GB2465589A

GB2465589A - Integrated Multilayer Network Restoration using Hybrid Network Elements.

Info

Publication number: GB2465589A
Application number: GB0821321A
Authority: GB
Inventors: George Tsirakakis
Original assignee: Individual
Current assignee: Individual
Priority date: 2008-11-24
Filing date: 2008-11-24
Publication date: 2010-05-26
Also published as: GB0821321D0

Abstract

A hybrid network element for establishing ports/interfaces from one layer (A) to another layer (B) during fault restoration, in order to re-route one (A) layer's failed traffic over these new interfaces to free bandwidth over Layer B, and back to Layer A again. Hybrid network elements are network elements having functionalities of multiple layers and they can process channel connections of multiple layers at the same time. They can be used to join the segmented free bandwidth scattered across layers by creating interfaces (interconnections) between layers dynamically. Therefore a complete connection could be formed to reroute some failed bandwidth. Multilayer network architectures are very common today. Each layer (IP, ATM, SDH, WDM etc) does not have knowledge and can not see the channel connections and the free bandwidth in them in the other layers. The result of this is that during fault restoration segmented free bandwidth may exist across different layers which could not be utilised. We are proposing the use of hybrid network elements for more efficient fault restoration.

Description

Integrated Multilayer Network Restoration using Hybrid Network Elements

Background

Nowadays mullipile communication networks exist, most often one using the connections of the other. For example IF on SDH, or SDH over WDM.

Each of these layers is defined by the nodes (of certain technology) on this layer (e.g. IF router, SDH DCS, WDM crossconnect) and the native connections between them.

Each layer's connections can not be seen it their content by other layers. For example an SDH DCS can not see what is inside a VC3, an WDM crossconnect can not see the VC3s in a wavelength. As a result the SDH DCS can not see if there is free bandwidth within a VC3 carrying IP or ATM traffic. Also the IP router does not know how many free VC3s exist on some physical connection nearby. In general, in a multilayer network we have native bandwidth channel pipes which themselves have inside, other technology's native bandwidth channel pipes and so on, each pipe being separate and isolate from the other. There is a segmentation of bandwidth, resulting to bandwidth inefficiencies. This is also seen on fault restoration.

When we have a fault restoration, each layer acts independently seeing the other as static, without knowing about the free bandwidth on the other layer and without being able to use it, if needed. In addition each layer has its own restoration scheme.

The above statements, describing the isolation of bandwidth channels, are withdrawn with the introduction of hybrid network elements which has functions of different layers. For example an IP/SDH hybrid could crossconnect SDH channels and at the same time process IF packets. It is expected that new hybrid element are going to appear in the future integrating other type of layers eg. SDH & WDM.

Statement of invention

We are proposing here a new scheme which uses hybrid nodes like these, for achieving better results during fault restoration. Because hybrid network element is one unit existing in two layers, it can see and access information and channels of both layers at the same time. In this way free bandwidth of one layer can he utilized if it is considered useful for restoration by the other layer.

How this could be done is described with reference to the accompanying figures in which: Fig 1: Explains how the bandwidth could be ananged with in a link (e.g. a fiber).

Fig 2: Describes the architecture of a hybrid network element Fig 3: Describes how the hybrid could help in rerouting the failed traffic.

Detailed description

Fig 1 gives an example of how bandwidth is channelized within a link, having isolated channels within other channels.

Reference 1 shows a Layer 1 channel carrying some traffic.

Reference 2 shows a Layer 1 channel carrying Layer 2 traffic.

Reference 3 shows a Free Layer 1 channel.

Reference 4 shows a Layer 2 channel Reference 5 shows a Layer 2 channel Reference 6 shows Free Layer 2 bandwidth As an example, Layer 1 could be SDH or WDM channels while Layer 2 could be e.g. IF.

Channels in layer 1 and 2 are isolate and separate normally. The free layer 1 channel is not accessible by layer 2 network elements, and vice-versa layer 2 free bandwidth is not accessible by layer 1 network elements.

Fig 2 describes a hybrid network element. Every input and output on the hybrid element could be a structure as that on Fig. 1.

OnFig2: Reference 1 shows an incoming connection having Layer 2 traffic in a Layer 1 channel (similar to that of Reference 2 of Fig 1).

Reference 2 shows an incoming connection having some other traffic in a Layer 1 channel (similar to that of Reference 1 of Fig 1).

Reference 3 shows an outgoing connection having Layer 2 traffic in a Layer 1 channel Reference 4 shows an outgoing connection having some other traffic in a Layer 1 channel Reference 5 shows the Layer 2 Functions Unit Reference 6 shows the Layer 1 Functions Unit As an example of a hybrid network element, an TP/SDH hybrid node is described in European patent number EP1701495 (2006) Hybrid digital cross-connect for switching circuit and packet based data traffic, https://publications.european-patent-office.org/PublicationServer/getpdf. jsp?cc=EP&pn=1701495&ki=A1 At Fig 2 it is seen that the hybrid network element has two elements / function units. One is the functions of layer 2 e.g. IP i.e. routing, the other is functions of layer 1 e.g. SDH i.e. TDM crossconnect. The ratio of how much traffic of all of its inputs is given to layer 2 varies, and it could be that 100% or 0% of traffic is given to layer 2 functions. The vertical lines between Layer 2 & Layer 1 represent ports/interfaces between the two layers. These could be created on demand as necessary.

Traffic of layer 2 (Reference 1) maybe crossconnected only in layer 1 by the Layer 1 Functions Unit (the complete Layer 1 channel as a single unit) without given to Layer 2 Functions Unit, or it could be given to Layer 2 Functions Unit to be crossconnected in Layer 2 (crossconnect individual Layer 2 channels).

The proposed use of hybrid for restoration Fig 3 describes how the hybrid could help in rerouting the failed traffic.

Connections AC, AB & BC could have similar structure to that of Fig. 1.

Let Li be the technology in layer i, and L2 be the technology in layer 2.

In Fig 3, A, B & C consist of Li and L2 collocated nodes. Connection AB has free Li channels but no free space within any of the Li channels that carry L2 traffic. Connection BC has no free Li channels but has free space in some of the Li channels that carry L2 traffic.

In the simplified case above, the Li recovery scheme will fail as connection BC has no free Li channels.

L2 recovery scheme will also fail as connection AB has no free L2 bandwidth. However there is free bandwidth on the connections.

Restoration of some of the failed L2 traffic on connection AC would be possible if A s Li node, cross-connect the failed AC Li channel carrying L2 traffic, to the free Li channel on connection AB, and B take the L2 traffic on this Li channel and cross-connect it to the free L2 bandwidth/channels on connection BC. This requires B to have the ability to process both L2 & Li traffic and also to create new Li/L2 ports on demand.

This is possible with the functions of the hybrid network elements.

Advantages We have found that using hybrid network elements bandwidth savings up to 20% could be achieved.