What Does a Framer Do?

The OTN electrical cross-connect technology uses ODU k as the granularity for mapping, multiplexing, and crossover, which is similar to the traditional SDH equipment VC crossover. SDH equipment has VC-12 low-order crossover capability and VC-4 high-order crossover capability. Corresponding to this, OTN electrical crossover equipment also introduces high-order / low-order optical channel data units (HO / LO ODU). HO / LO is similar to the concept of high-order / low-order containers in SDH. LO ODU is equivalent to the service layer and is used to adapt services of different rates and different formats. HO ODU is equivalent to the tunnel layer and is used to provide a certain bandwidth transmission capability. This hierarchical structure supports the separation of business cards and line cards, making network deployment more flexible and economical.

Chinese name: OTN optical cross-connect

Foreign name: OTN optical cross connect

Basic principles of OTN optical cross-connect electrical cross-connect technology

The core device of the electrical cross-connect device is a cross-connect matrix, which is used to achieve arbitrary cross-connection between various branches of a certain level in the N input signals.

To understand electrical crossover, you must first understand the concept of time slots. For OTN, it is like a conveyor belt that transfers large cars from one place to another. There are many small boxes in the large compartment as needed, and many smaller boxes in the small ones. The smallest box may be 64k, and upwards are 2M, 155M, 622M, ODU k , and light wavelength. Each box must be marked with the destination. When the box arrives at the destination, it must be unloaded, and then other boxes of the same specification are loaded into the unloaded box and then sent down.

OTN optical cross-connect matrix structure

Cross matrix currently has two commonly used structures, namely a square matrix and a CLOS matrix.

OTN optical cross-connect square matrix

A single-stage switching matrix is also called a square matrix. There are N inputs and N outputs. During the establishment of a connection, there can only be one cross joint action per horizontal row and each vertical row. Single-stage switching matrix, non-blocking, but requires many cross-connects. N × N matrix. The number of cross-connects is N × N. For example, N = 64, the total number of cross-connects is 64 × 64 = 4096. From N equivalent cross-slots to 2 N equivalent cross-slots, the required cross-matrix unit will be increased to 2 = 4 times, and so on, the matrix size will increase squarely.

CLOS OTN optical cross-connect three-level CLOS network

The non-blocking condition was first proposed by CLOS in 1954. Taking a three-level CLOS network as an example, it consists of an input stage, a center stage, and an output stage. The input and output stages perform space division (S) and time division (T) exchanges, while the center stage only performs space division exchanges. The typical configuration form of the matrix is (TS-S-ST). The capacity of the center level is configured according to the planned maximum matrix capacity. When expanding the capacity, it is only necessary to increase the matrix capacity of the input stage and output stage. Compared with the square matrix, the number of cross points that the three-level CLOS matrix needs to control is greatly reduced, which is suitable for the realization of large-capacity cross-connect matrices. When the broadcast service cannot exceed 25%, this matrix is non-blocking for unidirectional and bidirectional connections.

OTN OTN optical cross-connect OTN switching unit technology route analysis

From the perspective of the implementation of the OTN cross-connect equipment, some use dual-core or multi-core technology, and some use single-core technology. Dual-core and multi-core technologies use packet exchange, ODU k exchange, and VC exchange as their respective exchange disks (or chips) for processing. The service disk and the line disk are respectively dispatched to different switching disks (or chips) for switching scheduling according to the classification of signal services. The single-core technology uses a unified switching core to handle packet switching, ODU k switching, and VC switching.

As for the technical orientation of the switching core, there are cell switching and VC crossover. There are several ways to implement the exchange technology.

OTN optical cross-connect cell and packet switching technology

Cell and packet switching is to divide OTN services into cells, use packet switching chips for scheduling, and then resume loading into OTN services. Data packet services can be processed directly with packet switching chips.

OTN optical cross connection space division cross technology

Space division crossover can solve almost all TDM service crossovers, but it is powerless for packet switching. In the air separation switching network, there is an independent physical route for each connection. There is no common point for the routes for different connections, and this route is maintained during each connection establishment. A route is composed of a set of internal links, which are connected to each other through cross-points configured in a matrix. The cross contact has two stable states: open state and closed state. The space division matrix is actually a type of space connector. Usually, the input and output are physically connected at a certain point (or some point) in the form of electricity, so the connection time is extremely short.

Its biggest advantage lies in its business independence, and its biggest problem lies in the limitation of capacity.

OTN optical cross-connect time division technology

The route provided by the time division switching network to the user is only a time slot, that is, only in a given time slot, this route is assigned to a certain pair of users. It may be allocated to other users at other times. The time division matrix is actually a time slot switch, which is usually implemented by random memory. By storing the input signals in different spatial locations, and using the readout control circuit to control the time in different time slots to read out the contents of different columns, the time slot exchange between different input signals is achieved. The delay of this matrix is large, much larger than that of the space division matrix.

There are two main implementation technologies of current cross-chips. One is to adopt the cross-time-space-time-division (TST) structure to implement the cross technology. With this technology, the chip circuit scale is smaller, but the implementation algorithm of the cross connection is more complicated. There are some shortcomings in the implementation speed and reconfigurability of cross-connect. Another structure is a bit-cut-based implementation method that can achieve crossover between any port and any time slot. The implementation algorithm is simple and easy to cascade and expand. The cross connection speed is fast and the reconstruction is strong.

Foreign equipment manufacturers focus on ODU k and VC exchange, while the domestic market pays more attention to ODU k and packet exchange. From the perspective of technology development, compared to SDH switching, packet switching should give more attention and capacity allocation. From the perspective of device vendors and the development of the OIF standard, OTN support for packet functions is likely to be the development direction for a period of time in the future.

OTN OTN optical cross -connect solution

Like SDH equipment, OTN equipment also has a blocking phenomenon. Services are generally encapsulated by ODU k and then enter the cross matrix unit through the backplane bus. If the granularity of the service itself, such as ODU2, the bus bandwidth will not be wasted, but the service When the particles are small, such as ODU1 or ODU0, the bandwidth of the bus carrying a service alone will not be full, the number of services is large, the bus is exhausted, and other services cannot achieve cross-functions.

The traditional OTN equipment processing flow is an asynchronous processing flow. The service signals are mapped by the ODU, enter the cross matrix, and then pass through the ODU framer to become the OTU k optical port, and vice versa. This implementation method is firstly difficult to cross a large-capacity ODU k , and it is difficult to achieve unimpeded crossover for ODU k of all granularities.