DE10248059A1 - Optical add/drop multiplexer has controlled group switch unit for connecting group multiplex signal of channel multiplexer unit to one or other optical group wavelength multiplexer/demultiplexer unit - Google Patents

Abstract

The device has two optical group wavelength multiplexer/demultiplexer units with some optical group multiplex signal ports of each unit fed to the other unit and at least one channel multiplexer unit that multiplexes useful signals from optical transceiver units into an optical group multiplex signal. The group multiplex signal of the channel multiplexer unit can be connected to either unit by a group switch unit controlled by a control unit. The device (9) has first and second optical group wavelength multiplexer/demultiplexer units (3) with some optical group multiplex signal ports (3a,3b) of each unit fed to the other unit and at least one channel multiplexer unit that multiplexes the useful signals from optical transceiver units into an optical group multiplex signal. The group multiplex signal of the channel multiplexer unit can be connected to one or other unit by a group switch unit (13) controlled by a control unit. AN Independent claim is also included for the following: (a) a self-healing ring structure for optical messaging .

Description

The invention relates to an optical Add / drop multiplexer for implementing an optical wavelength division multiplex communication ring network and a ring structure using such an optical Add / drop multiplexer.

The transmission bandwidth more modern Telecommunications networks are growing dramatically today. Service providers offer data and voice services forever more customers based on the same - typically point-to-point or ring-shaped - network infrastructure on. With an increasing number of participants and increasing network utilization grows also the security requirements of the network. Especially inside of the densely populated urban area is the probability of failure for fiber optic routes very high. While Comparatively errors in transmission devices can be fixed quickly The downtimes of fiber optic cables are often in the range of several days.

The SDH technology currently offers established standard transmission technology all national Service providers have certain protection mechanisms that are capable are to switch individual SDH services to alternative ways. As a general rule is the transmission capacity of the SDH Technology limited by the data rates that can be implemented today. To comes that customers increasingly use data services based on non-SDH Applications are in demand. DWDM Technik offers one here Solution perspective.

In a DWDM network, optical Multiplexing wavelengths generated virtual transmission channels. Each of these wavelength channels can one or more SDH services but also other applications such as for example IP, video or SAN applications. A DWDM based Network is characterized in comparison to a pure SDH network increased capacity and application transparency. Because a DWDM network usually cannot access the protection mechanisms of SDH technology together with the DWDM technology, new protection mechanisms are also necessary introduce.

Because DWDM technology today on protocol-transparent transmission based on so-called transponder or channel cards stand for DWDM protection mechanisms principally exclusively Switchover criteria available, use the parameters of the optical layer. Typical switching criteria are "Loss Of Light" (LOL) and "Loss Of Clock" (LOC). That is, light level or light modulation at a port of the system to be defined have disappeared. As a result, the traffic that goes to this Port corresponds, switched to an alternative route.

• – Um Sprachdienste mit ausreichender Qualität übertragen zu können, müssen Schaltzeiten im Störfall ausreichend klein sein, beispielsweise kleiner als 50 ms. Standard-Betriebssysteme können derart schnelle Signalverarbeitung nicht gewährleisten und scheiden daher für eine Steuerung der Protection-Schaltung aus.
• – Es muss zu jeder Zeit gewährleistet sein, dass die „Protection" Trasse zur Übertragung bereit ist. Außerdem muss nach erfolgter Protection Schaltung detektierbar sein, wann die „Working" Trasse wieder hergestellt ist. Diese Randbedingung erfordert eine permanente Beleuchtung beider Trassen entweder mit Nutzsignalen oder mit einem Pilotsignal.
• – Aus Sicherheitsgründen muss die aus einem offenen Port emittierbare Leistung beschränkt werden. In der Praxis soll in der Regel die Laserklasse 1 erfüllt werden. Die maximal zulässige Lichtleistung beträgt hierfür bei einer Wellenlänge von 1550 nm etwa 10 mW. Ein DWDM System mit seinen zahlreichen Wellenlängen verarbeitet auf der Line Seite in der Regel höhere Lichtpegel. Um den noch Laserklasse 1 zu wahren, muss gewährleistet sei, dass keine offenen Ports vorhanden sind. Auch eine verletzte Fasertrasse gilt als offener Port. Fällt eine Trasse aus, muss daher der Protection Mechanismus den Verkehr auf dieser Trasse aktiv ausschalten. Die Forderung nach Laserklasse 1 tauglichem Gerät verlangt also, dass der Protection Mechanismus aktiv in Senderichtung schaltet. Die Forderung nach Laserklasse 1 ist dabei nicht zwingend, vereinfacht aber den Betrieb des Systems erheblich.

In principle, every protection concept must meet three important criteria in order to be used successfully:

• - In order to be able to transmit voice services with sufficient quality, switching times in the event of a fault must be sufficiently short, for example less than 50 ms. Standard operating systems cannot guarantee such fast signal processing and are therefore ruled out for controlling the protection circuit.
• - It must be guaranteed at all times that the "protection" route is ready for transmission. In addition, after the protection circuit has been completed, it must be possible to detect when the "working" route is restored. This boundary condition requires permanent lighting of both routes either with useful signals or with a pilot signal.
• - For security reasons, the power that can be emitted from an open port must be limited. In practice, laser class 1 should generally be met. The maximum permissible light output for this is about 10 mW at a wavelength of 1550 nm. A DWDM system with its numerous wavelengths generally processes higher light levels on the line side. In order to maintain laser class 1, it must be ensured that there are no open ports. An injured fiber route is also considered an open port. If a route fails, the protection mechanism must therefore actively switch off the traffic on this route. The requirement for a laser class 1 device requires that the protection mechanism actively switches in the direction of transmission. The requirement for laser class 1 is not mandatory, but it simplifies the operation of the system considerably.

Typically, in a DWDM System channels grouped together. Accordingly, there are usually two filter stages: a group and a channel filter stage. A Group typically includes four or eight channels. The applications are finally on wavelength-specific Channel cards terminated. Here is an optically electrical or electrical / optical conversion of the signals.

The following protection methods are most common for DWDM systems today:
In the case of a point-to-point line protection system, only the fiber infrastructure is designed redundantly. An optical switch switches to the alternative route in the event of a fiber fault. Filters, channel cards and the switch itself are not protected. This process is characterized by low equipment costs. Protection of the fiber infrastructure makes sense above all because the repair times are typically special here are long. The requirement for a laser class 1-compatible scheme requires switching in the direction of transmission. In order to be able to monitor the protection route at the same time, an additional so-called pilot laser is required, which is superimposed on the data signals and permanently illuminates both routes. Typical optical switches for point-to-point line protection that meet the three requirements mentioned above therefore work with pilot lasers.

With a point-to-point path protection per wavelength system fiber infrastructure, filters and channel cards are redundant. The equivalent circuit takes place at the level of individual wavelength paths through communication between two channel cards. Since there is no switch in the signal path, there is also no single Point of Failure ". The procedure ensures extremely high network availability with very high equipment costs.

DWDM ring networks are usually implemented as bi-directional two-fiber rings (Optical Bidirectional Line Switched Ring - O-BLSR) with path protection. 1 shows the example of a 4-node ring network 1 the fully redundant configuration of the entire path. group filter 3 , Channel filter 5 and channel cards 7 are configured redundantly in each node for OST and WEST direction. In the event of a fault, the channel cards switch to the other direction.

That through the add / drop multiplexer 9 formed nodes of the ring network 1 offer "pass through" and "add / drop" functionality for different proportions of traffic. As a rule, certain wavelength groups are assigned to the different traffic relationships of a DWDM ring. The reason for this lies in the optical parameters: The main task of the filters of a DWDM system is to isolate wavelengths as sharply as possible. Two adjacent wavelengths experience a certain finite isolation in a channel filter. The isolation between non-adjacent wavelengths is higher. The isolation between wavelengths belonging to different groups is particularly high. All applications that go through the same geometric traffic relationship within a ring (e.g. as in 1 of knots 1 by knot 3 ), have approximately the same optical power values at all points of the path. The wavelengths of such a traffic relationship are therefore relatively easy to separate. In contrast, wavelengths that belong to different traffic relationships can be difficult to separate from one another due to the possibly very different performance values and must be strongly isolated from one another. It is therefore common to use a wavelength group for only one traffic relationship in network planning. This ensures that wavelengths of different traffic relationships are strongly isolated from each other. Whole groups are therefore terminated or forwarded in a node (add / drop or pass through). Pass-through for a wavelength group can either be purely passive or with simultaneous amplification by an optical amplifier EDFA (Erbium doped Fiber Amplifier).

The protection mechanism of the path protection happens at the channel card level. If a channel card on the WEST side detects LOS or LOC, it switches to a replacement card on the OST side. Switching takes place through direct communication between two cards without a detour via an operating system and is therefore faster than the normally required 50 ms. Loss of a line-side receive signal (LOS of the Remote Rx) also leads to the immediate shutdown of the line-side transmitter (Remote Tx) and thus guarantees laser class 1 , As long as there is no LOS, the transmitters of the working and protection card are switched on, which means permanent monitoring of both routes.

The disadvantage of such a system is the high cost of redundantly providing the channel filter 5 and channel cards 7 ,

Based on this state of the art the invention has for its object an optical add / drop multiplexer for the realization of an optical wavelength division multiplex communication ring network and a ring structure using such an optical To create add / drop multiplexers, with the least possible effort sufficient operational security for the communication ring network guaranteed becomes.

The invention solves this problem with the features of claims 1 or 9.

The invention is based on knowledge from that by switching groups of wavelength channels (ring Group switching, RGS) on the optical layer opposite the full path protection achieved a drastic reduction in hardware expenditure becomes. It leaves show that for usual ring networks A sufficient structure in the metro area, even with such a structure network availability is achievable.

The optical concept of the Ring Group Switching (RGS) is based on a shifting of the switching mechanism behind the first DWDM filter stage. The switching granularity thus fits perfectly with the architecture of typical DWDM rings, since by default traffic relationships are associated with complete groups. A so-called group switch (GS) directs the data traffic of a single group in the OST (WEST) direction. In the event of loss of the OST (WEST) route, WEST (OST) Switched direction.

Only such are preferred Groups of a node that are intended for add / drop. All groups that just pass through a node (pass through) are not switched and continue from OST (WEST) to WEST (OST) connected. While Path Protection in the event of the loss of a route channels Switch individually at RGS is only one switching process per add / drop multiplexer and add / drop Groups necessary.

According to the preferred embodiment According to the invention, the add / drop multiplexer has at least one transceiver unit for one Management channel (Optical Supervisory Channel, OSC) on which for transmission management information within the ring structure. The transfer management information can either be immediately between the Add / drop multiplexers take place or via a network management control unit. So that can to each add / drop multiplexer by means of which the switching operations are effected the information required for at least one group switch unit is transmitted. The relevant add / drop multiplexer can either directly be communicated which switching operation is to be carried out, or only the required information about the state of the transmission paths, wherein the control unit of the add / drop multiplexer itself the necessary switching operations determined and executes.

According to one embodiment of the invention the control unit of each add / drop multiplexer using the management channel and internet protocol (IP) with the Control units of all other add / drop multiplexers and / or the Network management control unit connected.

According to the preferred embodiment carries the Control unit the information on the functionality the east and west transmission line in two Bit of a link vector and transmits the link vector to the rest Add / drop multiplexer and / or the network management control unit. The operations of the Writing / reading the link vector and determining the required ones switching operations from the link vector and the node or add / drop multiplexer known traffic relationships within the ring network in the control unit preferably immediately, i.e. without the influence an operating system.

According to the reading or writing process takes place of the link vector preferably between layer 3 and the MAC Layer 2, the control unit placing the link vector in a proprietary framework between two MAC frames. This allows the required short response time in a simple manner or switchover time after a malfunction has occurred.

Further embodiments of the invention result from the subclaims.

The invention will follow of the figures shown in the drawing. The drawing shows:

1 a schematic block diagram of a bidirectional ring network known structure with four nodes and full point-to-point path protection and

2 is a schematic block diagram of a bidirectional ring network according to the invention with four nodes using the protection by optical switching of wavelength groups.

2 shows schematically the use of optical group switch units for the protection of the data transmission channels in a WDM ring network 1 , The ring network 1 has four nodes A, B, C and D in the illustrated embodiment. Each node is through an add / drop multiplixer 9 formed, whose internal structure in 2 is shown using the example of node A.

The add / drop multiplexer 9 comprises two group filters or wavelength division multiplexer / demultiplexer units 3 whose multiplex ports 3 are each connected to the west or east link of the transmission medium via a monitoring channel filter unit OSF. One or two optical fibers, preferably single-mode fibers, are used as the transmission medium, depending on whether the data traffic is unidirectional via one optical fiber or bidirectional via a single optical fiber. The monitoring channel filter units OSF filter the wavelength of the monitoring channel out of the respectively received multiplex signal, which also contains the individual optical useful channels, and feed this received signal of the monitoring channel to a monitoring channel module OSCM-RGS. The monitoring channel module OSCM-RGS is part of the control unit 11 of the add / drop multiplexer 9 , which also includes the network management control unit NEMI. The monitoring channel module OSCM-RGS feeds the transmission signal of the monitoring channel to the monitoring channel filter units OSF, which couple this transmission signal into the east or west side link or optical fiber. For this purpose, the monitoring channel module has two transceiver cards or monitoring channel cards. The actual core of the monitoring channel module OSCM-RGS is preferably implemented as an FPGA (Field Programmable Gate Array), so that the actions of the OSCM-RGS take place very quickly and are not slowed down by a relatively slow-reacting operating system.

Every link between two nodes is So in every direction by the optical monitoring channel signal concerned transceiver card so that the respective Detect neighboring nodes using the monitoring channel signal can whether the link is functional is or not. As already explained can do this by detecting the "Loss of Light" (LOL) or "Loss of Clock "(LOC).

At the group multiplex signal ports 3b the wavelength division multiplexer / demultiplexer units 3 the demultiplexed group multiplex signals are present. These become the relevant group multiplex signal ports either directly or via optical amplifiers EDFA 3b the other wavelength division multiplexer / demultiplexer units 3 looped through. Only those group multiplex signals that are intended for termination in the relevant add / drop multiplexer each become a switched port 13a respectively. 13b one group switching unit each 13 fed. At the in 2 illustrated embodiment is only a single wavelength group in the add / drop multiplexer 9 of node A terminated.

The group multiplex signal received is from the group switch unit 13 (RGSM) the multiplex port of a channel filter unit 5 fed, which separates the individual, in the example shown four, optical channels of the group multiplex signal and each feeds a transceiver or channel card WCM1 to WCM4.

Conversely, the optical signals sent by the channel cards WCM1 to WCM4 are combined by the channel filter unit to form an optical group multiplex signal and the group switching unit 13 fed. This then carries the group multiplex signal to either the east-side or west-side wavelength division multiplexer / demultiplexer unit 3 to.

The filter units 3 . 5 and OSF can be implemented as purely passive optical couplers, for example AWGs. The group switch unit 13 In addition to a controllable optical 1 × 2 switch, it can also include a unit for determining the required switch state. For this, the group switch unit receives 13 from the monitoring channel module OSCM-RGS of the control unit 11 Information about the status of the individual links in the ring network 1 have, and from the network management unit NEMI the information which traffic relationships within the ring network are actually to be realized. From this, it can then be determined within the group switch unit which switching state has to be assumed. The receive and send paths are always switched at the same time.

The logical concept of use of group switch units is based on optical switching individual traffic relationships of an optical ring network. each Traffic relationship has its own - from the route in Ring dependent - switching criterion. In the following, these switching criteria and the influence of Switching to traffic routing defined generically in the ring. The number of ring nodes N can in principle scale from 2 to any height. In the Practice is not greater than N the number of available standing wavelength groups.

To describe the switching states and Traffic relations the following three sizes are introduced: link vector L, switch matrix S, traffic matrix T.

A ring with N nodes has N optical fiber sections or links between the nodes. Each link can either be uninterrupted ("up") or interrupted ("down"). The N-dimensional link vector characterizes the state of the set of all links in a ring. Each element of the link vector can have the value zero for "down" or the value one for "up". The state L = [0; 1; 1; 1] means, for example, that the direct connection between nodes 1 and 2 has failed.

Physically, the link vector is transported over the monitoring channel, which in 2 is shown dotted. A single ring node has the status of the immediately eastern or immediately western link seen from the node through the monitoring channel arriving from OST or WEST. Each node can therefore describe two bits of the link vector. This functionality of describing and evaluating the link vector is implemented in the monitoring channel module OSCM-RGS.

The monitoring channel typically spans an IP network between the network nodes by means of Ethernet, which connects the network management units NEMI of all nodes to one another. This enables central management of the entire network from a “gateway” node (of course, central management can also be carried out by a separate control unit connected to the nodes). The reading or writing process of the link vector is not found IP layer 3 , but already between layers 3 and MAC layer 2 instead of. The reason for this is that just below layer 3 fixed, defined switching times of less than 50 ms can be realized. The link vector is therefore written into a proprietary frame between two MAC frames. In addition to the link vector, the additional frame contains a start and a stop bit. Depending on the number of nodes in the ring, it has a length of (N + 2) bits. The maximum length of a MAC frame (extended format) can be up to 65536 bits. At a data rate of 125 Mbit / s, this corresponds to a time of less than 1 ms between two link vectors. The transmission area thus achieved Speed of the "switching criterion link vector" is therefore sufficient to switch SDH services.

mit S_ij = 0 für i = j, S_ij = L_i für i < j, S_ji = L _i für i > jThe switch matrix S describes and controls the status of all group switch units in the ring. The switch matrix is ranked (N × N). The base of the switch matrix is spanned by the network nodes' OST and WEST terminals. Ie a group switch unit that has a traffic relationship between the nodes 1 and 2 switches, is represented by the elements S ₂₁ and S ₁₂ . Both elements are always complementary. The state {S ₁₂ = 1; S ₂₁ = 0} means, for example, that all traffic between the nodes 1 and 2 from the node's WEST terminal 1 over the knot 4 and 3 away to the node's OST terminal 2 is running (see also Figure 5). The switch matrix clearly results from the link vector. All diagonal elements are zero, since the group switch units always switch between different nodes of a ring. The extra-diagonal elements have the value either above or below the diagonal.

with S _ij = 0 for i = j, S _ij = L _i for i <j, S _ji = L _i for i> j

The information about the traffic relationship to be switched is supplied by the network management unit NEMI and is part of the configuration data of the management. Through a direct connection to the monitoring channel module OSCM-RGS, the link vector is transmitted with a defined, high time resolution. The link vector becomes the group switch unit 13 the relevant switching bit S _{ij is} generated by a simple arithmetic operation and directly controls the optical switch.

While the connection to the network management unit NEMI is usually via a intrinsic bus and the intrinsic management operating system is established, the connection to the OSCM-RGS is hard-wired and does not take the slow detour through an operating system.

The traffic matrix T describes all traffic in the ring. It has the same basis as the switch matrix. Each entry in the traffic matrix that is different from zero means that - depending on the position of the entry in the matrix - data traffic is exchanged between two nodes. The traffic matrix is first configured by the management system, ie all elements are described in a first step - according to the specified network configuration: T ' ij = T ' ji = {(λ - group) ij, (λ x = Application n A, λ y = Application n B, ...) ij }

Each element contains information about the wavelength group occupied for the corresponding traffic relationship and the exact assignment of the individual wavelength channels with applications. As long as all links of a ring are uninterrupted (links are "up"), each node can be reached via two independent routes: either from the east or from the west. The traffic matrix is therefore not unique. Only the state of the switch matrix provides clear traffic guidance in a ring. In a second step, the elements of the traffic matrix are multiplied by those of the switch matrix:

So there are always two logical ones Steps that define the traffic matrix: This is done in the first step Load the logical traffic relations and in the second step the Alignment with the physical switch positions.

The link vector is transported to all nodes of the ring by means of the monitoring channel. In the monitoring channel module OSCM-RGS, the link vector is isolated from the MAC frames and two bits are overwritten in order to update the state of the links to the west and east of the node. The link vector thus updated is passed on to the group switch unit. The link vector is translated into a switching bit in the group switch unit. If a malfunction occurs in the ring, the group scarf decides unit, whether the corresponding wavelength group must be switched from the OST to the WEST terminal or vice versa. The optical switch of the group switch unit switches the transmission and reception direction in parallel. The network management unit NEMI on the one hand supplies configuration data to the monitoring channel module OSCM-RGS or the group switch unit RGS and on the other hand compares the traffic matrix with the switch matrix. While the link vector and switch matrix directly influence the traffic in the ring, the traffic matrix has no influence on the traffic and is only used to display all data streams in the ring.

Claims (13)

Optical add / drop multiplexer for implementation an optical wavelength division multiplex communication ring network a) with a first (west side) and a second (east side) optical Group wavelength multiplexer / demultiplexer, which each one fed to her Wavelength division multiplexed signal into multiple optical group multiplex signals demultiplex and / or several optical group multiplex signals fed to it multiplex to an outgoing optical wave multiplex signal, b) wherein part of the optical group multiplex signals the group multiplex signal ports the other wavelength division multiplexer / Demultiplexer unit supplied are, c) with at least one channel multiplexer unit, which an optical group multiplex signal can be supplied and what this demultiplexed into the individual optical useful signals and each an optical transceiver unit and which the of the optical transceiver units sent optical useful signals multiplexed an optical group multiplex signal, d) where the group multiplex signal port of the at least one channel multiplexer unit by means of an optical group switch unit either with a Group multiplex signal port of the first or second optical group wavelength division multiplexer / demultiplexer unit is connectable, and e) with a control unit for control the at least one optical group switch unit. Add / drop multiplexer according to claim 1, characterized in that the control unit is malfunctioning east and west transmission line recorded and this information for controlling the at least one Group switch unit used. Add / drop multiplexer according to claim 2, characterized in that the control unit interprets the relevant transmission link as disturbed, if there is no optical reception power or no modulation an optical carrier signal is detected. Add / drop multiplexer according to one of the preceding Expectations, characterized in that at least one transceiver unit for one Management channel is provided, which for the transmission of management information serves within the ring structure. Add / drop multiplexer according to claim 4, characterized in that a separate wavelength is assigned to the management channel. Add / drop multiplexer according to claim 5, characterized in that the signal of the management channel in each case by means of a separate optical filter decoupled from the optical path and into it is coupled, which with the multiplex port of the optical group wavelength multiplexer / Demultiplexer units is connected. Add / drop multiplexer according to one of claims 4 to 6, characterized in that the carrier signal of the management channel continuously transmitted to the adjacent add / drop multiplexer becomes. Add / drop multiplexer according to one of claims 4 to 7, characterized in that two transceiver units for implementation a management channel between the respective add / drop multiplexer and the add / drop multiplexers adjacent to the east and west are provided. Self-healing ring structure for optical communication in wavelength division multiplex, a) where each ring node goes from an add / drop multiplexer to a of the preceding claims is educated b) wherein each add / drop multiplexer over at least an optical fiber with one on the east and one on the west neighboring add / drop multiplexer connected is. Ring structure according to claim 9, characterized in that the control unit of each add / drop multiplexer has a fault in the east and west transmission line recorded and this information directly or via a central network management control unit transmits to all the other add / drop multiplexers and that the control unit Each add / drop multiplexer uses this information to control the at least one group switch unit used to predetermined Communication connections between the relevant add / drop multiplexer and to enable at least one further add / drop multiplexer. Ring structure according to claim 10, characterized in that the control unit of each add / drop multiplexer by means of the management channel and internet protocol (IP) with the control units of all other add / drop multiplexers and / or the network management control unit is connected. Ring structure according to claim 10 or 11, characterized in that the control unit has the information about the functionality the east and west transmission line in two Enter a bit of a link vector and the link vector to the rest Add / drop multiplexers and / or the network management control unit transmits. Ring structure according to claim 12, characterized in that the reading or writing of the link vector between the Layer 3 and the MAC layer 2 takes place, the control unit the link vector into a proprietary Between two MAC frames.