Transcript WDM

Introduction
Chapter 1
1
Outlines
•
•
•
•
•
•
What is a WDM-enabled optical network?
Why IP over WDM?
What is IP over WDM?
Next-generation Internet
IP/WDM standardisation
Summary and subject overview
2
1.1 What is a WDM-enabled Optical
Network?
• Conventional copper cables can only provide a
bandwidth of 100 Mbps (106) over a 1 Km distance
before signal regeneration is required.
• In contrast, an optical fibre using wavelength division
multiplexing (WDM) technology can support a
number of wavelength channels, each of which can
support a connection rate of 10 Gbps (109).
• Long-reach WDM transmitters and receivers can
deliver good quality optical signals without
regeneration over a distance of several tens of
kilometres. Hence, optical fibre can easily offer
bandwidths of tens of Tbps (1012).
3
What is a WDM-enabled Optical
Network?
• In addition to high bandwidth, fibre, made of glass (which is in
turn made mainly from silica sand), is cheaper than other
conventional transmission mediums such as coaxial cables.
• Glass fibre transmission has low attenuation.
• Fibre also has the advantage of not being affected by
electromagnetic interference and power surges or failures.
• In terms of installation, fibre is thin and lightweight, so it is
easy to operate.
• An existing copper-based transmission infrastructure can be
(and has been) replaced with fibre cables.
4
What is a WDM-enabled Optical
Network?
• In the fibre infrastructure, WDM is considered
as a parallel transmission technology to exploit
the fibre bandwidth using non-overlapping
wavelength channels.
• An individual optical transmission system
consists of three components:
– the optical transmitter
– the transmission medium
– the optical receiver.
5
WDM
• The transmitter uses a pulse of light to indicate the ‘1’ bit and the
absence of light to represent the ‘0’ bit.
• The receiver can generate an electrical pulse once light is detected.
• A single-mode fibre transmission requires the light to propagate in a
straight line along the centre of the fibre.
– a good quality signal,
– used for long-distance transmission.
• multimode fibre
– A light ray may enter the fibre at a particular angle and go through
the fibre through internal reflections.
– The basic optical transmission system is used in an optical network,
which can be a local access network (LAN),
– a metropolitan local exchange network (MAN), or a longhaul interexchange network (also known as Wide Area Network, WAN).
6
TDM vs. WDM
• There is a continuous demand for bandwidth in the construction of the
Internet.
• It is also relatively expensive to lay new fibres and furthermore to
maintain them.
• Time Division Multiplexing (TDM)
– is achieved through multiplexing many lower speed data streams
into a higher speed stream at a higher bit rate by means of
nonoverlapping time slots allocated to the original data streams.
• Wavelength Division Multiplexing (WDM)
– is used to transmit data simultaneouslyat multiple carrier
wavelengths through a single fibre, which is analogous to using
Frequency Division Multiplexing (FDM) to carry multiple radio
and TV channels over air or cable.
• TDM and WDM can be used together in such a way that TDM
provides time-sharing of a wavelength channel, for example, through
aggregating access network traffic for backbone network transport.
7
TDM vs. WDM
8
Cost of WDM
• Splitting the useable wavelength bandwidth into a
number of slots (wavelength channels) not only
demands sophisticated equipment but also increases
the likelihood of inter-channel interference.
• As such, WDM equipment cost may dominate the
total cost in a LAN and MAN environment.
• As of November 2001, commercial WDM optical
switches are able to support 256 wavelength channels,
each of which can support a data rate of OC-192 (10
Gbps).
9
A point-to-point WDM-enabled optical
transmission
• Wavelength Add/Drop Multiplexer (WADM)
• Wavelength Amplifier (WAMP)
10
Double-ring WDM network
• A WDM-enabled optical network employs point-topoint WDM transmission systems and requires
– Wavelength Selective Crossconnect (WSXC) that is
able to switch the incoming signal unto a different fibre
possibly a different optical frequency.
11
1.1.2 WDM Optical Network
Evolution
• The first generation of WDM provides
–
–
–
–
only point-to-point physical links that are confined to WAN trunks.
configurations are either static or use manual configurations.
only supports relatively low-speed end-to-end connectivity.
The technical issues include design and development of WDM lasers
and amplifiers, and static wavelength routing and medium access
protocols.
• The WADM can also be deployed in MANs,
• To interconnect WADM rings, Digital Cross Connects (DCX)
are introduced to provide narrowband and broadband
connections.
• Generally these systems are used to manage voice switching
trunks and T1 links.
12
1.1.2 WDM Optical Network
Evolution
• The second generation of WDM
– is capable of establishing connection-orientated end-to-end lightpaths
in the optical layer by introducing WSXC.
– The lightpaths forma virtual topology over the physical fibre topology.
– The virtual wavelength topology can be reconfigured dynamically in
response to traffic changes and/or network planning.
• The technical issues include
– the introduction of wavelength add/drop and cross-connect devices,
– wavelength conversion capability at cross-connects, and
– dynamic routing and wavelength assignment.
• network architecture begins to receive attention,
• In particular the interface for interconnection with other
networks.
• Their cost efficiency in long-haul networks has been widely
accepted.
13
WDM Optical Network Evolution
• The third-generation of WDM
– offers a connectionless packet-switched optical network, in which
optical headers or labels are attached to the data, transmitted with the
payload, and processed at each WDM optical switch. Based on the ratio
of packet header processing time to packet transmission cost, switched
WDM can be efficiently implemented using label switching or burst
switching.
– Pure photonic packet switching in all optical networks is still under
research.
• The bufferless, all-optical packet router brings a new set of
technical issues for network planning:
–
–
–
–
–
contention resolution;
traffic engineering;
over-provisioning;
over-subscription;
interoperability with conventional IP (Internet Protocol) routers
destination-based routing).
14
WDM Optical Network Evolution
• Examples of third-generation WDM devices are:
– optical label switch routers
– optical Gigabit routers
– fast optical switches.
• Interoperability between WDM networks and IP networks
becomes a major concern in third-generation WDM networks.
• Integrated routing and wavelength assignment based on
MultiProtocol Label Switching (MPLS), also known as
Generalised MPLs (GMPLS), starts to emerge.
• bandwidth management, path reconfiguration and restoration,
and Quality of Service (QoS) support.
15
Evolution
16
Switching method
• Optical circuit switching
– is used for large-sized aggregated traffic (such as data trunks), so that
once circuits are set up, the formed topology does not change often.
– This provides cost-efficiency in the long haul network because a few
add/drop points are needed by the traffic and only physical transport
link services are required.
• Optical packet switching
– is used for small-sized data packets. It offers efficient, flexible resource
sharing by introducing complexity to the control system.
• Optical burst switching
– A compromise between packet and circuit switching,
– which switches traffic bursts over a packet-orientated network.
17
cut-through paths
– Packet switching offers cut-through paths, known as ‘layer
2 switching’.
– The cut-through paths lower the network latency by
avoiding intermediate node layer 3 functions.
– A packet routed network employs a store-and-forward
paradigm, where each node maintains a routing table and a
forwarding table.
– Once a packet arrives at the node, by comparing the packet
header with the local forwarding table, the packet is sent to
the next hop on the routing path.
18
1.2 Why IP over WDM?
• IP provides the only convergence layer in the global and
ubiquitous Internet.
• IP, a layer 3 protocol, is designed to address network level
interoperability and routing over different subnets with
different layer 2 technologies.
• Above the IP layer, there are a great variety of IP-based
services and appliances that are still evolving from its infancy.
• Hence, the inevitable dominance of IP traffic makes apparent
the engineering practices that the network infrastructure should
be optimised for IP.
• Below the IP layer, optical fibre using WDM is the most
promising wireline technology, offering an enormous network
capacity required to sustain the continuous Internet growth.
19
WDM
• WDM-based optical networks have been deployed not only in the
backbone but also in metro, regional, and access networks.
• In addition, WDM optical networks are no longer just point-to-point pipes
providing physical link services, but blend well with any new level of
network flexibility requirements.
• The control plane is responsible for transporting control messages to
exchange reachability and availability information and computing and
setting up the data forwarding paths.
• The data plane is responsible for the transmission of user and
• application traffic. An example function of the data plane is packet
buffering and forwarding.
• IP does not separate the data plane from the control plane, and this in turn
requires QoS mechanisms at routers to differentiate control messages from
data packets.
20
IP over WDM
• A conventional WDM network control system uses a separate control
channel, also known as a data communication network (DCN), for
transporting control messages.
• A conventional WDM network control and management system, e.g.
• according to the TMN framework, is implemented in a centralised fashion.
• To address scalability, these systems employ a management hierarchy.
• Combining IP and WDM means, in the data plane, one can assign WDM
optical network resources to forward IP traffic efficiently, and in the control
plane, one can construct a unified
• control plane, presumably IP-centric, across IP and WDM networks. IP
over WDM will also address all levels of interoperability issues on intraand inter-WDM optical networks and IP networks.
21
The motivation behind IP over WDM
• WDM optical networks can address the continuous growth of the Internet
traffic by exploiting the existing fibre infrastructure. The use of WDM
technology can significantly increase the use of the fibre bandwidth.
• Most of the data traffic across networks is IP. Nearly all the end-user data
application uses IP. Conventional voice traffic can also be packetised with
voice-over-IP techniques.
• IP/WDM inherits the flexibility and the adaptability offered in the IP
control protocols.
• IP/WDM can achieve or aims to achieve dynamic on-demand bandwidth
allocation (or real-time provisioning) in optical networks.
– By developing the conventional, centralised controlled optical networks into a
distributed, self-controlled network, the integrated IP/WDM network can not
only reduce the network operation cost, but can also provide dynamic resource
allocation and on-demand service provisioning.
22
The motivation behind IP over WDM
• IP/WDM hopes to address WDM or optical Network Element
(NE) vendor interoperability and service interoperability with
the help of IP protocols.
• IP/WDM can achieve dynamic restoration by leveraging the
distributed control mechanisms implemented in the network.
• From a service point of view, IP/WDM networks can take
advantage of the QoS frameworks, models, policies, and
mechanisms proposed for and developed in the IP network.
• Given the lessons learned from IP and ATM integration, IP and
WDM need a closer integration for efficiency and flexibility.
For example, classical IP over ATM is static and complex, and
IP to ATM address resolution is mandatory to translate
between IP addresses and ATM addresses.
23
What is IP over WDM?
• IP/WDM network is designated to transmit IP traffic
in a WDM-enabled optical network to leverage both
IP universal connectivity and massive WDM
bandwidth capacity.
• IP, as a network layer technology, relies on a data link
layer to provide:
– framing (such as in SONET or Ethernet);
– error detection (such as cyclic redundancy check, CRC);
– error recovery (such as automatic repeat request, ARQ).
24
All-optical network
• An objective of optical networking is to provide optical transparency frome
nd to end so that the network latency is minimised.
• This requires all-optical interfaces and all-optical switching fabric for the
edge and intermediate network elements.
• Transponders are used to strengthen the optical signal.
– all-optical transponders (tunable lasers) and
– Optical-Electrical-Optical (O-E-O) transponders.
• The figure shows two types of traffic, IP (e.g. Gigabit Ethernet) and
SONET/SDH, which in turn requires Gigabit Ethernet and SONET/SDH
interfaces.
• In the case of multiple access links, a sublayer of the data link layer is the
Media Access Protocol (MAC) that mediates access to a shared link so that
all nodes eventually have a chance to transmit their data.
• The definition of a protocol model to efficiently and effectively implement
an IP/WDM network is still an active research area.
25
26
Possible approaches for IP over WDM
27
IP/ATM/SONET/WDM
• transports IP over ATM (Asynchronous Transfer Mode), then
over SONET/SDH and WDM fibre.
• WDM is employed as a physical layer parallel transmission
technology.
• The main advantage of this approach by using ATM is
– to be able to carry different types of traffic onto the same pipe with
different QoS requirements.
– its traffic engineering capability and the flexibility in network
provisioning, which complements the conventional IP best effort traffic
routing.
• Disadvantage
– offset by complexity, as IP over ATM is more complex to manage and
control than an IP-leased line network.
28
IP/ATM
• ATM uses a cell switching technology. Each ATM cell has a
fixed 53-byte (5-byte header and 48-byte user data) length, so
application traffic has to be packetised into cells for transport
and reassembled at destination.
• ATM cell packetisation is the responsibility of the ATM SAR
(Segmentation and Reassembly) sublayer.
• SAR becomes technically difficult above OC-48.
• Having an ATM circuit layer between IP packet and the WDM
circuit seems superfluous.
• The statement is strengthened by the emergence of the MPLS
technique of the IP layer.
29
key features of MPLS
• Use of a simple, fixed-length label to identify flows/paths.
• Separating control from data forwarding, control is used to set
up the initial path, but packets are shipped to next hop
according to the label in the forwarding table.
• A simplified and unified forwarding paradigm, IP headers are
processed and examined only at the edge of MPLS networks
and then MPLS packets are forwarded according to the ‘label’
(instead of analysing the encapsulated IP packet header).
• MPLS provides multiservice. For example, a Virtual Private
Network (VPN) set up by MPLS has a specific level of
priority indicated by the Forwarding Equivalence Class (FEC).
30
key features of MPLS (count.)
• Classification of packets is policy-based, with packets being
aggregated into FEC by the use of a label. The packet-to-FEC
mapping is conducted at the edge, for example, based on the
class of service or the destination address in the packet header.
• Providing enabling mechanisms for traffic engineering, which
can be employed to balance the link load by monitoring traffic
and making flow adjustments actively or proactively. In the
current IP network, traffic engineering is difficult if not
impossible because traffic redirection is not effective by
indirect routing adjustment and it may cause more congestion
elsewhere in the network. MPLS provides explicit path routing
so it is highly focused and offers class-based forwarding. In
addition to explicit path routing, MPLS offers tools of
tunneling, loop prevention and avoidance, and streams
merging for traffic control.
31
IP/SONET/WDM
• IP/MPLS over SONET/SDH and WDM.
• SONET/SDH provides several attractive
features to this approach:
– SONET provides a standard optical signal
multiplexing hierarchy by which low-speed signals
can be multiplexed into high-speed signals.
– SONET provides a transmission frame standard.
– the SONET network protection/restoration
capability, which is completely transparent to
upper layers such as the IP layer.
32
IP/SONET/WDM
• SONET networks usually employ a ring topology. SONET
protection scheme can be provided:
– as 1 + 1 meaning data are transferred in two paths in the opposite
direction and the better signal is selected at the destination;
– as 1:1 indicating there is a separate signalled protection path for the
primary path;
– or as n:1 representing where primary paths share the same protection
path.
• The design of SONET also enhances OAM&P (Operations,
Administration, Maintenance, and Provisioning) to
communicate alarms, controls, and performance information at
both system and network levels.
33
SONET/SDH
• However, SONET carries substantial overhead
information, which is encoded in several levels.
– Path overhead (POH) is carried from end-to-end.
– Line overhead (LOH) is used for the signal between the
line terminating equipment, such as OC-n multiplexers.
– Section overhead (SOH) is used for communication
between adjacent network elements, such as regenerators.
• For an OC-1 pipe with 51.84 Mbps transmission rate,
its payload has the capacity to transport a DS-3 with
44.736 Mbps digital bit rate.
34
3G
• IP/MPLS directly over WDM,
• the most efficient solution among the possible approaches.
• It requires that the IP layer looks after path protection and
restoration.
• It also needs a simplified framing format for transmission error
handling.
– Several companies are developing a new framing standard known as
Slim SONET/SDH, which provides similar functionality as in
SONET/SDH but with modern techniques for header placement and
matching frame size to packet size.
– adopt the Gigabit Ethernet framing format. The new 10-Gigabit
Ethernet is especially designed for dense WDM systems. Using the
Ethernet frame format, hosts (Ethernet) on either side of the connection
do not need to map to another protocol format (e.g. ATM) for
transmission.
35
Signal
• Conventional IP networks use in-band signalling so
data and control traffic is transported together over
the same link and path.
• A WDM optical network has a separate data
communication network for control messages. Hence,
it uses out-of band signalling.
• In the control plane, IP over WDM can support
several networking architectures, but the architecture
selection is subject to constraints on existing network
environments, administrative authority, and network
ownership.
36
Signalling
37
1.4 Next-generation Internet
• US Internet-related research and development partnerships
include not only entities that are directly focused on Internet
development such as IETF but also general standard
organisations such as IEEE and ANSI and federal government
agencies such as DARPA (Defense Advanced Research
Projects Agency) (www.darpa. mil) and NSF (National
Science Foundation) (www.nsf.gov).
• The Next Generation Internet (NGI) initiative (www.ngi.gov)
was established in 1998 for the period of 5 years, through
which government agencies will cooperate to create next
generation Internet capabilities to allow for enhanced support
for their core missions, as well as to advance the state-of-theart in advanced networking.
38
NGI initiative will
• develop new and more capable networking
technologies to support Federal agency missions;
• create a foundation for more powerful and versatile
networks in the 21st century;
• Form partnerships with academia and industry that
will keep the US at the cutting edge of information
and communication technologies;
• enable the introduction of new networking services
that will benefit businesses, schools, and homes.
39
NGI goals
• conducting research in advanced end-to-end
networking technologies, including differentiated
services, particularly for digital media, network
management, reliability, robustness, and security;
• prototyping and deploying national-scale testbeds that
are able to provide 100 to 1000 times current Internet
performance;
• developing revolutionary new applications requiring
high performance networks.
40
SuperNet testbed
41
SuperNet streamline networking
protocol stacks
42
CANARIE
• the Canadian Network for the Advancement of
Research, Industry, and Education, is a non-profit
corporation supported by its members, project
partners and the Canadian government to accelerate
Canada’s advanced Internet development and
facilitate the widespread adoption of faster, more
efficient networks and enable the next generation of
advanced products, applications and services.
• In February 1998, the Canadian government provided
CANARIE with a $55 million grant towards the
$120 million project to develop a national optical
R&D Internet, known as CA*net III. Industry
members provided the remaining part of the funding.
43
Canada
• Using new fibre optic-based technology and Dense
Wavelength Division Multiplexing (DWDM),
CA*net 3 intends to deliver unrivalled network
capability with a potential for OC768 (40 Gbps) to
Canadian research institutions and universities.
• Phase I was completed in October 1998, where an
optical Internet backbone was set up between Toronto,
Ottawa and Montreal. Currently Phase II is in
progress,
• through which the optical Internet Backbone will be
extended west from Toronto to Vancouver and east
from Montreal to Atlantic Canada.
44
European &Asia
• European ACTS (Advanced Communications
Technologies and Services) program.
• In addition, many European NRNs (National
Research Networks) have established national
high-performance advanced network
infrastructures.
• In Asia Pacific, a number of countries have
participated in the APAN (Asia Pacific
Advanced Networking) initiative.
45
1.5 IP/WDM Standardisation
• the Internet Engineering Task Force (IETF)
(www.ietf.org) and
• the International Telecommunication Union,
Telecommunication Standardisation Sector
(ITU-T) (www.itu.org) respectively.
46
• In particular, IETF has been focusing on these IP/WDMrelated issues:
– MPLS/MPl S (Multiprotocol Lambda Switching)/GMPLS (Generalized
MPLS).
– Layer 2 and layer 3 functionalities within optical networks.
– NNI (Network to Network Interface) standard for optical network.
47
ITU-T
• In particular, ITU-T has been focusing on these
IP/WDM related issues:
– Layer 1 features of the OSI model.
– Architectures and protocols for next-generation
optical networks, also known as optical transport
network (OTN), defined in G.872.
– Architecture for the automatic switched optical
network, defined in G.ason.
48