Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Look up keyword
Like this
1Activity
0 of .
Results for:
No results containing your search query
P. 1
Modeling GMPLS Paper 2

Modeling GMPLS Paper 2

Ratings: (0)|Views: 10|Likes:
Published by Yassir Gr

More info:

Published by: Yassir Gr on Feb 03, 2012
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

02/03/2012

pdf

text

original

 
 
1
Modeling GMPLS domains in MPLS networks
 
Henrik Christiansen
&20%OGJ97HFKQLFDO8QLYHUVLW\RI'HQPDUN .JV/\QJE\'HQPDUN(PDLOKF#FRPGWXGN
 
Abstract
A consequence of migrating the existing Internet architecture toan all-optical one is that the network will consist of a mixture of equipment, ranging from electrical routers to all-optical packetswitches. Hence, future networks will consist of multipledomains employing different technologies. The MPLS conceptis attractive because it can work as a unifying control structurecovering all technologies. This paper describes how opticalcircuit switched, GMPLS-based networks can be incorporated insuch multi-domain, MPLS-based scenarios and how it could bemodeled with the OPNET MPLS model. GMPLS nodes areimplemented and routing and path set up by GMPLS signaling isdemonstrated
.
Introduction
In the old days, the vision was to create one single technologyfor multi service networks. This was one of the drivers behinddeveloping and deploying ATM. However, the technologiesbeing developed today are of a different nature. It is no longerlikely with a network based on one single technology, simplybecause the vast amount of equipment in e.g., the global Internetmakes instant upgrade/replacement impossible. Migration tofuture technologies will be seen as islands popping up and thisgradual upgrade creates heterogeneous networks consisting of anumber of different technologies. Currently, for instance, opticaltechnologies are being introduced into the networks, butelectrical routers/switches are still present. Thus, the networks of the future will be multi technology, multi service networks. Addto that the requirements of traffic engineering capabilities andyou will end up with a very complex network.
Figure 1: A multi-domain network comprising differenttechnologies
This have had an impact on the structure of modern networks,but also this has created a requirement for special adaptationdevices that are able to propagate traffic between network domains running different technologies and for a commoncontrol plane structure able to unify all these technologies andcreate a useful network (See figure 1) A closer look at theadaptation devices can be found in [Chr2001]. In this paper theemphasis is on the control part of the network.This paper is organized as follows. Firstly, a brief MPLS tutorialis provided, where after a short comparison of packet andwavelength switch leads us the way to GMPLS. Then theintegration of those technologies is treated and it is describedhow to model these combined MPLS / GMPLS networks. TheGMPLS OPNET model is then presented along with somesimulation results that verify the functionality and illustrate howthe GMPLS model interoperates with the OPNET MPLS model.
MPLS
MPLS [Ros2001a] is a networking concept that is based mainlyon a shift of all complex functionality to the edge of the network,leaving only simple operation for the core network and henceenabling fast and efficient operation. The control plane (thattakes care of e.g., routing) and switching (packet forwarding) arecompletely decoupled, which yields the advantageous propertythat they can be chosen independently. This is the main reasonwhy we in this paper can consider routing and structural issueswithout treating e.g., packet forwarding explicitly. MPLS isdesigned as a pure ’everything over everything’ concept, henceits name. In reality, however, its predominant use and themajority of standardization work are focused on carrying IPtraffic with MPLS, which is due to the importance of theubiquitous Internet.In MPLS packets are forwarded along routes called LabelSwitched Paths (LSPs) that may be determined by routingprotocols based on predefined traffic classes called ForwardEquivalent Classes (FECs). An FEC can be equivalent to asingle entry in a conventional IP routing table or it can be anaggregation of multiple entries. An FEC can also be specifiedbased on a number of additional constraints such as originatingaddress, receiving port number and QoS parameters. These LSPsare defined in the switches by using labels, which are distributedby a Label Distribution Protocol (LDP) responsible for mappingbetween routing and switching. The MPLS standard doesn’tspecify one specific label distribution protocol; it just highlightsthe required properties. Currently, four protocols of which twoare new and two are modifications of existing protocols arementioned in the standards[And2001][Rekh2000][Jamo1999][Brad1997].One of the major benefits of the MPLS concept is its ability toperform traffic engineering, i.e., to be able to control how trafficflows through the network, which is one of the prerequisites forproviding QoS guarantees on connections. Generally, trafficengineering implies to route along non-shortest paths andutilizes Constraint Based Routing (CBR) where the routes are
 
 
2
calculated subject to performance- and administrativeconstraints, which are assigned by the network managementsystem, based on e.g., traffic measurements.In MPLS, switches are generally called Label Switch Routers(LSRs). Ingress edge LSRs take care of attaching short, fixedlength labels to packets when they enter the MPLS domain,which includes the non-trivial task of determining to which FECa given packet belongs. Within the core of the network forwarding will be based on the label only, and before leavingthe MPLS domain packets have their label removed by theegress edge LSR (see figure 2).
Figure 2: The label is used only within one MPLS domain.By attaching different labels at the ingress LSR, differentroutes through the network for the same destination can beselected, which allows for traffic engineering.
The labels are generally not kept constant along a LSP and thusa path through the network is defined by a sequence of labels, allof which are assigned by the LDP. Within the core switches onlythe labels are examined, and what distinguishes this methodfrom that of conventional IP routing are the loose couplingbetween the label and the destination address as well as thelookup scheme within the switches themselves. The labels usedby MPLS require exact match in the lookup tables, which is amuch simpler operation than LPM [Rekh1995]. I.e., OSPFwould build a routing table is each LSR and based on thisinformation and possibly additional information the labeldistribution protocol builds another table in which the
label
isused as the key. The outcome of a table lookup is informationabout outgoing port number and the outgoing label, which isused to replace the label contained within the packet as well asexpediting the packet to the designated output port. The labelreplacement operation is usually called
label swapping
and is themost common packet modification operation in MPLS. Inaddition, when working with multiple domains in a network, thesingle label might be replaced by a stack of labels with only thetop label being used within one particular domain. At domainboundaries label swapping is insufficient and must be exchangedfor more complex operations such as label pushing and popping.
Packet and Wavelength switching
MPLS was designed for packet switched networks. However,when considering all-optical devices, packet switching is not yeta mature technology. The main difference between electrical andoptical packet switching is in the data path where the opticalpacket switch matrix operates on purely optical signals andtherefore is capable of switching at very high bit rates[Dan1997][Hun2000][Chi1998]. The optical switches canpotentially be fully transparent (with respect to bit rate andpayload) but 2R or 3R regeneration is required when cascadingseveral switches [Wol1999]. To resolve contention on the outputports, buffering is required, but the only optical bufferingavailable today is fiber delay lines, which control the delay of the optical packets. Optical buffer space is bulky and hence verylimited compared to memory sizes in conventional, electricalswitches. In some optical buffer designs the wavelength domainis exploited for contention resolution, thus increasing theeffective buffer size [Dan1997].In optical switches, electronic circuits control the packet headeranalysis and switch-matrix configuration so the switchthroughput, measured in packets per second, is normally limitedby packet processing and reconfiguration times. To summarizethe properties of electrical and optical switching the followingshould be emphasized:
The packet length measured in bytes is longer for the opticalpacket than for the electrical packet, which means thatseveral electrical packets must be bundled into one opticalpacket. Hence, when mixing electrical and optical packetswitches considerations on traffic aggregation are called for.
Optical buffer sizes are considerably smaller than theirelectrical counterparts, which mandate shaping of the trafficto avoid buffer overflow.
GMPLS
GMPLS is a generalization of MPLS that allows a seamlessintegration of a multitude of technologies, especially circuitswitched systems, with packet switched networks. Thus,interfacing with traditional telecom TDM systems (e.g. SONET SDH) and wavelength routed optical networks is possible withthe use of GMPLS. GMPLS is in widespread use and have beenimplemented by several manufacturers [Ber2002].
Figure 3: GMPLS in a typical usage scenario where GMPLS
is used as ‘islands’ in the network.
Figure 3 depicts a likely usage scenario for GMPLS
– GMPLS isforming ‘islands’ within an MPLS network. It is exactly themodeling of this kind of network that this paper covers.
 
 
3
 
Optical wavelength switching (or circuit switching) on the otherhand is now becoming available and is an attractive alternativein high capacity backbone networks. By using mixed-technology, multi domain networks the advantages of differenttechnologies can be combined. What is lacking is a unifiedcontrol system, which is exactly what GMPLS provides. I.e., theintegration of MPLS and GMPLS with these circuit-switchedsystems is advantageous because is offers:
Traffic engineering capabilities,
High capacity core
Flexible, controllable edge
Protocol independence (i.e., e.g. IP interoperability)
Modeling GMPLS
Real GMPLS networks are highly complex and may coverdevices such as optical wavelength switches and SONETnetwork nodes, i.e. GMPLS can operate with as well electronicas optical technologies. Hence, GMPLS networks can get verycomplex since a multitude of technologies are hidden there,implying a vast number of protocols, devices and configurationoptions.The real-life network must be simplified greatly in order to beable to build a model that can produce results within anacceptable timeframe. A brute-force modeling methodology that just tries to model the real network in every detail isinappropriate. Below the goals for the simulation are identifiedand based on that the simplified simulation model can be set up.Obviously, the model must be simple enough to achieve theidentified goals, while representing a fair model of the realnetwork.
Requirements to the model
The goal of this simulation study is to build a model of howGMPSL interacts with an MPL S based network. With the modelit should be possible to measure/study:
Call setup probability
Optical signal quality
Network topology / routing issuesA list of input parameters is provided below:
Attribute Description
Topology generation parameters- Number of nodes- Number of links- Maximum distanceSize and connectivity of thenetwork Path constraints Bandwidth constraintsType of network SONET / pure optical
OPNET implementation
The GMPLS implementation has been made with OPNETmodeler 8.0 and the MPLS model suite. The MPLS model hasbeen extended/modified in order to create a GMPLS network element that can be built into MPLS network. This GMPLSmodels element represents the entire GMPLS network, i.e. acomplete topology can be built with this single node. Figure 4illustrates how the GMPLS network can interoperate with MPLSdevices, i.e., LSPs can be setup through the GMPLS domain inthis mixed environment. .
Figure 4: A GMPLS model, which can interoperate withMPLS, has been built into OPNET.
A number of modification to the OPNET MPLS models areneeded. As well the user as the control plane need to bemodified.Modification needed:
control plane (path/LSP setup)
o
Interacts with the label distribution protocol
o
Setup / tear down of connections
o
Path constraints handling
data plane (packet forwarding)
o
Interacts with the label swap/push/pop operations
o
Delay
o
Packet loss
o
Signal degradation (for all-optical networks)
 
 More details of the implemented model.
In order to minimize the modifications needed in the OPNETcode, GMPLS has been implemented as a separate processwithin the network nodes. The LDP process has then just beenmodified to detect whether this GMPLS process is present or not(and hence whether this is a MPLS or GMPLS node)
Figure 5: GMPLS has been implemented as a separateprocess in the MPLS node model
The details of the process model is shown below (figure 6)

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->