to remove optical signaling header 210 from the packets dropping out of the

0351network via header remover 322.

0352Generally, encoding/removing module 132 is placed where the IP traffic is

0353interfaced into and out of the WDM network, which is between the client interface of the

0354network element and the IP routers. The client interfaces can be either a CCI-type or a

0355non-compliant client interfaces (NCI)-type. At these interfaces, header encoder 321 puts

0356optical header 210 carrying the destination and other information in front of data payload

0357211 as the IP signal is transported into network 201. Optical header 210 is encoded in the

0358optical domain by an optical modulator (discussed later). Signaling header remover 322

0359deletes header 210 from the optical signal dropped via a client interface, and provides an

0360electrical IP packet to IP router 111.

0361More specifically, module 132 accepts the electrical signal from IP router

0362111, converts the electrical signal to a desired compliant wavelength optical signal, and

0363places optical header 210 in front of the entire packet. Module 132 communicates with NC&M 220 and buffers the data before optically converting the data if requested^

0364NC&M 220. Module 132 employs an optical transmitter (discussed later) with the

0365wavelength matched to the client interface wavelength. (As indicated later but instructive

0366to mention here, module 132 is also compatible with NCI 404 of FIG. 4 since the

0367wavelength adaptation occurs in the NCI; however, the bit-rate-compatibility of NCI

0368wavelength adaption and the IP signal with optical headers must be established in

0369advance.)

0370FIG. 4 depicts a second type of 'Plug-and-Play' module, optical element

0371410, which is associated with each WDM network element 121-125, say element 121 for

0372discussion purposes. Module 410 is inteφosed between conventional network element

0373circuit switch controller 420 and conventional switching device 430. Module 410 detects

0374information from each signaling header 210 propagating over any fiber 401-403, as

0375provided to module 410 by tapped fiber paths 404-406. Module 410 functions to achieve

0376very rapid table look-up and fast signaling to switching device 430. Switch controller

0377420 is functionally equivalent to the conventional "craft interface" used for controlling

0378the network elements; however, in this case, the puφose of this switch controller 420 is

0379to accept the circuit-switched signaling from NC&M 220 and determine which control

0380commands are to be sent to tag switch controller 410 based on the priority. Thus, tag

0381switch controller 410 receives circuit-switched control signals from network element

0382circuit switch controller 420, as well as information as derived from each signaling each

0383header 210, and intelligently choose between the circuit-switched and the tag-switched

0384control schemes. The switches (discussed later) comprising switching device 430 also

0385achieve rapid switching. The delay imposed by fibers 415, 416, or 416, which are placed

0386in input paths 401-403 to switching device 430, are such that the delay is larger than the total time it takes to read signaling header 210, to complete a table look-up, and toϋeffect

0387switching. Approximately, a 2 km fiber provides 10 microsecond processing time. The

0388types of WDM network elements represented by elements 121-125 and which encompass

0389switching device 430 include: Wavelength Add-Drop Multiplexers (WADMs);

0390Wavelength Selective Crossconnects (WSXCs); and Wavelength Interchanging

0391Crossconnects (WIXCs) with limited wavelength conversion capabilities.

0392In operation, module 410 taps a small fraction of the optical signals

0393appearing on paths 401-403 in order to detect information in each signaling header 210,

0394and determine the appropriate commands for switching device 430 after looking up the

0395connection table stored in module 410. The fiber delay is placed in paths 401-403 so that

0396the packet having header 210 and payload 211 reaches switching device 430 only after

0397the actual switching occurs. This fiber delay is specific to the delay associated with

0398header detection, table look-up, and switching, and can typically be accomplished in

0399about 10 microseconds with about 2 km fiber delay in fibers 415-417.

0400Since there is no optical-to-electrical, nor electrical -to-optical conversion

0401of data payload 211 at network elements 121-125, the connections are completely

0402transparent. Contrary to IP routing, where a multiplicity of bit-rates and lower-level

0403protocols increases the number of different interfaces required and consequently the cost

0404of the router, routing by WDM tag switching is transparent to bit-rates. By way of

0405illustration, optical routing by network elements 121-125 is able to achieve 1.28 Tb/sec

0406throughput (16x16 cross-connect switching device 430 with 32 wavelengths/fiber at

04072.5Gb/sec per wavelength) which is much larger than any of the current gigabit routers.

0408Each network element 121-125 in combination with NC&M 220 effects a

0409routing protocol which is adaptive; the routing protocol performs the following functions: (a) measures network parameters, such as state of communication lines, estimatedHaffic,

0410delays, capacity utilization, pertinent to the routing strategy; (b) forwards the measured

0411information to NC&M 220 for routing computations; (c) computes of the routing tables

0412at NC&M 220; (d) disseminates the routing tables to each network element 121-125 to

0413have packet routing decisions at each network element. NC&M 220 receives the network

0414parameter information from each network element, and updates the routing tables

0415periodically, then (e) forwards a connection request from an IP router such as element

0416111 to NC&M 220, and (f) forwards routing information from the NC&M 220 to each

0417network element 121-125 to be inputted in optical signaling header 210.

0418Packets are routed through network 201 using the information in signaling

0419header 210 of each packet. When a packet arrives at a network element, signaling header

0420210 is read and either the packet (a) is routed to a new appropriate outbound port chosen

0421according to the tag routing look-up table, or (b) is immediately forwarded through an

0422already existing tag-switching originated connection within the network element. The

0423latter case is referred to as "flow switching" and is supported as part of optical tag-

0424switching; flow switching is used for large volume bursty mode traffic.

0425Tag-switched routing look-up tables are included in network elements

0426121-125 in order to rapidly route the optical packet through the network element

0427whenever a flow switching state is not set-up. The connection set-up request conveyed

0428by optical signaling header 210 is rapidly compared against the tag-switch routing look¬

0429up table within each network element. In some cases, the optimal connections for the

0430most efficient signal routing may already be occupied. The possible connection look up

0431table is also configured to already provide an alternate wavelength assignment or an

0432alternate path to route the signal. Providing a limited number of (at least one) alternative wavelength significantly reduces the blocking probability. The alternative wavelength

0433routing also achieves the same propagation delay and number of hops as the optimal case,

0434and eliminates the difficulties in sequencing multiple packets. The alternate path routing

0435can potentially increase the delay and the number of hops, and the signal-to noise-ratio of

0436the packets are optically monitored to eliminate any possibility of packets being routed

0437through a large number of hops. In the case where a second path or wavelength is not

0438available, contention at an outbound link can be settled on a first-come, first-serve basis

0439or on a priority basis. The information is presented to a regular IP router and then is

0440reviewed by higher layer protocols, using retransmission when necessary.

0441Routing Example

0442An illustrative WDM circuit-switched backbone network 500 for

0443communicating packets among end-users in certain large cities in the United States is

0444shown in pictorial form in FIG. 5 ~ network 500 is first discussed in terms of its

0445conventional operation, that is, before the overlay of WDM optical tag switching in

0446accordance with the present invention is presented.

0447With reference to FIG. 5, it is supposed that New York City is served by

0448network element 501, Chicago is served by network element 502, ..., Los Angeles is

0449served by network element 504, ..., and Minneapolis by network element 507. (Network

0450elements may also be referred to a nodes in the sequel.) Moreover, NC&M 220 has

0451logical connections (shown by dashed lines, such as channel 221 to network element 501

0452and channel 222 to network element 507) to all network elements 501-507 via physical

0453layer optical supervisory channels; there is continuous communication among NC&M

0454220 and network elements 501-507. NC&M 220 periodically requests and receives information about: (a) the general state of each network element (e.g., whether it is?

0455operational or shut down for an emergency); (b) the optical wavelengths provided by

0456each network element (e.g., network element 501 is shown as being served by optical

0457fiber medium 531 having wavelength Wl and optical fiber medium 532 having

0458wavelength W2 which connect to network elements 502 (Chicago) and 505 (Boston),

0459respectively); and (c) the ports which are served by the wavelengths (e.g., port 510 of

0460element 501 is associated with an incoming client interface conveying packet 520, port

0461511 is associated with Wl and port 512 is associated with W2, whereas port 513 of

0462element 502 is associated with Wl).

0463Thus, NC&M 220 has stored at any instant the global information

0464necessary to formulate routes to carry the incoming packet traffic by the network

0465elements. Accordingly, periodically NC&M 220 determines the routing information in

0466the form of, for example, global routing tables, and downloads the global routing tables

0467to each of the elements using supervisory channels 221, 222, .... The global routing

0468tables configure the ports of the network elements to create certain communication links.

0469For example, NC&M 220 may determine, based upon traffic demand and statistics, that a

0470fiber optic link from New York City to Los Angeles (network elements 501 and 504,

0471respectively) is presently required, and the link will be composed, in series, of: Wl

0472coupling port 511 of element 501 to port 513 in network element 502; Wl coupling port

0473514 of element 502 to port 515 of element 503; and W2 coupling port 516 of element 503

0474to port 517 of element 504. Then, input packet 520 incoming to network element 501

0475(New York City) and having a destination of network element 504 (Los Angeles) is

0476immediately routed over this established link. At network element 504, the propagated

0477packet is delivered as output packet 521 via client interface port 518. In a similar manner, a dedicated path between elements 506 and 5Q 2*(St.

0478Louis and Minneapolis, respectively) is shown as established using W2 between network

0479elements 506 and 502, and W3 between elements 502 and 507.

0480Links generated in this manner — as based upon the global routing tables -

0481- are characterized by their rigidity, that is, it takes several seconds for NC&M 220 to

0482determine the connections to establish the links, to download the connectivity

0483information for the links, and establish the input and output ports for each network

0484element. Each link has characteristics of a circuit-switched connection, that is, it is

0485basically a permanent connection or a dedicated path or "pipe" for long intervals, and

0486only NC&M 220 can tear down and re-establish a link in normal operation. The benefit

0487of such a dedicated path is that traffic having an origin and a destination which maps into

0488an already-established dedicated path can be immediately routed without the need for any

0489set-up. On the other hand, the dedicated path can be, and most often is, inefficient in the

0490sense that the dedicated path may be only used a small percentage of the time (e.g., 20%-

049150% over the set-up period). Moreover, switching device 430 (see FIG. 4) embedded in

0492each network element which interconnects input and output ports has only a finite

0493number of input/output ports. If the above scenario is changed so that link from St. Louis

0494to Minneapolis is required and a port already assigned to the New York to Los Angeles

0495link is to be used (e.g., port 514 of network element 502), then there is a time delay until

0496NC&M 220 can respond and alter the global routing tables accordingly.

0497Now the example is expanded so that the subject matter in accordance

0498with the principles of the present invention is overlaid on the above description. First, a

0499parameter called the "tag-switched state" is introduced and its use in routing is discussed; then, in the next paragraph, the manner of generating the tag-switch state is elucidated.

0500The tag-switch state engenders optical tag switching.

0501NC&M 220 is further arranged so that it may assign the tag-switch state to

0502each packet incoming to a network element from a client interface ~ the tag-switch state

0503is appended by Plug & Play module 132 and, for the puφoses of the present discussion,

0504the tag-switch state is commensurate with header 210 (see FIG. 2). The tag-switch state

0505is computed by NC&M 220 and downloaded to each network element 501-507 in the

0506form of a local routing table. With reference to FIG. 6, there is shown network element

0507501 and its embedded switch 601 in pictorial form. Also shown is incoming optical fiber

0508602, with delay loop 603, carrying packet 620 composed of header 210 and payload 211

0509- payload 211 in this case is packet 520 from FIG. 5. Fiber 6022 delivers a delayed

0510version of packet 620 to network element 501. Also, a portion of the light energy

0511appearing on fiber 602 is tapped via fiber 6021 and inputted to optical module 410 which

0512processes the incoming packet 620 to detect header 210 — header 210 for packet 620 is

0513shown as being composed of the tag-switch state ' 11101011000', identified by reference

0514numeral 615. Also shown in FIG. 6 is local look-up table 610, being composed of two

0515columns, namely, "Tag-Switch State" (column 611), and "Local Address" (column 612).

0516The particular tag-switch state for packet 620 is cross-referenced in look-up table 610 to

0517determine the routing of the incoming packet. In this case, the tag-switch state for packet

0518620 is the entry in the fourth row of look-up table 610. The local switch address

0519corresponding to this tag-switch state is "0111", which is inteφreted as follows: the first

0520two binary digits indicate the incoming port, and the second two binary digits indicate the

0521output port. In this case, for the exemplary four-input, four-output switch, the incoming

0522packet is to be routed from input port "01" to output port "11", so switch 601 is switched accordingly (as shown). After the delay provided by fiber delay 603, the incomings

0523packet on fiber 6022 is propagated onto fiber 604 via switch 601.

0524The foregoing description of tag-switch state indicates how it is used. The

0525manner of generating the tag-switch state is now considered. NC&M 220, again on a

0526periodic basis, compiles a set of local look-up tables for routing/switching the packet

0527through each corresponding network element (such as table 610 for network element

0528501), and each look-up table is then downloaded to the corresponding network element.

0529The generation of each look-up table takes into account NC&M 220's global knowledge

0530of the network 500. For instance, if incoming packet 620 to network 501 is destined for

0531network 504 (again, New York to Los Angeles), if port 510 is associated with incoming

0532port "01" and serves fiber 602, and if outgoing port 511 is associated with outgoing port

0533"11" and serves fiber 604, then NC&M 220 is able to generate the appropriate entry in

0534look-up table 610 (namely, the fourth row) and download table 610 to network element

0535510. Now, when packet 520 is processed by electro-optical module 132 so as to add

0536header 210 to packet 520 to create augmented packet 620, NC&M 220's knowledge of

0537the downloaded local routing tables as well as the knowledge of the destination address

0538embedded in packet 520 as obtained via module 132 enables NC&M 220 to instruct

0539module 132 to add the appropriate tag-switch state as header 210 — in this case

0540' 11101011000'.

0541It can be readily appreciated that processing a packet using the tag-switch

0542state parameter is bursty in nature, that is, after switch 601 is set-up to handle the

0543incoming tag-switch state, switch 601 may be returned to its state prior to processing the

0544flow state. For example, switch 601 may have interconnected input port '01 ' to output

0545port ' 10' prior to the arrival of packet 620, and it may be returned to the '0110' state after processing (as determined, for example, by a packet trailer). Of course, it may be_^that the

0546circuit-switched path is identical to the tag-switch state path, in which case there is no

0547need to even modify the local route through switch 601 for processing the tag-switch

0548state. However, if it is necessary to temporarily alter switch 601, the underlying circuit-

0549switched traffic, if any, can be re-routed or re-sent.

0550As discussed so far, tag switching allows destination oriented routing of

0551packets without a need for the network elements to examine the entire data packets. New

0552signaling information ~ the tag ~ is added in the form of optical signal header 210 which

0553is carried in-band within each wavelength in the multi-wavelength transport environment.

0554This tag switching normally occurs on a packet-by-packet basis. Typically, however, a

0555large number of packets will be sequentially transported towards the same destination.

0556This is especially true for bursty data where a large block of data is segmented in many

0557packets for transport. In such cases, it is inefficient for each particular network element

0558to carefully examine each tag and decide on the routing path. Rather, it is more effective

0559to set up a "virtual circuit" from the source to the destination. Header 210 of each packet

0560will only inform continuation or ending of the virtual circuit, referred to as a flow state

0561connection. Such an end-to-end flow state path is established, and the plug-and-play

0562modules in the network elements will not disrupt such flow state connections until

0563disconnection is needed. The disconnection will take place if such a sequence of packets

0564have come to an end or another packet of much higher priority requests disruption of this

0565flow state connection.

0566The priority aspect of the present invention is also shown with respect to

0567FIG. 6. The local look-up table has a "priority level" (column 613) which sets forth the

0568priority assigned to the tag-switching state. Also, header 210 has appended priority data shown as the number '2' (reference numeral 616). Both the fourth and fifth row iπ3he

0569"tag-switch state" column 611 of table 610 have a local address of ' 0111. ' If an earlier

0570data packet used the entry in the fifth row to establish, for example, a virtual circuit or

0571flow switching state, and the now another packet is processed as per the fourth row of

0572column 611, the higher priority data ('2' versus '4', with ' 1 ' being the highest) has

0573precedent, and the virtual circuit would be terminated.

0574Detailed Illustrative Embodiment

0575In order to achieve ultra-low latency IP over WDM tag switching,

0576processing of the optical header at each optical switch must be kept to a minimum during

0577the actual transmission of the optical packet. To achieve this end, a new signaling

0578architecture and packet transmission protocol for performing optical WDM tag switching

0579is introduced.

0580The signaling and packet transmission protocols decouple the slow and

0581complex IP routing functions from the ultra-fast WDM switching functions. This

0582decoupling is achieved via the setup up an end-to-end routing path which needs to be

0583performed very infrequently. To send IP packets from a source to a destination, the

0584following steps are executed:

0585(a) End-to-end routing path setup, where the IP layer software invokes the

0586signaling protocol between the network elements and the NC&M to set up an end-to-end

0587routing path for the IP packets. This step will also configure the WDM network elements

0588along the routing path to support subsequent packet forwarding. The tags for optical tag

0589switching to be inserted in the optical headers during actual packet transmission are also

0590determined. (b) Optical packet transmission, where the arrival of the optical packet triggers the

0591local header processing which among other things looks up the output port for forwarding

0592the packet on to the next hop based on the optical tag inside the optical header.

0593Although routing path setup involves invoking the routing function which

0594is generally a slow and complicated procedure, it is performed prior to packet

0595transmission handling, and hence it is not in the critical path that determines transmission

0596latency.

0597Routing Path Setup

0598During routing path setup, the internal connection table of a WDM packet

0599switch will be augmented with a tag-switch look-up table, and contains the pertinent

0600packet forwarding information. In particular, in the interest of achieving ultra-low

0601latency and hardware simplicity, the inventive scheme produces tag-switch states that

0602remain constant along the flow path. For example, tag-switch assignments include the

0603following techniques:

0604(1) Destination-based flow tag assignment — In this scheme the destination, e.g. a

0605suitable destination IP address prefix can be used as the tag-switch state in next hop look¬

0606up. In addition to having no need to modify the optical header, the same header can be

0607used in the event of deflection routing.

0608(2) Route-based flow tag assignment ~ In this scheme the tag-switch state

0609assigned refers to the end-to-end route that is computed dynamically at the tag-switch

0610state setup phase. The advantage of this scheme is that it can be specialized to meet the

0611Quality-of-Service requirements for each individual tag-switched states. Switching Conflict Resolution ^

0612The present-day lack of a viable optical buffer technology implies that

0613conventional buffering techniques cannot be used to handle switching conflicts. As

0614previously described, the invention embodiment utilizes fixed delay implemented by an

0615optical fiber to allow switching to occur during this time delay, but not to achieve

0616contention resolution as electrical buffers do in conventional IP routers. To resolve

0617switching contentions, in accordance with the present invention, the following three

0618methods are used:

0619(a) Limited wavelength interchange ~ where a packet is routed through the same

0620path but at a different wavelength. Since this wavelength conversion is utilized just to

0621avoid the contention, it is not necessary that the network elements must possess the

0622capability of converting to any of the entire wavelength channels. Rather, it is sufficient

0623if they can convert some of the entire wavelength channels. This wavelength conversion

0624converts both the signaling header and the data payload. Care must be taken to prevent a

0625packet from undergoing too many wavelength conversions which will result in poor

0626signal fidelity. A possible policy is to allow only one conversion, which and can easily

0627be enforced by encoding the original wavelength in the optical header. This way an

0628intermediate WDM switch will allow conversion if and only if it is carried on its original

0629wavelength.

0630(b) Limited deflection routing ~ where a packet may be deflected to a

0631neighboring switching node from which it can be forwarded towards its destination. Care

0632again must be taken to prevent a packet from being repeatedly deflected, thereby causing

0633signal degradation, as well as wasting network bandwidth. A solution scheme is to record a "timestamp" field in the optical header, and allow defections to proceed ifand

0634only if the recorded timestamp is no older than a maximum limit.

0635(c) Prioritized packet preemption — where a newly arrived packet may preempt a

0636currently transmitting packet if the arriving packet has a higher priority. The objective is

0637to guarantee fairness to all packets so that eventually a retransmitted packet can be

0638guaranteed delivery. In this scheme, each packet again has a timestamp field recorded in

0639its optical header, and older packets have higher priority compared to newer packets.

0640Furthermore a retransmitted packet assumes the timestamp of the original packet. This

0641way, as a packet "ages," it increases in priority, and will eventually be able to preempt its

0642way towards its destination if necessary.

0643It is noted that in all these schemes the optical header always remains

0644constant as it moves around in the network. This is consistent with the desire to keep the

0645optical switching hardware fast and simple. It is also possible to consider combinations

0646of these schemes.

0647Routing Protocol

0648For a network the size of the NGI, centralized routing decisions are quite

0649infeasible, so the approach needs to be generalized to distributed decision making.

0650Hierarchical addressing and routing are used as in the case of IP routing. When a new

0651connection is requested, NC&M 220 decides whether a WDM path is provisioned for this

0652(source, destination) pair within the WDM-based network. If it is, the packets are

0653immediately sent out on that (one-hop IP-level) path. If no such path is provisioned,

0654NC&M 220 decides on an initial outbound link for the first WDM network element and a

0655wavelength to carry the new traffic. This decision is based on the rest of the connections in the network at the time the new connection was requested. NC&M 220 then use?

0656signaling, through an appropriate protocol, to transfer the relevant information to the

0657initial WDM network element to be placed in the signaling header. After the initial

0658outbound link is determined, the rest of the routing decisions are taken at the individual

0659NE's according to the optical signaling header information. This method ensures that the

0660routing tables at each switching node and the signaling header processing requirements

0661are kept relatively small. It also enables the network to scale easily in terms of switching

0662nodes and network users. It is noted, too, that multiple WDM subnetworks can be

0663interconnected together and each subnetwork will have its own NC&M.

0664When a path is decided upon, within a WDM NE, the optical switches can

0665be set in that state (i) for the duration of each packet through the node and then revert

0666back to the default state (called optical tag-switching), or (ii) for a finite, small amount of

0667time (called flow switching). The former case performs routing on a regular packet-by-

0668packet basis. The system resources are dedicated only when there is information to be

0669sent and at the conclusion of the packet, these resources are available for assignment to

0670another packet. The latter case is used for large volume bursty mode traffic. In this case,

0671the WDM NE only has to read a flow state tag from the optical signaling header of

0672subsequent packets arriving at the NE to be sure such packet is bound for the same

0673destination, without the need to switch the switching device, and forward the payload

0674through the already existing connection through the NE as previously established by the

0675optical tag-switching.

0676The packets are self-routed through the network using the information in

0677the signaling header of each packet. When a packet arrives at a switching node, the

0678signaling header is read and either the packet is forwarded immediately through an already existing flow state connection or a new appropriate outbound port is chosen*

0679according to the routing table. Routing tables in each node exist for each wavelength. If

0680the packet cannot follow the selected outbound port because of contention with another

0681packet (the selected outbound fiber is not free), the routing scheme will try to allocate a

0682different wavelength for the same outbound port (and consequently the signal will

0683undergo wavelength translation within the switching node). If no other eligible

0684wavelength can be used for the chosen outbound port, a different outbound port may be

0685chosen from another table, which lists secondary (in terms of preference) outbound links.

0686This routing protocol of the inventive technique is similar to the deflection

0687routing scheme (recall the Background Section), where the session is deflected to some

0688other outbound link (in terms of preference) if the preferred path cannot be followed. The

0689packet is not allowed to be continuously deflected. In traditional routing protocols, a hop

0690count is used to block a session after a specified number of hops. In the new scheme, in

0691case no header regeneration is allowed at the switching nodes, then the hop count

0692technique cannot be used. Alternatively, the optical signaling header characteristics (i.e.,

0693the signaling header's SNR) can be looked upon to decide whether a packet should be

0694dropped.

0695IP Routing Algorithm in WDM layer

0696The technique used by NC&M 220 to determine the routing tables is

0697based upon shortest path algorithms that route the packets from source to destination over

0698the path of least cost. Specific cost criteria on each route, such as length, capacity

0699utilization, hop count, or average packet delay can be used for different networks. The

0700objective of the routing function is to have good performance (for example in terms of low average delay through the network) while maintaining high throughput. Minimum

0701cost spanning trees are generated having a different node as a root at each time, and the

0702information obtained by these trees can then be used to set-up the routing tables at each

0703switching node. If deflection routing as outlined above is implemented, the k-shortest

0704path approach can be used to exploit the multiplicity of potential routing paths. This

0705technique finds more than one shortest path, with the paths ranked in order of cost. This

0706information can be inputted into the switching node routing tables, so that the outbound

0707link corresponding to the minimum cost path is considered first, and the links

0708corresponding to larger cost paths are inputted in secondary routing tables that are used to

0709implement deflection routing.

0710Description of Plug-and-Play Modules

0711The present invention is based upon two types of Plug-and-Play modules

0712to be attached to the WDM network elements. Introduction of these Plug-and-Play

0713modules add optical tag switching capability to the existing circuit-switched network

0714elements.

0715In FIG. 3, both header encoder 321 and header remover 322 were shown

0716in high-level block diagram form; FIGS. 7 and 8 show, respectively, a more detailed

0717schematic for both encoder 321 and remover 322.

0718In FIG. 7, IP packets or datagrams are processed in microprocessor 710

0719which generates each optical signaling header 210 for tag switching. Optical signaling

0720header 210 and the original IP packet 211 are emitted from microprocessor 710 at

0721baseband. Signaling header 210 is mixed in RF mixer 720 utilizing local oscillator 730.

0722Both the mixed header from mixer 720 and the original packet 211 are combined in combiner 740 and, in turn, the output of combiner 740 is encoded to an optical ^

0723wavelength channel via optical modulator 760 having laser 750 as a source of

0724modulation.

0725In FIG. 8, the optical channel dropping out of a network element is

0726detected by photodetector 810 and is electrically amplified by amplifier 820. Normally,

0727both photodetector 810 and the amplifier 820 have a frequency response covering only

0728the data payload but not the optical signaling header RF carrier frequency provided by

0729local oscillator 730. Low-pass-filter 830 further filters out any residual RF carriers. The

0730output of filter 830 is essentially the original IP packet sent out by the originating IP

0731router from the originating network element which has been transported through the

0732network and is received by another IP router at another network element.

0733Block diagram 900 of FIG. 9 depicts the elements for the detection

0734process effected by Plug-and-Play module 410 of FIG. 4 to convert optical signal 901,

0735which carries both tag-switching signaling header 210 and the data payload 211, into

0736baseband electrical signaling header 902. Initially, optical signal 901 is detected by

0737photodetector 910; the output of photodetector 910 is amplified by amplifier 920 and

0738filtered by high-pass filter 930 to retain only the high frequency components which carry

0739optical signaling header 210. RF splitter 940 provides a signal to local oscillator 950,

0740which includes feedback locking. The signal from local oscillator 950 and the signal

0741from splitter 940 are mixed in mixer 960, that is, the high frequency carrier is subtracted

0742from the output of filter 920 to leave only the information on tag-switching signaling

0743header 210. In this process, local oscillator 950 with feedback locking is utilized to

0744produce the local oscillation with the exact frequency, phase, and amplitude, so that the

0745high frequency component is nulled during the mixing of this local oscillator signal and the tag-switching signaling header with a high-frequency carrier. Low-pass filter 9 0,

0746which is coupled to the output of mixer 960, delivers baseband signaling header 210 as

0747electrical output signal 902.

0748The circuit diagram of FIG. 10 shows an example of a more detailed

0749embodiment of FIG. 4. In FIG. 10, each header detector 1010, 1020, ..., 1030, ..., or

07501040 processes information from each wavelength composing the optical inputs arriving

0751on paths 1001, 1002, 1003, and 1004 as processed by demultiplexers 1005, 1006, 1007,

0752and 1008, respectively; each demultiplexer is exemplified by the circuit 900 of FIG. 9.

0753The processed information is grouped for each wavelength. Thus, for example, fast

0754memory 1021 receives as inputs, for a given wavelength, the signals appearing on lead

07551011 from header detector 1010, ..., and lead 1034 from header detector 1030. Each fast

0756memory 1021-1024, such as a content-addressable memory, serves as an input to a

0757corresponding tag switch controller 1031-1034. Each tag switch controller 1031-1034

0758also receives circuit-switched control signals from network element switch controller 420

0759of FIG. 4. Each tag switch controller intelligently chooses between the circuit switched

0760control as provided by controller 420 and the tag switched information supplied by its

0761corresponding fast memory to provide appropriate control signals the switching device

0762430 of FIG. 4.

0763Flow diagram 1100 of FIG. 11 is representative of the processing effected

0764by each tag-switch controller 1031 - 1034. Using tag-switch controller 1031 as exemplary,

0765inputs from circuit-switched controller 420 and inputs from fast memory 1021 are

0766monitored, as carried out by processing block 1110. If no inputs are received from fast

0767memory 1021, then incoming packets are circuit-switched via circuit-switched controller

0768420. Decision block 1120 is used to determine if there are any inputs from fast memory 1021. If there are inputs, then processing block 1130 is invoked so that tag-switch^

0769controller 1031 can determine from the fast memory inputs the required state of

0770switching device 430. Then processing block 1160 is invoked to transmit control signals

0771from tag-switch controller 1031 to control switching device 430. If there are no fast

0772memory inputs, then the decision block 1140 is invoked to determine if there are any

0773inputs from circuit-switched controller 1140. If there are inputs from circuit-switched

0774controller 420, then processing by block 1150 is carried out so that tag-switch controller

07751031 determines from the inputs of circuit-switched controller 420 the required state of

0776switching device 430. Processing block 1160 is again invoked by the results of

0777processing block 1150. If there are no present inputs from circuit-switched controller

07781140 or upon completion of procession block 1160, control is returned to processing

0779block 1110.

0780By way of reiteration, optical tag-switching flexibly handles all types of

0781traffic: high volume burst, low volume burst, and circuit switched traffic. This occurs by

0782interworking of two-layer protocols of the tag-switched network control. Thus, the

0783distributed switching control rapidly senses signaling headers and routes packets to

0784appropriate destinations. When a long stream of packets reach the network element with

0785the same destination, the distributed switching control establishes a flow switching

0786connection and the entire stream of the packets are forwarded through the newly

0787established connections.

0788A tag switching method scales graciously with the number of wavelengths

0789and the number of nodes. This results from the fact that the distributed nodes process

0790multi-wavelength signaling information in parallel and that these nodes incoφorate predicted switching delay in the form of fiber delay line. Moreover, the tag switching

0791utilizes path deflection and wavelength conversion for contention resolution.

0792Optical Technology

0793Optical technologies span a number of important aspects realizing the

0794present invention. These include optical header technology, optical multiplexing

0795technology, optical switching technology, and wavelength conversion technology.

0796(a) Optical Header Technology

0797Optical header technology includes optical header encoding and optical

0798header removal as discussed with respect to FIGS 3 and 4. In effect, optical header 210

0799serves as a signaling messenger to the network elements informing the network elements

0800of the destination, the source, and the length of the packet. Header 210 is displaced in

0801time compared to the actual data payload. This allows the data payload to have any data

0802rates/protocols or formats.

0803As previously described with respect to FIGS. 7and 8, the header encoding

0804is subcarrier based. This method allows header 210 to be separated in modulation

0805frequency so that header detection can be relatively simple. Header 210, which precedes

0806the data payload in the time domain, also has higher frequency carrier than the highest

0807data rate. This allows reading of header 210, and eventually removal of header 210

0808without affecting the data payload.

0809(b) Optical Multiplexing Technology

0810Optical multiplexing may illustratively be implemented using the known

0811silica arrayed waveguide grating structure. This waveguide grating structure has a number of unique advantages including: low cost, scalability, low loss, uniformity nd

0812compactness.

0813(c) Optical Switching Technology

0814Fast optical switches are essential to achieving packet routing without

0815requiring excessively long fiber delay as a buffer.

0816Micromachined Electro Mechanical Switches offer the best combination

0817of the desirable characteristics: scalability, low loss, polarization insensitivity, fast

0818switching, and robust operation. Recently reported result on the MEM based Optical

0819Add-Drop Switch achieved 9 microsecond switching time

0820(d) Wavelength Conversion Technology

0821Wavelength conversion is resolves packet contention without requiring

0822path deflection or packet buffering. Both path deflection and packet buffering cast the

0823danger of skewing the sequences of a series of packets. In addition, the packet buffering

0824is limited in duration as well as in capacity, and often requires non-transparent methods.

0825Wavelength conversion, on the other hand, resolves the blocking by transmitting at an

0826alternate wavelength through the same path, resulting in the identical delay.

0827Illustratively, a WSXC with a limited wavelength conversion capability is deployed.

0828Although various embodiments which incoφorate the teachings of the

0829present invention have been shown and described in detail herein, those skilled in the art

0830can readily devise many other varied embodiments that still incoφorate these teachings.